Microsoft has announced a "definitive agreement" to acquire artificial intelligence (AI) speech recognition firm Nuance Communications for $19.7 billion (£14.3 billion).

The deal is intended to close later in the year with Nuance boss Mark Benjamin remaining as CEO but reporting to Microsoft's executive vice president of Cloud & AI, Scott Guthrie.

Nuance is an American AI firm based just outside of Boston that sells audio recognition and transcription tools. The company was founded in 1992, has over 7,000 employees and reported $346 million in fourth-quarter revenues. It's thought that Microsoft is attempting to expand via acquisitions, with an interest in Nuance's work in healthcare, customer services and voicemail.

Microsoft and Nuance already have a professional relationship, collaborating on technologies that allow doctors to capture voice conversations and enter them into electronic medical records.

This will be Microsoft's immediate focus once the deal closes, with the company saying that by augmenting the Microsoft Cloud for Healthcare with Nuance's solutions, it will be better able to empower healthcare providers through the power of ambient clinical intelligence and other Microsoft cloud services.

"Nuance provides the AI layer at the healthcare point of delivery and is a pioneer in the real-world application of enterprise AI," said Microsoft CEO Satya Nadella.

"AI is technology's most important priority, and healthcare is its most urgent application. Together, with our partner ecosystem, we will put advanced AI solutions into the hands of professionals everywhere to drive better decision-making and create more meaningful connections, as we accelerate growth of Microsoft Cloud for Healthcare and Nuance."

Nuance also specialises in AI, speech recognition and virtual assistants, areas in which Microsoft already has extensive expertise with its developer tools that enable transcription and functions to incorporate speech recognition into its products. Both Teams and the Bing search engine have some forms of speech recognition and audio transcription, for example.

The tech giant is seemingly in takeover mode with a $7.5 billion deal for video game maker Zenimax completed last month and wide reports that it recently considered buying social media app TikTok. The firm was also said to be chasing an acquisition of gaming chat app Discord, for around $10 billion.

Both Discord and TikTok deals were thought to be more about the large volumes of users data that Microsoft would have access to if the acquisitions were successful. The deal for Nuance is Microsoft's biggest acquisition since buying LinkedIn in 2016.

EasyJet has introduced a new feature to its app that will allow customers to search for flights using voice recognition technology.

The search feature uses AI and Dialogflow, Google Cloud's natural language understanding tool for building conversational experiences, to support voice-based flight queries about destinations, dates and airports.

While typing a search query takes 12 taps, EasyJet said that with the vocal steps in Speak Now it could reduce this process to a few seconds.

Daniel Young, head of digital experience at EasyJet said the airline was aiming to improve the travel experience, including making booking flights more accessible, "especially for those who are visually impaired, and Speak Now helps us achieve this objective."

"We continue to place innovation at the heart of our industry-leading mobile travel app to give our customers the tools they need to have a smooth experience booking flights with us. Embracing this latest technology makes preparing for travel, easy and fun," he said.

Cormac Reilly, VP & global head of Travel Partners Solutions, who developed the technology for EasyJet at Travelport, said that this would make booking travel "as easy as possible, giving EasyJet customers a helpful digital experience".

"Conversational voice search is a great example of how we're using the latest technologies, including AI and Google Cloud's Dialogflow, to make the experience of buying and managing travel continually better for everyone," he added.

The voice search functionality will be available in English on iOS and follows other advances developed by the airline including a visual search tool using Instagram images (Look&Book), and a cabin bag-sizer tool, which includes augmented reality technology to enable passengers to check the size of their hand-luggage before heading to the airport.

The Speak Now feature will be launched on EasyJet's app in the coming weeks.

Hundreds of contractors who were hired to listen to Siri recordings are being removed from their roles after Apple shut down the programme earlier this month.

More than 300 contract workers in Apple's Cork facility were placed on paid leave from 2 August when Apple suspended its 'grading' project pending a review. These workers have now been informed their contracts are being terminated, according to the Guardian.

The controversial programme saw criticism from privacy advocates after it emerged Apple workers were listening to recordings users had made via Apple's flagship voice assistant. This included things like sensitive medical information.

Contractors were hired through third-party firms as part of the grading process to examine audio recordings for errors.

"Our goal with Siri, the pioneering intelligent assistant, is to provide the best experience for our customers while vigilantly protecting their privacy," Apple said in a statement.

"We know that customers have been concerned by recent reports of people listening to audio Siri recordings as part of our Siri quality evaluation process - which we call grading.

"We heard their concerns, immediately suspended human grading of Siri requests and began a thorough review of our practices and policies. We've decided to make some changes to Siri as a result."

Following its review, Apple has decided to resume its grading programme later in the year, but will no longer retain audio recordings of Siri interactions by default. The firm will also allow users to opt-in to allowing humans to examine their interactions. For customers who opt-in, too, only full Apple employees will be allowed to listen to audio samples.

The more than 300 contractors previously hired to work on the programme, however, now find themselves without work. The graders were initially sent home when reports first emerged as they were told there was no work due to "technical errors". They were abruptly told last week, however, that their contracts would be terminated.

The story follows news that Microsoft also hired contractors to listen to personal Skype calls made using the app's translation function. It was also reported these workers reviewed voice commands made to the company's Cortana voice assistant.

IT Pro approached Apple for comment but the company did not reply at the time of writing.

Who would have thought that an online bookshop would end up winning the AI game? Google and Apple are playing catch-up, but Amazon's Alexa is way out in front. The company's almighty retail presence, and its ability to sustain deep price cuts, has helped make Alexa unstoppable.

Once the Echo is settled in your home, though, what exactly do you use it for? For many of us, Alexa is little more than an interactive replacement for the radio. Yet there's a lot more that your always-listening, ever-ready assistant can do. Here's a guide to some of the clever ways you can customise the Echo and make the most of its extensive capabilities.

Change Alexa's wake word

Out of the box, your virtual assistant is called Alexa - but opening vowels can be problematic for some stammerers, and there's plenty of scope for confusion if you have a child or pet with the same name. Happily, while Google, Apple and Microsoft are quite precious about what you call their assistants, Amazon gives you options: instead of addressing your Echo as "Alexa", you can choose to say "Echo", "Amazon" or - for Star Trek fans - "Computer".

To change this setting, open the Alexa smartphone app, click Settings and tap on the device whose wake word you want to change. Click Edit on the wake word line and choose from the drop-down list. We're using "Alexa" throughout this feature, but obviously if you've chosen a different name then swap it as appropriate.

Set up Multi-Room Music

If you have multiple Echo devices in your home, you can easily stream one playlist or radio station to several (or all of them) at once. First though you need to put them into groups; you can do this within the Alexa app, by clicking Settings followed by Multi-Room Music.

You can create as many groups as you like, but note that each device can only appear in one group. So, if you have two Echo units downstairs and two upstairs, you could create a "downstairs" group for the first two - but you couldn't then also add them to a "whole house" group.

To set up your first group, all you need to do is tap the Create Group button, followed by Multi-Room Music, give the group a name and tick the boxes beside the devices that should be included.

Now, if you want to play music on your two downstairs Echo devices, you can simply say "Alexa, play Kylie downstairs". If Kylie's not to your taste, modify the instruction as appropriate.

Play podcasts with Alexa

You probably know that Alexa can stream music and digital radio stations - but did you know that she can also play podcasts? Alexa has a built-in ability to play podcasts through the TuneIn app - just say, for example; "Alexa, play Serial".

This skill, however, may be a bit limited for the true podcast enthusiast's tastes, lacking the ability to do things like subscribe to certain podcasts. If you want a bit more functionality, check out the AnyPod skill, which allows you to do things like subscribe to podcasts and play specific episodes.

Create multiple profiles

Alexa will listen to anyone - but that doesn't mean she has to treat everyone the same. If you set up separate profiles for each person in your household, you can switch between them to ensure that any music played, calendars accessed, and accounts used for shopping will be appropriate to that user.

Creating a new profile has to be done by the registered owner of the Echo device. Open Settings in the Alexa app, click Household Profile in the Settings section, then enter your account password. Now get the other member to log in using the same device and link your accounts.

One word of warning: when you tell Alexa to order an item from Amazon, the system will use whichever payment method is set up for the active profile - so to avoid mix-ups, it's worth checking before you place the order. To do so, just ask "Alexa, which profile is this?"

Protect your purchases

On the subject of voice purchases, if you've got kids in the house you'll probably want to set up a PIN code for online shopping, to ensure they don't order a box load of Lego every two weeks.

To do this, scroll down to Voice Purchasing in the Alexa app and add the code in the "Require voice code" field. This will need to be spoken when making a purchase -so make sure they don't overhear you.

Connect external devices

Everyone knows that the Echo, Echo Plus and Echo Show can stream music from various online sources. What you might not realise is that these devices can all also be used as external Bluetooth speakers, so you can - for example - use them to play music directly from your phone.

To make the connection, either search for the Echo device in your phone's Bluetooth menu, or make your phone discoverable and say "Alexa, pair Bluetooth" or "Alexa, pair phone". (Note that the Dot lacks this capability.)

Once connected, you can control the music with all the usual voice commands, including "play", "pause", "previous" and "next", and volume controls. To close the connection, say "Alexa, disconnect Bluetooth" or "Alexa, disconnect phone".

Use IFTTT with Alexa

You probably already know about the free-to-use automation service IFTTT - there's a guide to some useful business applications here. What's perhaps more handy is the numerous Alexa integrations the service now offers.

To get started, point your browser at ifttt.com/amazon_alexa and click Connect. Provide your password on the Amazon page that appears and authorise the connection. You can now use the pre-rolled applets to link Alexa to a huge range of services and devices - from a Roomba vacuum cleaner or a WeMo coffee maker to Facebook Messenger and Google spreadsheets.

In our view, one of the most useful integrations is between Alexa and Todoist: although Alexa has a native shopping-list feature, Todoist is more flexible and works across almost every platform imaginable.

To set it up, search IFTTT for "Todoist"; you should see an result entitled "Add your new Alexa to dos to Todoist". Click the Turn On toggle switch, and on the next screen, sign into your Todoist account to link it with IFTTT.

Next, select which Todoist project you'd like Alexa to add new entries to - we'd suggest the all-purpose 'inbox' list - and hit save. Now every time you ask Alexa to add an entry to your to-do list, it'll automatically appear in your Todoist inbox.

Create your own routine

If you frequently fire off the same string of commands in a row, why not create a routine, so you can launch the sequence with a single instruction?

To do this, tap Routines in the Alexa Account section of the Alexa app, followed by the "+" to get started. Tap "When this happens", then "When you say something" and type the trigger phrase. We're using "Alexa, open the office" to set up a routine that will run when we get to our desk in the morning.

Save the phrase, then tap "Add action" and choose what should happen from the News, Smart Home, Traffic and Weather options. For example, our routine turns on two lights and asks Alexa to tell us what the day's weather is.

Use Alexa remotely

Did you know that you can communicate with Alexa even when you aren't within range of its microphone? The Alexa app which is used to set up the device can also be used to launch remote commands from anywhere in the world, along with a host of other tasks.

Simply open the app and tap the Alexa logo in the centre of the navigation bar at the bottom to launch Alexa's voice recognition service, and tell her what you want. You can use any command that you'd use with the Echo itself, including turning on connected home devices like lights or plugs, playing music or answering general queries.

Make phone calls with Alexa

Setting up phone-call capabilities is only slightly more involved than setting up the Drop In function. First, open the Alexa app on your phone and tap the speech bubble on the toolbar. When you originally set up the app you'll have been asked for your phone number, and for authorisation to pull in your contacts from your address book - and now you should see them here.

To call or text one of these contacts, tap the person icon at the top of the screen, then tap their name. You'll see options to send a text or call. A texts will be read out by the recipient's Echo device, as well as appearing on their phone, while a phone call make their device ring so you can speak hands-free.

What if you don't want to be disturbed? Tap a name and use the "Allow Drop In" switch to enable or disable their access. Scroll to the bottom of their card and there's a link to block them if you need.

Make Skype calls with Alexa

If you're a business user that relies on Skype rather than traditional phone calls, don't worry - Alexa's got you covered. In the Alexa app, open up the settings and tap 'communication'. Tap on 'Skype' and then click the link to sign in to your Skype account.

Once your account is linked and authenticated, you can simply say "Alexa, call Frankie on Skype", and Alexa will initiate the call. If you've got an Echo device with a screen - i.e. the Echo Show, Echo Spot or a Kindle Fire tablet with Show Mode - you can even use it to make Skype video calls.

Personalise your Flash Briefing

A Flash Briefing is Amazon's name for a quick info dump that draws content from multiple sources, such as news publishers, weather forecasters and exchange rate trackers.

To set one up that's personalised just for you, open the Alexa app and click Flash Briefing in the Settings section. Click Get more Flash Briefing content and select the elements you'd like to add: you'll find options like BBC World Service, The Guardian, MTV and the Joke of the Day.

Each one you add will be automatically enabled, but you can remove any source from the briefing if you choose: just return to the Flash Briefing section and toggle the switch beside each one's name.

You can enable sports content within the Flash Briefing too - but Alexa already knows a lot about football and other sports. Click Sports Update on the Settings screen, then use the search box to find the teams you're interested in. As well as Premier League clubs, you'll also find local teams like Leatherhead and Taunton Town.

Tell Alexa where you are

Another way to personalise the information Alexa feeds you is by telling her where you live. The setup procedure asks you for a street, but a postcode works just as well, and will ensure you get accurate weather and travel information. This should be set for each device by clicking its name in the app's Settings section and entering the required details in the Device location section.

If you frequently make the same journey between two fixed points, you can also setup a custom traffic report. To set your start and end points, step back from the device settings and click Traffic in the Alexa Preferences section. Then, when you ask Alexa "what's the traffic like?", she'll respond with a full report.

Create custom Alexa skills with Blueprints

Have you ever wished you could program Alexa to give certain specific responses? Well, now you can! Amazon has launched a feature called 'Skill Blueprints', which essentially enables you to roll your own custom Alexa skills. There are various templates for creating skills for things like games and quizzes, inspirational quotes and chore rotas.

Our favourite Blueprint is one for letting guests know all the essential information they'll need. The example is based around information for house-guests, such as how to turn the TV on and where the toilet paper's kept, but it works equally well as an information point for offices.

For instance, you can configure it to respond to questions like "Alexa, what's the Wi-Fi password?" or "Alexa, how does the coffee machine work?". No more having to answer the same questions over and over again!

To set up a Blueprint, go to blueprints.amazon.co.uk and sign in with your Amazon account, then pick the Blueprint you want to use as a starting point, then following the step-by-step setup and configuration guide.

Say goodbye, Alexa

If you're upgrading from a Dot to a Plus, or from a regular Alexa to a Show, you might be tempted to pass your old device on to a friend, or sell it online. Before you do, make sure you deregister it so that the new owner can't place online orders using your account.

To do this, open the Alexa app, click Settings, then click on the name of the device you're getting rid of. You'll find the deregistration option in the About section. You don't have to worry about the record that Alexa keeps of thing you've said to her: this won't follow the device to its new home.

But if you ever want to purge this information from your own account, you can delete individual recordings from the home page of the Alexa app, or switch to your Amazon account to delete the whole lot.

To do this, log in at amazon.co.uk, click Your Account, and find the link to Manage Your Content and Devices. Switch to the Your Devices tab, click the three dots beside the names of each of your Echo speakers, and choose Manage voice recordings on each one. Read the disclaimer and, if you want to proceed, click Delete to wipe the slate clean.

Voice technology will be the next major skills gap in the UK IT industry as demand climbs over the next five years, a report has claimed.

The vocal recognition talent pool comprises just 6,000 workers across the country, who are already demanding rising salaries for analytical and developmental work, with businesses already finding it difficult to recruit.

Based on a 2018 forecast the global industry will grow 19.8% to reach $8.3 billion in the next five years, with UK organisations expected to struggle to meet the rising demand unless they adapt the hiring process.

The authors of the report titled 'The rise of voice technology', published by IT resourcing firm Experis, explored how UK businesses might cope as vocal recognition, and its intersection with AI, becomes more widespread.

"While voice technology has been around in the consumer world for the past few years, enterprise adoption will be the next major focus for organisations," said Martin Ewings, director of specialist markets at Experis.

"It will impact every technology interface and could transform the employee and customer experience. As a result, the available talent pool will soon be exhausted.

"Today, the top salaries for Technical Architects in voice technology are in line with similar roles in other technology disciplines. But these salaries will rise as candidate availability becomes stretched."

The report has dubbed the integration of voice technology into businesses, and the talent shopping list it will generate, as a "potential game changer". But the industry is in its infancy, with hiring demand still relatively low for individuals with voice technology skills.

Average salaries, according to findings, start at 18,000 and rise to 75,000 per year for more senior roles, such as for technical architects.

But location is the most crucial element for firms looking to hire given how lacking the vast majority of regions across the UK are in the right talent pool. According to the Industry Insiders report, most talent is based in the South East in London, with a moderately-sized pool also based in the East and North West of England.

Also reflecting the wider industry, the gender breakdown is weighed heavily towards men, with an 84% to 16% split of men to women. The report warned that excluding women from setting norms and parameters of this new technology will raise the likelihood of bias being developed into the software.

Organisations must also extend beyond software developers to business analysts and enterprise architects, the report said, with voice being integrated into a business' cloud strategy, security architecture and data model, among other areas.

"Businesses need to get ahead of the potential talent shortage by targeting those with adjacent skills in coding languages like C, C++, Java, and PHP," Martin Ewings continued.

"Employers must look both within and beyond their businesses for talent and consider cross-training those with relevant coding skills to fill the anticipated void in voice technology talent."

The report recommended that turning to candidates with transferrable skills could pose a solution to the looming talent crisis, such as hiring workers proficient in similar coding languages. Those with C, C++, Java and PHP skills are particularly valuable, as it could expand a talent pool from some near-6,000 to almost 85,000 candidates.

Several recommendations, including upskilling workers from declining industries and providing an attractive route for workers to acquire voice skills, were issued that could help businesses preemptively plug any future skills gap.

Experis, for instance, highlighted one case in which hundreds of under-employed textile workers were retrained to fill a skills crisis one of the most advanced motorsport manufacturers was experiencing.

With the help of technical schools, universities and the government, the workers were able to fabricate stronger, lighter-weight components for the car manufacturer.

"Since voice technology remains in its infancy, there are a limited number of specialists to meet demand," the report concluded. This means it is important to consider how individuals from IT specialisms that are experiencing a reduction in hiring demand can be redeployed in this emerging field.

"What's more, it is also important to remember that voice technology is not a standalone project or initiative. To be effective, it must seamlessly dovetail into every element of your IT strategy, which will require significant training efforts to ensure your IT department at large is prepared for this next wave in computing."

Dictation software has come a long way since the early days of dodgy speech transcriptions. Today there are incredibly sophisticated tools available that never fail to surprise with how accurate they are at translating even the most accented of human speech.

What's more, systems now offer users a dizzying array of functions, whether that's pinpoint accuracy, PC controls, collaboration tools, or even the option to sync your transcriptions to your PC so you can talk while you're on the go.

With all this added complexity, choosing the right platform for your needs is understandably difficult – so we've done the hard work and pulled together the most impressive software available.

Google Voice Typing

docs.google.com

Getting started with voice typing in Google Docs is as straightforward as you'd expect from a company whose home page is a white screen with a search bar: once you've opened a document, and selected 'Voice typing' from the drop-down Tools menu, just click the on-screen microphone and speak.

The tool is reasonably fast – there's a slight lag, because it uses cloud processing to figure out exactly what you're saying – but after a second or so, your words, punctuation and even formatting appear on the page. And Google is, by far, the most accurate of all contenders in this roundup.

We tested each tool by reading the opening paragraph of Raymond Chandler's 'The Big Sleep' five times, using both built-in and external microphones. Out of 84 words, Google Docs scored an average three errors; typically two typos and a dropped word.

With any free voice dictation software, it's recommended you run through the content afterwards, manually correcting mistakes – Google Voice Typing's impressive accuracy makes that job a lot easier, especially since right-clicking underlined words displays possible suggestions for misheard words. You can also use the 'Delete [word or phrase]' command to make corrections on the fly.

There's more to Voice Typing than actual typing. You can highlight words, insert tables and shift around the page by ordering the cursor to relevant sections. This helps maintain your flow, rather than breaking it as you pause to grab the mouse. Visit Google's support page for the full list of voice instructions.

How it can be improved

The biggest irritation is that Voice Typing is locked to Chrome. Try to access the option in any other browser and you'll find it greyed out. Currently, the tool's only available for Docs and for adding speaker notes to Slides. We'd like to see speech recognition rolled out to other Google products – it seems a logical evolution, considering the rise of smart speakers such as Google Home.

Verdict

Google Voice Typing delivers an effortless and impressively accurate speech-to-text experience. Despite being blocked on non-Chrome browsers, its sleek presentation and admirable performance makes it far less frustrating than similar tools, and a joy to use.

Features: 4

Performance: 5

Ease of use: 5

Overall: 5

Microsoft Dictate

dictate.ms

At first glance, you'd be forgiven for thinking that Dictate – a Microsoft Garage project created by developers in their downtime – is a needless add-on. After all, Windows already builds Speech Recognition into various Office products.

But this voice-recognition tool, which is compatible with Word, Outlook and PowerPoint, harnesses the might of Microsoft's Cognitive Services, the online AI that fuels Cortana and Translator. That power means Dictate does what Speech Recognition has always struggled with: recognising speech.

Selecting the Dictation tab on the ribbon reveals a sparse interface featuring a Start/Stop button – click this and your spoken words whip along a Response bar, processed online, before appearing on-screen without too many mistakes; an average hit-rate of 80 out of 84 words.

Elsewhere, you'll find an option to switch to manual punctuation. But the stand-out feature is the translator that converts speech into all major languages as you speak.

How it can be improved

Dictate isn't the quickest horse in the stable, although it's fun to see your words gallop across the page three seconds after uttering them. There are some peculiar word choices, and, like Google, it occasionally typed 'comma' and 'full stop' where punctuation marks would suffice. The real deal-maker-or-breaker is that, in running on Microsoft Cognitive Services, Dictate only works online.

Verdict

Dictate is so good, you'd be forgiven for thinking it's a fully fledged, Microsoft-supported product; not a homebrew piece of kit. Translation integration is expertly done and we wouldn't be surprised to see this become a true feature in later Office iterations.

Features: 4

Performance: 4

Ease of use: 5

Overall: 4

Braina Lite

Brainasoft.com

Braina Lite – the free version of an otherwise paid-for tool – is an AI-based virtual assistant, which is capable of opening files and programs on your computer. It might not steal the crowns of Google and Cortana, but that's down to poor design, rather than performance.

The tool barely stumbles when transcribing. Faster than its rivals, Braina Lite delivered the same overall accuracy as Google. But the tool is flawed, since you need to copy and paste captured words into your word processor separately.

Dictation mode should let you directly dictate into third-party software such as Word, but the software demands a full purchase to use this feature – along with several others. At least if your microphone isn't up to scratch, you can grab the Siri-style app (iOS and Android).

This easily connects to a computer on the same network, and even packs in a virtual keyboard and mouse, should you need to give your jaws a rest.

How it can be improved

We rarely come across software that so completely locks down the menu bar; almost every button slotted at the top of the screen is out-of-bounds in the free version. Instead, you need to manually type such commands as 'Dictation mode on'. This is stunningly counter-intuitive for a voice-controlled tool.

Verdict

Voice-activated computer control is a breeze here, and far more responsive than Windows Speech Recognition. But the constant harassment to buy the product buries otherwise excellent features, and we found Braina Lite too limited in its free form.

Features: 3

Performance: 5

Ease of use: 4

Overall: 3

Best of the rest

Windows Speech Recognition

products.office.com

Built into various Microsoft Office products, Windows Speech Recognition offers voice typing, as well as letting you entirely control your computer. However, you'll need to dedicate time to training the tool to learn your vocal idiosyncrasies to get the most from it. Our experience was hit-and-miss – although it never failed to deliver the correct punctuation, plenty of the words in-between were horribly garbled.

Speech Notes

speechnotes.co

This browser-based speech recognition tool is powered by Google's own back-end servers, but the only similarity with Google Voice Typing is the requirement to use Chrome.

Elsewhere, expect poorly transcribed words and disdain for proper grammar. We did find it worked okay when speaking slowly and pausing after each sentence – at which point, it's probably quicker to just write it down.

Speech to Text Converter

microsoft.com

Speech to Text Converter does what it says on the tin. Available from the Microsoft Store, it's a no-frills, no-nonsense app. The simple interface, which imitates the dark Windows 10 theme, features just a text box, Start button and an option to clear text. It's perfect for making a quick note that can be pasted into a word processor, but the overall accuracy just isn't dependable enough.

Millions of taxpayers' voice recordings have been logged and stored by HMRC without their consent, according to a privacy campaign group.

The tax office has collected 5.1 million 'voiceprints' from UK citizens who found they were unable to access its services without repeating the phrase 'My voice is my password' on an automated line, Big Brother Watch claimed today.

"Taxpayers are being railroaded into a mass ID scheme that is incredibly disturbing," said Big Brother Watch director Silkie Carlo. "The tax man is building Big Brother Britain by imposing biometric ID cards on the public by the back door."

Its Freedom of Information (FoI) request revealed that HMRC started recording voiceprints of callers to its tax credits and self-assessment helplines in January 2017.

Callers were unable to speak to advisors without recording their voice, Big Brother Watch claimed. Meanwhile, HMRC declined to provide Big Brother Watch with a copy of the privacy impact assessment for the scheme, and claimed the department has yet to consult with the UK data protection watchdog, the Information Commissioner's Office (ICO), about the data collection.

Capturing people's voice recordings without their freely given consent, or making clear what it's being used for, would be in contravention of the Data Protection Act 2018 and the EU's General Data Protection Regulation (GDPR) rules, unless HMRC can demonstrate a legitimate interest in collecting such data.

However, Pat Walshe, data protection law expert and director of Privacy Matters, accused HMRC of "failing to meet basic data protection principles".

HMRC's automated phone line told callers: "I'll need you to say exactly those words", BBW found after members of the public drew its attention to the practice.

There was no way to opt out of recording your voice, Big Brother Watch found, with callers saying 'no' repeatedly being told: "It's important you repeat exactly the same phrase. Please say 'My voice is my password'."

Big Brother Watch managed to get around it by repeating 'no' three times, but this process wasn't made clear by HMRC's phone line.

A Big Brother Watch spokesperson told IT Pro: "It's such a remarkable ID database nationally [but] there's no privacy impact assessment [that we can see] and they haven't engaged with the ICO about it."

HMRC has also not consulted with the Biometrics Commissioner about its storage of voiceprints.

While people have a right to the erasure of their data under the Data Protection Act 2018, HMRC declined to answer the Freedom of Information request to find out what process people could follow to have their voiceprint deleted, citing the prevention or detection of crime exemption.

"This suggests that taxpayers' voiceprints are being used in ways we do not know about," BBW said.

However, HMRC refused to disclose which other government departments it has shared the voiceprints with, where and how they're stored, or what they're used for.

"These voice IDs could allow ordinary citizens to be identified by government agencies across other areas of their private lives," Carlo added. "HMRC should delete the five million voiceprints they've taken in this shady scheme, observe the law and show greater respect to the public."

The Information Commissioner's Office (ICO) is now looking into the issue, and Walshe said it could order the tax office to suspend the scheme.

An HMRC spokesperson said: "Our Voice ID system is very popular with customers as it gives a quick and secure route into our systems. The Voice ID data storage meets the highest government and industry standards for security."

Image: Shutterstock

Microsoft CEO Satya Nadella yesterday highlighted the key role ethics must play in developing future AI technology.

Speaking at Microsoft's Leading Transformation in AI event in London, Nadella predicted AI will render enormous change in "every walk of life", but warned that developers, including those in his own company, would need to take their ethical responsibilities seriously.

"We're at that stage where the choices we make are grounded in the fact that technology development doesn't just happen - it happens because us humans make design choices," he said in a speech spanning a broad range of topics, including the Internet of Things (IoT) and quantum computing. "Those design choices need to be grounded in principles and ethics - and that's what's the best way to ensure the future we all want."

Speaking favourably of a recent House of Lords report into the need for an ethical code of practice in AI, he revealed that Microsoft has formed its own ethics committee, also stressing that the tech giant is working to ensure its technology reflected the high standards Nadella said Microsoft aspires to.

One of the key challenges in language, he said, "is the models that pick up language and learn from the corpus of human data available", lightly referencing Redmond's infamous Tay bot scandal - in which the company's Twitter-based chatbot, designed to emulate a teenage girl, began parroting racial slurs and conspiracy theories after being exploited by trolls.

"Unfortunately the corpus of human data is full of biases," he conceded, "so you need to invest in tooling that allows us to de-bias when you model language from the corpus of human data."

Joined by fellow Microsoft executives and customers from Unilever and Great Ormond Street Hospital (GOSH), Nadella outlined his vision for how AI could augment human capability, not replace it, mainly in language, audio and visual applications across a range of industries.

He cited the use of EmpowerMD at Pittsburgh Medical Centre, a software designed to cut down transcription time by intelligently transcribing doctor-patient conversations.

"The doctor is more efficient, and the patient is getting more of the doctor's attention and time, which is, in fact, the benefit of AI in the workspace," the CEO said. "And these design choices around we make in how we create AI and bring to market underline how we hope to make impact."

Unilever's CIO, Jane Moran, spoke at the event about how the use of machine learning has led to improvements in analysing and democratising data, before she demonstrated how its AI-powered chatbot, Unibot, is being used across the whole organisation to cut 90% of HR time spent on manual tasks like answering basic queries.

"Crucially, it frees up the HR professional's time to really focus on productive ways to support the staff in the business," she said, adding that the platform can be built out into other divisions in the company including legal, finance and manufacturing.

Meanwhile, Nadella outlined his vision for how AI is naturally integrated with cloud and edge computing, highlighting recent shifts in how the computing fabric itself is becoming distributed from the cloud to the edge - allowing developers to build multi-sense and multi-device experiences.

"What is perhaps at the core of the experiences we create, and the applications that we create, as well as the platform itself, is AI," said Nadella. "AI is the runtime that is going to shape all of what we do going forward in terms of the applications as well as the platform advances."

"We're reconceptualising the core products we build," he said, explaining the way the company's Office, Management and Security suites were rolled into Office 365, with the same principles applying to Dynamics 365, as well as its gaming arm, and LinkedIn.

Microsoft has acquired machine learning startup Semantic Machines to push conversational AI into Cortana and its Azure Bot Service.

California-based Semantic Machines uses machine learning and adds context to conversations with chatbots, taking information received by the AI and applying it to future dialogue.

The company is run by a mix of conversational AI pioneers, including entrepreneur Dan Roth, UC Berkeley professor Dan Klein and Stanford University professor Percy Liang, as well as former Apple chief speech scientist Larry Gillick. Its speech recognition team previously led automatic speech recognition for Apple's personal assistant Siri.

Microsoft hopes to add Semantic Machines' technology into its current work in conversational AI, improving speech recognition and natural language understanding. Today, it reports that one million developers use its Cognitive Services and 30,000 developers use Azure Bot Services.

"With the acquisition of Semantic Machines, we will establish a conversational AI centre of excellence in Berkeley to push forward the boundaries of what is possible in language interfaces," said David Ku, chief technology officer at Microsoft AI and research.

"Combining Semantic Machines' technology with Microsoft's own AI advances, we aim to deliver powerful, natural and more productive user experiences that will take conversational computing to a new level."

Like Microsoft, Google and Amazon have take steps to improve voice assistants and conversational AI. Earlier this month Google showed off its own human-sounding virtual assistant Google Duplex, which booked a hair appointment over the phone by mimicking human speech.

Amazon has also looked into improving conversational capabilities with Alexa and have said it is developing a 'memory' for it, which will enable it to remember bits of information it is given by users while in a conversation.

Picture: Shutterstock

Microsoft yesterday set out to open up its cloud platform to developers hoping to build AI applications, revealing a host of machine learning initiatives at its annual Build conference.

Redmond wants its Azure cloud to be the backbone of developers' AI innovations, with CEO Satya Nadella saying:"The era of the intelligent cloud and intelligent edge is upon us. These advancements create incredible developer opportunity and also come with a responsibility to ensure the technology we build is trusted and benefits all."

Project Kinect for Azure,is a package of sensors, including its next-generation depth camera, with onboard compute designed to allow local devices to benefit from AI capabilities.

Meanwhile, Microsoft's Speech Devices SDK aims to enable developers to build a variety of voice-enabled scenarios like drive-through ordering systems, in-car or in-home assistants, smart speakers, and other digital assistants.

The tech giant also previewed its Project Brainwave, an architecture fordeep neural net processingthat Nadella said "will make Azure the fastest cloud for AI". This is now available on Azure and for edge computing, and is fully integrated with Azure Machine Learning with support for Intel FPGA hardware and ResNet50-based neural networks.

The firm is also open sourcing the Azure IoT Edge Runtime, allowing customers to modify, debug and have more transparency and control over edge applications.

Azure IoT Edge also runs Custom Vision technology, a new service that Microsoft says enables devices such as drones and industrial equipment to work without cloud connectivity. This is the first Azure Cognitive Service to support edge deployment, with more coming to Azure IoT Edge over the next several months, according to the vendor.

"With over 30 cognitive APIs we enable scenarios such as text to speech, speech to text, and speech recognition, and our Cognitive Services are the only AI services that let you custom-train these AI capabilities across all your scenarios," said Nadella.

Also announced was a new SDK for Windows 10 PCs in partnership with drone company DJI. Using Azure cloud, the SDK brings real-time data transfer capabilities to nearly 700 million Windows 10 devices. As part of the commercial partnership, DJI and Microsoft will co-develop tools leveraging Azure IoT Edge and Microsoft's AI services to make drones available for agriculture, construction, public safety and other verticals.

Nadella also announced the company's new AI for Accessibility, a $25 million, five-year programme aimed at using AI to help more than one billion people around the world who have disabilities.

The programme consists of grants, technology investments and expertise, and will also incorporate AI for Accessibility innovations into Microsoft Cloud services. Microsoft said that the initiative is similar to its previous AI for Earth scheme.

Intelligent apps

Microsoft also handed developers the ability to introduce more customisation forMicrosoft 365 applications, so organisations can tailor them to their needs.

Developers whose businesses use the Microsoft 365 bundle of Office 365, Windows 10 and Enterprise Mobility and Security can now benefit from wider integrations in collaboration app Microsoft Teams, and even publish custom apps on the Teams app store.

There is also deeper SharePoint integration within Microsoft Teams to enable people to pin a SharePoint page directly into channels to promote deeper collaboration. Developers can use modern script-based frameworks like React within their projects to add more pieces that can be organised within SharePoint pages.

The vendor also revealed new Azure Machine Learning and JavaScript custom functions that let developers and organisations create their own additions to the Excel catalogue of formulas.

Picture: Satya Nadella/Credit: Microsoft Press Centre

Intel and Amazon are working together on a developer kit, aimed at making it easier for third party developers to include Amazon Alexa Voice Service on their own devices.

The Intel Speech Enabling Developer Kit aims to provide the hardware components for voice control so that all developers will have to do is write the applications for the hardware, which is said to be the first "audio front end solution" for the combining of voice processing and wake work detection on one chipset.

The kit includes high-performance algorithms for echo cancellation and noise reduction, allowing the device to work even in noisy environments.

It also has Intel's mic circular array and its dual DSP (digital signal processing) with inference engine.

"People speak and hear in 360 degrees, not just in a direct line of sight. Devices need array microphones and complex noise mitigation technology. A quality voice interaction means devices identify the speaker's location, mitigate and suppress ambient noise, and understand spoken commands on the mics, even while playing music, as well as waking up when it hears the wake word," said Miles Kingston, the general manager of the Smart Home Group at Intel.

With Amazon wanting to see its Alexa technology in more and more devices, this new collaboration seems to be pushing the world further into the smart home era.

"We can all look forward to a wave of innovation from the developer community as we transition from being simply connected to being truly smart," Kingston said.

The new kit is available for pre-order starting today for $399, although the cost will raise $100 after the presale window closes.

Who isn’t guilty of having used a simple password like ‘football' or '123456'? In a world that demands authentication for each and every service, how can users be expected to memorise all the different combinations of keys?

However, what might seem like an easy way out is actually a security nightmare from an IT professional’s point of view. Weak passwords are the cause of a significant number of preventable data breaches, with many so ill-thought that it is unsurprising to see them cracked.

On the other side of the spectrum, there are also users who set a really good, strong password – and then proceed to reuse it for all their logins across multiple devices and systems, undermining its power in one fell swoop.

These unfortunate, yet familiar, scenarios show that we, as human beings, are prone to undermining our security strategies – regardless of how foolproof they seem.

However, while our memory can be futile, humankind has also been equipped with unique physical characteristics which might just be the answer to our IT security crisis.

Fingerprints and facial features already offer the benefits of not having to be memorised nor stored in a password manager app, resolving many of the problems caused by passwords. These are the most common examples of biometrics – the statistical analysis of human characteristics which, in tech, is most often used for identification purposes. Considered to be the most secure method of authentication, biometrics use a unique piece of your own biology, such as iris shape or papillary lines, in order to verify whether you are who you claim to be.

Nevertheless, despite its wide-ranging popularity in recent years, there are also a number of ethical concerns surrounding biometrics, from the impact of facial recognition technology on our privacy to a few morbid examples of unlocking devices with a dead person's finger.

Biometric authentication

Facial recognition is often used in CCTV evidence gathering

Biometric authentication is the process of turning captured human characteristics and behaviour, into a digital format. It is by far the most common form of biometric identification, and the most common authentication method is fingerprint scanning; although this can also include face, iris, voice, DNA, and even the way we move or perform tasks, such as typing.

Unlike traditional measures, biometric authentication requires the person to be present while data is captured. This means that it's almost impossible for a hacker to steal a person's credentials, as they are intrinsically tied to an individual.

What's great about biometrics is that not only is it considered the most secure form of authentication, it's also the simplest from an end-user perspective. It frees the user from having to remember ID numbers or complex password strings, and best of all, there's no need to come up with ridiculous password recovery questions and answers - there's no way to lose your DNA, outside of some adventurous scientific experimentation.

Can biometrics be hacked?

Although the idea of biometrics sounds incredible in principle, it hasn’t always been the pillar of success. In fact, early attempts at rolling out biometric authentication managed to generate rather mixed results.

Despite fingerprint scanners being widely available for some time now, innovations such as voice-based authentication for banking have proven less feasible. In 2017, BBC reporter Dan Simmons famously discovered he was able to sign into HSBC's mobile banking system by mimicking his twin brother’s voice.

Facial recognition isn’t without faults either. In April 2017, it was reported that a Samsung Galaxy S8 user was able to log in through the device’s facial scanner using a printed-out selfie – a trick which could be very easily exploited by criminals.

A more recent incident involving Samsung devices occurred last year when it was discovered that the Galaxy S10's ultrasonic sensor could be easily unlocked using a 3D-printed fingerprint.

Apple's iPhone X onwards boasts the latest in smartphone facial recognition technology

Apple promised to solve such problems with its now-discontinued iPhone X, which was released in late 2017. Using infrared and visible light scans, the device’s Face ID technology was able to almost instantaneously analyse a person's face and hide the onscreen notification messages if the phone was picked up by a stranger. In fact, Apple stated that the chance of a stranger being able to unlock someone else’s phone was 1 in 1,000,000, making it increasingly safer than other methods of biometric authentication.

By contrast, the same likelihood of unlocking a phone using a fingerprint scanner was estimated to be 1 in 50,000.

The PIN is still king

While biometrics seems to be slowly taking over industries such as banking, with NatWest expected to roll out behavioural biometrics technology this year, it might still be outdone by the good old PIN.

This is largely due to the fact that, although the underlying principles of biometric authentication are incredibly effective, its development has been dictated by the pace of technology. Unfortunately, the tech industry has yet to come up with a highly-accurate system that's also cost-effective for mass deployment.

For better or worse, passwords and two-factor authentication still rules the day – at least for now.

Whatever you use to perform web searches, whether it's your mobile, tablet or desktop, it's likely it's all about to change. Soon, you'll be scrapping the keyboard and reaching for the voice activation button instead - that's if predictions from analyst firm Gartner and Comscore are correct.

For example, Gartner thinks that by 2020, 30% of web browsing sessions will be carried out without a screen and this means voice will begin its quest to take over from the keyboard (whether virtual on a smartphone or hardware on a desktop).

"New audio-centric technologies, such as Google Home and Amazon's Echo, are making access to dialogue-based information ubiquitous and spawning new platforms based on "voice-first" interactions," the company said in its report into 2017 trends.

Comscore is a little more optimistic, saying it thinks 50% of searches will be made by voice by 2020.

Thanks to the introduction of voice assistants such as Amazon Echo, Apple Siri, Google Now and Microsoft's Cortana, tech innovators are deeming the keyboard defunct, instead encouraging everyone to use their voice to search the internet.

What is voice search?

Voice search allows you to simply read out what you'd like to search for. It's usual for a variety of things, whether you want to find out information using a search engine, directions to a meeting or want your virtual voice assistant to call someone for you, rather than having to type a search into Google, Bing or any other search engine of your choice, find a phone number or address of your next meeting manually.

You'll usually need to say a command first, such as "OK, Google" to trigger the speech recognition software to start running, but once you get confirmation it's listening, you can simply speak your request.

Your words are sent to servers, de-scrambled using speech recognition algorithms working in tandem with machine learning to identify the pattern of speech and then this information is sent to the search engine to perform the search, trawl your device for the meeting you're supposed to be traveling to or your contacts to find who you want to call.

The response then pings back to the server, where it's either presented as a web page to your device, it performs the action, or, if you're using a speaker such as Google Home or Amazon Echo, it'll transform back into natural speech and be amplified for your entire family to hear. All of this happens in seconds - you won't even notice the process behind it even happened.

How popular is voice search?

Although technology firms would have us believe voice search is going to soon outstrip the number of people using typed search to find what they're looking for, Google has revealed the technology still only accounts for approximately 20% of all searches performed on its Android and mobile apps. Overall, across all platforms, just 12% of searches are made using voice, according to the tech giant.

However, Google voice search queries increased by 35x in 2016 compared to 2008, showing it is an increasingly used technology.

Why is voice search becoming so ubiquitous?

The main reason why voice search is set to grow is the introduction of keyboardless devices and assistants powering up your smart home. We're referring to Google Home, Amazon Echo and other devices including your smartphone and tablet that connect to your media, the internet and generally help you plan your life.

Accuracy is also improving considerably, which makes more people want to use voice search. Google claims its voice recognition accuracy is now at 92% (last year it was 75%) and although this means you're still likely to get some inaccurate results filtering through, the more people that use voice search to power their lives, the more the algorithms behind them will learn what you're asking for, producing relevant and better results.

Right after its millennial Tay bot turned genocidal might not seem like the best time for Microsoft to pin its future on bots, conversations and artificial intelligence, but that's exactly what Satya Nadella announced at Build 2016 last night.

The Redmond CEO claimed that "we are on the cusp of a new frontier that pairs the power of natural human language with advanced machine intelligence".

Microsoft wants to bring conversation into so many places that it becomes the next stage of the GUI; "we want to take the power of human language and apply it more pervasively to all computing interfaces," Nadella said.

Conversations as a Platform, as Microsoft calls it, envisions a world where you ask Cortana to block out the week you'll be at a conference in your calendar and she tags in the bot from your favourite hotel to book a room for the right dates.

This bot, by the way, already knows what kind of room you prefer, and suggests a message to send to a friend who lives nearby letting them know when you'll be in town.

That extends the way Cortana can already index your email, looking for meetings you ought to go to and things you've promised to do to services like Skype and Slack, so she can extract information from video voicemail (which comes with a handy transcript using voice recognition since Microsoft is extracting the information for the bots, you can see it too).

If you're working late, Cortana might ask if you want to order a takeaway and pops up the Just Eat bot or the Dominos bot to take your order depending on what she knows about your eating habits.

Behind the scenes, the Dominos bot will be using Microsoft's Cognitive Services APIs to understand what you're saying or typing or talking about, although it's up to the developer whether they want to accept suggestions from the text recognition API that their bot should know that send a pizza to my crib' means deliver a pizza to my house'.

The intelligent underpinnings of conversational interfaces draw on AI and machine learning research Microsoft has been doing for a couple of decades. They also rely on on what they know about how businesses work, the information in your mail, and documents stored in the Microsoft Graph.

This is the big data processing that Azure is so good at and is more evidence of the cross-platform mobility of experience Nadella has been preaching from day one, because although Microsoft wants Windows to be the best place to have a voice conversation with Cortana, this will work in Cortana on your iPhone or Nexus too.

Intelligence is not just voice recognition for talking to bots and language understanding for typing messages. It's also image recognition (which down the line is going to include recognising people, places and things in video, along with emotions), and the other 20 APIs Microsoft is already offering.

It's also holograms and the way HoloLens maps the real world to mix virtual into reality. It's also the smart handwriting recognition any developer can put in their Windows apps - when you write tomorrow' and Cortana recognises it as a date, it's another way you could start a chat about what you want to do or automatically get a reminder when you need it.

Really, it's the Natural User Interfaces we've been promised for about a decade, and they might finally become real.

Plausible, polite conversations

If conversations with bots sound automated and dehumanising, like a future of brands shouting their wares at you, look at projects like Seeing AI, a service for smartphones and smartglasses to help the visually impaired understand who and what is around them.

The app can guide a blind person to take a photo of the menu (move the phone up and to the right') and then read it out to them and rather than listing that menu, it lets them pick and choose the sections they want to hear about.

Alternatively you can swipe a finger over the earpiece of your smartglasses to snap a photo and have it automatically captioned, so you know what's happening ( I think it's a girl throwing an orange frisbee in the park' says the system, using the kind of image captioning that was a research project only last year).

Cleverly, Microsoft demoed this capability by recognising the age and gender of the people you're talking to, instead of trying to do facial recognition on strangers walking down the street. That's a perfect of example of what Nadella calls "intelligence that augments human capabilities and experience", saying that "it's not going to be about man versus machines, it's going to be about man with machines".

Similarly, bots that talk to you or that Cortana deals with automatically for you - don't get to know your location unless you allow that. For example, Cortana might ask do you want the courier to know where you are this afternoon so the delivery comes to the right place?'.

Cortana can also securely pass along your credit card details, an innovation that might even work with something like the FIDO standards that underpin Windows Passport, so Cortana can do your banking for you, once you've used the Windows Hello biometrics to sign in.

AI is one of the biggest buzzwords around at the moment. It's also often hype. What Microsoft is promising is more achievable than some AI futures. Think of a bot as very simple app, or better yet a more useful search result that lets you do what you want rather than clicking on a link and hunting through a web page looking for the right tool.

That doesn't mean the idea of pervasive conversational interfaces isn't ambitious and it won't necessarily be easy to get right. It's not that the effects will be so much worse if it goes wrong than we have problems with other areas of technology; this is image recognition and language understanding, not robots with lasers and getting a hotel booking wrong is as easy on a website as it will be talking to a bot.

But because of the personal, intuitive, conversational interface, it might feel a lot more upsetting if a hotel bot gets our room wrong or a restaurant bot books the table for the anniversary of your first wedding when you're divorced and remarried.

That's one reason Nadella was careful to position AI, bots, assistants, conversational interfaces and intelligence as part of the future landscape of technology, but also something society needs to have a discussion about. He specifically mentioned the problems with Tay (while subtly noting that Microsoft's chat bots in China and Japan haven't run into the same malicious pranking by users), promising Microsoft would take "a principled approach" to adding intelligence to interfaces, and make technology that's "more inclusive and respectful" with those principles including privacy as well as security and compliance. "We want to build technology so we get the best of humanity, not the worst," he notes.

That makes conversational interfaces part of the ongoing discussion about accessibility and diversity in technology. Nadella hinted at other common questions like where the balance between security and privacy lie with things like encryption and whether technology is actually contributing to GDP.

Conversational interfaces that actually work could be great for productivity as well as extending access to computing to a lot more people than are comfortable with laptops and smartphones. That makes the old Microsoft promise of a computer on every desktop' seem almost unambitious.

The latest version of Nuance's speech recognition software, Dragon Professional Individual extends and enhances the program to make it still more flexible and convenient as an alternative to typing.

Designed for the busy professional a home worker, sole trader or anybody who doesn't require centralised administration of their speech recognition it's also geared to work with Nuance's new cloud-based product, Dragon Anywhere, when it launches early in 2016.

Dragon PI builds on the speech recognition engine of earlier versions, which was already extremely accurate, and concentrates on its automation and transcription abilities to improve productivity still further, from a wider range of devices.

Getting going

The key advantage to using Dragon for text dictation and control is the speed and accuracy it provides. Even if you're a competent touch typist, and many people aren't, you should notice at least a three-fold increase in creation speed when you dictate text, rather than typing it.

If you're an existing Dragon NaturallySpeaking user, upgrading to the new product is very straightforward. The software installs over the previous version and updates existing profiles to work with the new application. It also requests and then downloads and installs browser extensions, to enable it to be used with Internet Explorer, Chrome or Firefox.

The software works with a wired or Bluetooth headsets, as it always has, but now it also works with many microphones built into laptops and other devices, so you don't have to dress up' before using the software.

The program is controlled via the Dragon Bar, a small control panel which sits at the top of the screen by default, and zooms itself down to not much more than a button when you're using other applications. Conventional menus give access to tools, vocabulary and your speech profiles and there is also access to Dragon's copious help system. There's a command cheat sheet included in the online help, but it's a shame this isn't provided as a printed sheet, too.

You don't have to train the software and can use it straight out-of-the-box with a recognition accuracy of more than 99 per cent, or less than one word in a hundred will be incorrect. This review was dictated, and we only had six corrections to make. Spelling errors are a thing of the past and only homonyms, like to' for too' can occasionally cause a problem, though even these are rare.

If you are new to Dragon speech recognition, you will welcome the improved help via the simple command What can I say?' which calls up a quick menu showing the most frequently-used spoken commands. This menu is context-sensitive, so you're not drowned in long lists of all available commands.

Auto-text and custom commands

If you use phrases repeatedly, you can further increase your productivity by creating Auto-text boilerplates. Using the redesigned Dictation Box, quite intricately structured paragraphs can be entered and assigned an Auto-text name, which can then be used when dictating, to insert the standard text.

This technique can now be extended by the use of variables in the Auto-text name, so you can, for example, add a person's name to a salutation Auto-text, simply by saying it after calling the Auto-text.

In early 2016, Nuance will be launching Dragon Anywhere, a new speech recognition application which works from the cloud and can be used on iOS and Android devices. Dictation files can be moved between different versions of Dragon's software, while customisations, such as preferences and auto-texts, can be easily synced between the two products.

You could, for example, create a simple form and attach it to an Auto-text command, which would then insert it into any new document. As well as being available within Dragon PI, any new command will also be available immediately to any device using Dragon Anywhere, simply by setting the two applications to sync automatically.

Transcription

You don't have to dictate directly into your PC, as Dragon can transcribe from audio files created with a variety of different recorders. You may want to record notes while on a train, for example, and then have the software transcribe these into Word, RTF or text files when you get home or back to your office.

You can have several sources set up within the program, so the software adapts to the particular recording device you're transcribing from, such as a digital recorder, phone or tablet. With the last two, though, Dragon Anywhere may prove a more convenient choice.

As an extension to the program's transcription facilities, you can specify a watch folder', where any new audio files will automatically be transcribed; simply copying files there will create corresponding text files.

Controlling applications

Dragon is not restricted to taking dictation in Word. It can also be used for controlling other Office applications, like Excel and Outlook and works with the three main Internet browsers: Internet Explorer, Chrome and Firefox, using extensions which are available as free downloads.

Within a browser, the software can be used to control applications like Gmail and Wordpress and create and edit emails and blogs. Wherever you use Dragon, you can combine spoken commands and text creation with the standard keyboard and mouse use to get the most productive combination of control methods for your personal requirements.

Word processing and browser use are not isolated features, but work together in useful ways, such as when sending an email. Giving the command send e-mail' from within Word wraps up the current document as an attachment, opens Outlook and creates a new email, for you to simply add recipient and any covering text; much quicker than the conventional route.

This product continues Dragon's evolution as a productivity tool for anybody concerned about or inundated with text creation in their day-to-day work. A very high level of accuracy is taken as read and this latest iteration of the product concentrates on ease-of-use and facilities to automate regular tasks. A little time invested in customising the product to your own needs can produce benefits which you'll appreciate for much longer.

Buy Dragon Professional Individual (for PC) online now.

This article is brought to you in association with Nuance.

Facebook is buying a natural language recognition startup as it tries to muscle in on the territory of personal assistant tools like Siri and Cortana.

The social network hopes to improve its offering with the services of Wit.ai, an open platform that allows developers to build apps that people can talk to.

The startup, founded in 2013, claims to have 6,000 developers using its technology to power hundreds of apps and devices.

In a blog post on its site, Wit.ai's team wrote: "It is an incredible acceleration in the execution of our vision. Facebook has the resources and talent to help us take the next step.

"Facebook's mission is to connect everyone and build amazing experiences for the over 1.3 billion people on the platform technology that understands natural language is a big part of that, and we think we can help."

Wait.ai's tool can take audio of the user's requests recorded by the app and turn it into structured data, telling the app to perform whatever task was requested by the user.

While the value of the deal has not been revealed, it gives Facebook a chance to catch up to rivals like Google's Google Now, Apple's Siri and Microsoft's Cortana, all of which are popular voice assistants.

Microsoft is also making other moves in speech and machine learning, with the Preview release of Skype Translator, which is currently able to translate conversations held between English and Spanish speakers in near real-time.

Wit.ai stated it will continue to work with developers, with the technology remaining open and becoming completely free to use.

The "OK Google" voice search command, available to users of Chrome OS, is now being ported to the search giant's internet browser.

Once enabled, users will be able to open a new window or tab, or visit Google's homepage, and conduct a voice search. Saying "Ok Google" will activate the feature, and the software will then listen to what is said afterwards and treat it as a search query.

Chrome and Google have offered voice search for some time, but before this would require the clicking of a microphone icon next to or near the search bar. With the new update, no user input is required to activate the voice search other than the phrase "Ok Google".

Those opting for the latest Chrome distribution will get the feature immediately, compatible with Windows, Mac and Linux. The command is also available in the latest version of Chrome OS.

The command word only works in English for now. Google has not specified when it might expand the feature to include other languages and regions.

Other updates that came along with the voice commands included bug fixes and improved stability.

For Chrome OS users, support for folders has been added to the launcher, enabling them to organise their apps on a main screen rather than having to search through multiple slides.

Also bundled in the new version for Chrome OS is captive portal detection, which will allow users to easily gain access to public Wi-Fi networks in places like cafes, hotels and airports.

Apple has reportedly splashed the cash on UK-based speech recognition company Novauris, as part of its ongoing work to improve the responsiveness of its voice assistant Siri.

The company has been going since 2002, and specialises in technology, dubbed NovaSearch, that uses voice commands to help people sift through large databases for information.

"It does this by coupling novel, advanced speech recognition with proprietary, fast search techniques," the Novauris website states.

"NovaSearch provides core technology for applications over telephone and data networks and in embedded devices.

"By providing convenient voice input together with modest computation and memory requirements, NovaSearch enables voice-access applications on a wide range of platforms," the website adds.

At the time of writing, Apple was remaining tight-lipped about the acquisition, which was brought to light by a report on TechCrunch late last week.

Apple issued the publication with a short statement that said: "Apple buys smaller technology companies from time to time, and we generally do not discuss purpose or plans."

The deal is thought to have been struck during the second half of 2013.

The Novauris listing on Companies House gives an address for the firm that previously housed Apple's UK retail operations. The listing suggests Novauris moved into the premises around October time last year.

While Apple's plans for the technology remain sketchy, it's highly likely the Novauris technology will be used to bolster the capabilities of its voice assistant Siri.

The tool made its debut on the iPhone 4S in 2011, and has blazed a trail for similar technologies on mobile devices. For instance, arch-rival Microsoft debuted its Siri challenger Cortana last week.

Greater adoption of speech recognition software across the NHS could speed up diagnosis and treatment times, but integrating the tools with legacy IT systems is preventing this from happening.

Several healthcare professionals shared their experiences of using the firm's tools and the benefits they bring to clinical environments, at an event organised by speech technology vendor Nuance in central London yesterday.

Nuance also used the event to showcase Florence, the firm's healthcare-focused virtual assistant, which can respond to doctor requests for patient data and issue prescriptions based on voice commands.

Dr. Ivan Camphor, a senior partner at the Heatherlands Medical Centre, told attendees how speech recognition software has transformed the way his GP surgery operates over the past two years, especially when it comes to writing up appointment notes.

This is a process all doctors are required to do by the General Medical Council (GMC), he explained, so that they have an account of what's been said to a patient should their judgement ever be called into question.

"Most GPs use two fingers to type and it takes them forever to do," Camphor explained.

However, using Nuance's Dragon dictation software makes it easier for doctors to keep track of exactly what the patient has said during their appointment by converting their speech into a written record.

"I have freedom of talking into something and it would deliver it word perfect into EMIS Web [a software suite widely used in general practice across the UK] and more importantly it would be done with the patient sitting there next to me and they can correct me if I make a mistake," Camphor explained.

"It's a communal thing. It's not just me writing the patient record. We're writing it together, and affirming that is the true record," he added.

The software has also sped up the time it takes to issue referral letters for patients, as these would traditionally be dictated by a doctor onto a cassette tape and then transcribed by a medical secretary.

Now, doctors can dictate the letters directly into the web-based Choose and Book service, which is the system used to process referral requests, and send them off while the patient is still in the room.

"What this has done is made people redundant in Primary Care. The old fashioned narrative of a medical secretary working in a surgery no longer exists. My secretary has to find alternative things to do and this is where the true potential in all this is," he said.

"Primary Care is expanding and developing and there are increasing demands, financially and...in other areas, and GPs need to expand but they can't do that without resources. This releases resources in Primary Care."

Up to now, adoption of voice recognition tools by the medical community has been slow, with Camphor claiming the technology has not been marketed well enough and that some people have tried similar systems in the past and been disappointed with the results.

Speaking at the event, Dean Harmse, departmental head of dermatopathology at Plymouth Hospital NHS Trust, said take-up has also been hindered by problems integrating the tools with existing IT systems.

His organisation received a 50,000 NHS innovation award for its use of speech recognition technology to reduce the time it takes to analyse tissue samples. This in turn has sped up the diagnosis of medical conditions.

"The biggest challenge, in my opinion, are integration problems with existing software. That will stop [the use of this software] dead in its tracks," he warned.

However, this is a challenge that can be overcome, though, with the right deployment partner, he added.

Arguably, Siri is one of the most powerful an innovative tools built into the iPhone product in years. The only problem is, no one wants to look like an idiot talking to their phone – and so the product isn’t used.

I decided to challenge the embarrassment of talking to my phone in public and got to grips with using Siri for day-to-day business actions.

With a quick visit to the Apple site and a few Google searches, I was well on the way to making full use of Siri functionality. And do you know what? I found out that it’s a heck of a lot faster than using the iPhone in the usual way. After several hours of practice, I was sending emails, dictating notes and updating Facebook far quicker than I could with the use of my digits. (I even dictated large parts of this article into an email that I then sent myself)

So put aside the embarrassment and follow this simple how to guide to put Siri to work on a wide array of business processes.

Dictate a lengthy email message or note

Using Siri functionality with a few simple commands it is possible to create long copy detailed notes and emails with perfect punctuation, grammar and paragraphs.

Below is a list of commands to add punctuation, spaces and paragraphs:

• question mark = ?
• exclamation point = !
• dot/point/period = .
• comma = ,
• next line = same as tapping the enter/return key once
• next paragraph = same as tapping the enter/return key twice
• quote/end quote = “”
• cap = capitalise the first letter of the next word
• caps on = capitalise every word in the sentence
• all caps = capitalise every letter in the word
• apostrophe = ‘
• open parenthesis = (
• close parenthesis = )
• dash = -

Below is a to-do list I created and the phrase I used to create it:

In the notes section I tapped the microphone icon on the bottom row and dictated the following to-do list (I could make it longer but I think it’s fairly self explanatory):

“to-do list next line check emails next line call Steve Jobs” and here’s what I got:

Set a meeting reminder

Got an important meeting you don’t want to miss, or a deadline you need to meet. Siri can create that for you in a matter of seconds.

1. Hold down the “Home Button”
2. Speak into the phone “set reminder (title of reminder)”
3. Siri will then prompt you for the date you wish to set the reminder “(date month)”

Here’s an example of what I created: “set reminder meeting with John Smith”, Siri them prompted with the date and I responded “first of July”.

Set an alarm

Want to set an alarm for an early flight or a meeting you wish, it’s simple.

1. Hold down the “Home Button”
2. Speak into the phone “Set an alarm for Xam/pm”

Send a Text

If you want to send a quick text or don’t wish to type it out for fear of poor spelling errors, with a few simple commands you can dictate a message that will then be sent.

1. Hold down the “Home Button”
2. Speak into the phone “Send a text to (insert contact) (text message body)”
3. Send the text by saying “send”, cancel the text by saying “cancel”. Alternatively use touch functionality on the appropriate button to send or cancel.

In the below example, I used the phrase: “send a text to John Smith. John we have a meeting tomorrow”

Find a location

If you need to find the location of an office, or you’ve only been given a postcode, using Siri you can quickly bring up the Maps application with the location of your requested target.

1. Hold down the “Home Button”
2. Speak into the phone “Find (name of location)”

In the below example, I used the phrase: “find the Tower of London”

Get directions

If you are on the way to a meeting or event and find you are hopelessly lost, Siri can bring up the Maps application and give you directions to your destination of choice, with verbal commands, TomTom style.

1. Hold down the “Home Button”
2. Speak into the phone “Get directions to (name of location)”

In the below example, I used the phrase: “Get directions to the Tower of London”

Find a Contact

This is great for people like me who have 2000+ contacts in their phone and don’t want to endlessly scroll or use the search functionality.

1. Hold down the “Home Button”
2. Speak into the phone “Find contact (name of contact)”

In the below example, I used the phrase: “find contact John Smith”

Access an App

Need to open an application you use for work, it’s quicker than finding it.

1. Hold down the “Home Button”
2. Speak into the phone “Open (name of App)”

For example, if you say “open yammer”, surprise surprise the Yammer application on your phone will open. Or say “open app store” and the App store will open.

Send an Email to a contact

Following the dictation guidelines I gave earlier, you can quickly create an email for a selected contact and send a message. Whether it be short and sharp or long winded.

1. Hold down the “Home Button”
2. Speak into the phone “ Send an email to (name of contact) subject (name of subject)”
3. Siri will open up an email and then request the content of the email which you can then dictate.

For ease of the example below, I used a quick email which used the following phrase: “send an email to John Smith subject meeting”. When Siri opens the email I said “Don’t forget we have a meeting”

Create and find notes

Create

Sat in a meeting and you want to take some notes or out and about and you want to remember a genius though. Bring up Siri and with a few simple words it will remember your thoughts, so you don’t have to.

1. Hold down the “Home Button”
2. Speak into the phone “Take a note (content in body)”

As before you can dictate at length or take short sharp notes, in this example I used the following phrase: “Take a note, this meeting was awesome”

Find

Know you took down some notes in a meeting and can’t remember quote where they are? Search your notes by doing the following:

1. Hold down the “Home Button”
2. Speak into the phone “Find note (content that may be in the note)”

So I had a to-do list in my phone and conducted the following search, “find to-do list”. Up pops my to-do list.

Conduct a search on the internet

You can use Siri to quickly bring up Safari and conduct a search for you. The search functionality automatically defaults to Google; and not Bing, so it’s likely you will find what you are looking for.

1. Hold down the “Home Button”
2. Speak into the phone “Search (topic of search/name of company etc.)

I said “Search Apple”, Siri searched Google for Apple. So simple and so quick.

Update a Facebook or Twitter status

Ok this isn’t perhaps work related, unless of course you manage the company’s social media profiles for Twitter and Facebook, and happen to have the applications on your phone (long shot I know)

To tweet:

1. Hold down the “Home Button”
2. Speak into the phone “tweet (content of tweet)”

To update a Facebook status:

1. Hold down the “Home Button”
2. Speak into the phone “Update Facebook (content of content of status)”

So here’s how I did it:

Twitter:

“Tweet having a rubbish day”

Facebook:

“Update facebook having a great day”

So there you have it, a whole host of functions to make you more pro-active in the work environment. Have I missed any vital ones? Let me know by getting in touch.

Edward Jones works for provider of IT and Project Management certification Firebrand Training. Edward writes on a series of IT and project management related topics including How to Guides, product reviews and industry news.

Telecoms research firm TeleGeography's annual state of the market report has shown that voice over IP (VoIP) technology has become the method of choice for making international voice calls.

In 2011, total international voice traffic grew 9 per cent. However, when broken down by type of technology, traditional phone-based traffic grew 3 per cent, to 317 billion minutes, while VoIP traffic grew 25 per cent, to 150 billion minutes.

This trend towards cloud-based communication was even starker in 2012, with TeleGeography estimating global traffic grew 5 per cent to 490 billion minutes, 34 per cent of which were transported as VoIP.

"While market demand for cross-border communications has not declined, hundreds of millions of consumers have discovered that they can communicate without the service of a telco," said TeleGeography.

"The share of international traffic routed via computer-to-computer VoIP services has skyrocketed. Cross-border traffic route via Skype, by far the largest provider of over-the-top' (OTT) communications service, is projected to grow by an astonishing 51 billion minutes in 2012," it added.

TeleGeography also said, while Skype is "long past its start-up phase, its traffic growth remains astonishing", estimating its international traffic to have grown by 44 per cent (51 billion minutes) in 2012. This is more than twice the volume growth achieved by all the phone companies in the world combined.

The organisation said: "Not all of Skype's traffic represents a loss for telcos. Over 40 per cent of Skype's traffic is now video, and it is likely that a meaningful share of this is new' traffic, rather than a direct replacement for a telephone call.

"However, given their enormous traffic volume, it is difficult not to conclude that at least some of Skype's growth is coming at the expense of traditional carriers. If all of Skype's on-net traffic had been routed via traditional telcos, global cross-border telephone traffic would have increased 14 per cent in 2011 and 13 per cent in 2012 in line with growth rates of a decade ago."

Microsoft has demoed new software that translates English into Mandarin in real-time, copying the speaker's own voice.

In a video posted online, head of Microsoft research, Rick Rashid, demonstrated how the software works. The clip shows Rashid speaking in English with a Mandarin translation spoken a few seconds later copying the tonality of his own voice.

He said the company hopes to have "systems that can completely break down language barriers" in the next few years.

At the core of the software is a technology called Deep Neural Net (DNN) translation. The technology has already been offered as a commercial service called inCus.

The system records the words in English, translates them into Mandarin and then matches them to a recorded set of words in that language.

Rashid said in a blog post that Microsoft was poised to take the system one step further.

The demo was performed at Microsoft Research Asia's 21st Century Computing event. He said his team had been able to reduce the word error rate for speech by over 30 per cent compared to previous methods.

"This means that, rather than having one word in four or five incorrect, now the error rate is one word in seven or eight," said Rashid.

"While still far from perfect, this is the most dramatic change in accuracy since the introduction of hidden Markov modelling in 1979, and as we add more data to the training we believe that we will get even better results."

He said that we would not have to wait until the 22nd century for a usable equivalent of Star Trek's universal translator.

"We can also hope that as barriers to understanding language are removed, barriers to understanding each other might also be removed," he said.

The demo can be watched in the video below.

Comms services provider Kcom claims to have saved the NHS Business Services Authority (NHSBSA) 600,000 by overhauling its European Health Insurance Cards (EHIC) renewals service.

UK residents travelling to European Economic Area countries have to apply for an EHIC, which entitles them to receive state healthcare for free or at a reduced cost.

The cards need renewing, via NHSBSA, every five years, and this process was previously handled by call centre operators at a cost of 1.40 per application.

However, since Kcom developed a voice recognition system to ease some of the financial burden on NHSBSA, it is claimed the cost of renewing an EHIC has fallen by 70 per cent.

The company also claims, in the two years since the system was first installed, it has saved NHSBSA around 600,000.

Suzette Bouzane-Meadows, head of contact at Kcom, said: "We've designed a service that is focused on improving the customer experience as well as delivering significant cost savings.

"We'll continue to work closely with NHSBSA to further develop the solution to meet the increasing demand," she added.

Brendan Brown, head of customer service and innovation at NHSBSA, said the system has made it easier for people to apply for EHIC.

"There's less [pressure] on the operator to assist. As a result, the cost savings have continued to be impressive, as a majority of the transactions are done via self-service," he explained.

Google has beefed up its Android Voice Search feature by adding support for a further 13 languages bringing the total supported to 42 and extending its reach to some 100 million additional users.

Speed is key when it comes to trying to search for places using voice recognition on a mobile device, according to Google which claims it is catering for that demand with the enhanced arrival.

The updated software is already making its way to users and will improve in accuracy the more people use it, thanks to Google's cloud-based approach, according to product manager Bertrand Damiba.

"'Norwegian restaurants in New York City.' I can type that phrase fast, but I can say it even faster - and when I'm on the go, speed is what I'm looking for," he said in a blog post.

"With Voice Search, you can speak into your phone to get search results quickly and easily... Each new language usually requires that we initially collect hundreds of thousands of utterances from volunteers and, although we've been working on speech recognition for several years, adding these new languages led our engineers and scientists to tackle some unique challenges.

"While languages like Romanian follow predictable pronunciation rules, others, like Swedish, required that we recruit native speakers to provide us with the pronunciations for thousands of words. Our scientists then built a machine learning system based on that data to predict how all other Swedish words would be pronounced."

Multilingual users need not get too excited just yet, however, as Google has stressed that the app can only handle one language at a time so users will need to tweak their language settings each time they use a different lingo.

Nuance Communications has launched a virtual assistant service to help bolster the appeal of customer service-type mobile apps by making them respond to users' voices.

The Nina service uses speech recognition and voice biometrics, and Nuance claims it is the first product of its type to understand what users say, as well as who said it.

The cloud-hosted service is aimed at people who make customer service apps, who can embed Nina in their products using Nuance's open software development kit.

App designers can also customise the look and feel of Nina to suit their own branding, regardless of whether they work with the Android or iOS mobile operating system.

The offering has already reportedly been adopted by US financial services provider USAA. The firm plans to pilot the service within its mobile app this month before rolling it out to customers early next year.

Robert Weideman, executive vice president and general manager of Nuance's Enterprise Division, said Nina will allow developers to introduce another level of interactivity into their apps.

"Nina provides our customers [with] a major competitive differentiator by enabling more successful self-service through their mobile apps," he added.

The product is available now in the UK.

Despite the initial furore surrounding its infamous antenna issues, the iPhone 4 is undoubtedly Apple's most successful smartphone so far and it's still the standard by which we judge all other smartphones here at IT Pro.

The new iPhone 4S therefore has a lot to live up to. The vast majority of the new features in iOS 5 and iCloud are shared between both iPhones, but there's still plenty in the new phone that leaves a favourable first impression, despite the fact that both iPhones look almost identical externally.

Wireless

The controversy surrounding the iPhone 4's antenna issues need no introduction for most of us. According to Apple, the iPhone 4S can automatically switch between its two antennas for better call quality whether this now puts the antenna issue to bed remains to be seen.

The iPhone 4S isn't an LTE-capable phone so it's not 4G, but it does have faster 3G speeds - 14.4Mbps download and 5.8Mbps upload. Whether we here in the UK will ever achieve those speeds depends on our mobile phone networks and their infrastructure. Frequent travellers to the US will appreciate the fact that the iPhone 4S works on both GSM and CDMA networks, unlike the iPhone 4 which came in two separate versions for each standard.

Voice control

The latest Android smartphones have fairly sophisticated dictation abilities, although its efficacy greatly depends on ambient noise and the speed and reliability of your data connection since the actual voice recognition is carried out by Google's servers. By comparison, the iPhone 4 and iOS 4 are limited to a handful of simple phrases for dialling phone numbers by voice or controlling music playback.

The new Siri feature on the iPhone 4S looks impressive. It's a natural language interface that should be able to understand and reply to most questions, even if they can be worded in numerous different ways. In a demonstration at Apple's press conference, Siri was used to check an incoming text message, check the calendar for any upcoming appointments on a particular day before replying back to the sender. Android-style dictation is also available.

Siri is likely to be exclusive to the iPhone 4S - its dual-core A5 processor is almost certainly needed for Siri to work. Even if Siri doesn't catch on with the general public, it's likely to be greatly appreciated by the blind and visually impaired.

Processing power

It was always a forgone conclusion that the 1GHz single core A4 processor used in the iPhone 4 would be superseded by an iPhone equipped with the dual-core A5 processor as seen in the iPad 2. The benefits of the A5 are clear it allows the use of apps that are off-limits to the original iOS devices, such as Garageband and Siri. What's not clear is if the iPhone 4S A5 is clocked at 1GHz just like the iPad 2.

Cameras

We have to admit that we were hardened sceptics about the utility of camera phones when they first appeared on the market seven years ago. Now, we're the most prolific smartphone camera users the most useful camera is the one you have on you. Especially if you can share photos simply by uploading to Flickr and Facebook, or shoot and edit HD footage then and there. The iPhone 4's five megapixel camera doesn't have dual-flashes or shoot in 3D, but it produces some of the most balanced looking images around.

The iPhone 4S has an eight megapixel rear-facing camera that can also shoot 1080p HD video as opposed to the 720p video recording found on the iPhone 4. The improvements aren't just about higher specs though. As well as features such as image stabilisation, noise reduction, face detection and better white balance, Apple claims the iPhone 4S camera is better at producing good looking pictures in dimly lit conditions than its predecessor. Apple claims the iPhone 4S camera is faster at taking a new photo from sleep and taking multiple shots one after the other - 1.1 and 0.5 seconds respectively which is around half as long as the HTC Sensation according to Apple.

Storage

Unlike most other competing smartphones, the iPhone has never had a memory card slot for easily adding more storage. This is needed now more than ever, since the iPhone 4S can shoot 1080p video which can easily gobble up lots of storage if you're not careful. The new 64GB iPhone 4S goes some way to compensating for this though the iPhone 4 had previously been limited to a maximum of 32GB.

Screen

One of the most impressive things about the iPhone 4 is its Retina Display. Rather than using the 960x640 pixels to simply display more information, Apple chose to use all those pixels to create incredibly smooth onscreen text and images that's very easy on the eyes. Only the 960x540 pixel screen of the Android-powered Motorola Atrix comes close. The iPhone 4S has the exact same screen which is to be expected smartphone screens with even higher pixel densities would likely be prohibitively expensive and difficult to produce, if they even exist at the moment.

Appearance

One of the most controversial aspects of the iPhone 4's design, besides is antenna, is its all-glass design. Although it looks and feels seductively great and despite Apple's claims of extreme durability, we've seen more than a few examples of easily cracked iPhones. Rumours of a MacBook Air-style unibody design will probably never come to fruition given HTC's use of a similar design in its smartphones. The look of the iPhone 4S is almost identical to the iPhone 4, so we don't expect any changes here. In other words, a case will probably still be a good idea.

So what are our initial impressions?

Verdict

The iPhone 4S isn't radical enough to be called the iPhone 5, but it's a compelling upgrade, especially for switchers and owners of the older iPhone 3GS. The Siri natural language voice control and dictation system look impressive and is highly likely to be exclusive to the iPhone 4S and its faster, dual core A5 processor – an example of Apple's tight hardware-software integration. For iPhone 4 users, it's trickier dependant on how much you'll use it, whether it’s a SIM-free purchase, or a contract upgrade. The 16GB iPhone 4S will cost £499 ex VAT SIM-free with pricing for the 32GB and 64GB models yet to be confirmed. The new phones will go on sale on October 14th, although pre-orders will be available on October 7th. We'll bring you a full review as soon as we can.

Google has made some significant changes to its search business, making it faster to access results and adding speech recognition.

The Silicon Valley giant yesterday announced Instant Pages, which prepares the top search result in the background whilst the user chooses which website to visit.

This will save users another two to five seconds on typical searches, Google said.

"Instant Pages will pre-render results when we're confident you're going to click them," said Google Fellow, Amit Singhal, in a blog post.

"The good news is that we've been working for years to develop our relevance technology, and we can fairly accurately predict when to pre-render."

Instant Pages will feature first in the next beta release of Chrome, although the developer version has been made available today.

Google has also brought the voice search capability it introduced to mobiles into the desktop.

"If you're using Chrome, you'll start to see a little microphone in every Google search box. Simply click the microphone, and you can speak your search," Singhal said.

Speech recognition on desktop is rolling out now on Google.com in English, he confirmed, claiming Google's voice search traffic has grown six-fold in the last year alone.

Google also introduced Search by Image, which lets users upload any picture or plug in an image URL to figure out what it is.

The feature is also rolling out from today. It can be accessed via a button next to the microphone on images.google.com.

Apple is a very secretive company. Usually, it refuses to even confirm whether it's working on a new product right up to the Steve Jobs keynote when it's publicly announced. It's a little different for iOS 5, the next version of the operating system for the iPhone and iPad, which Apple acknowledges will be unveiled at the Worldwide Developer's Conference (WWDC) on Monday June 6th less than a week away.

Jobs and co still aren't divulging any specific details in advance though, fuelling speculation about the new iOS release even more than usual. Although iOS is a pretty mature mobile operating system, there are still some things we'd personally like to see added or addressed in version 5.

Better Notifications

The way notifications works in iOS has remained virtually changed since the release of the original iPhone back in 2007. When you receive a notification, whether it's a SMS, alarm or an alert from an app, it appears on screen as a modal dialog box that must be dismissed before you can do anything else. If lots of alerts come in all at once, the earlier alerts are silenced with only the most recent alert visible until you dismiss it.

The simplistic notifications in iOS are in need of a drastic overhaul

It's a very clunky, clumsy system, especially when compared to the pull-down notifications drawer in Android and the webOS pop up system both of which queue up notifications for your perusal. This iPhone-using writer has missed more than one alarm for an important event after it was silenced by a notification from another app.

A freelance designer has blogged extensively on the topic and has even come up with his own notifications system for jailbroken iPhones.

Accessible file system and smoother file handling

Unlike Android, the iOS file system isn't user accessible despite the fact it's based on Apple's desktop operating system, MacOS X. Any files used by apps are stored within the app itself, so the developer has to come up with their own file manager. Files can be exchanged between apps by tapping the Share/Open in button which then shows options for opening that file in another app.

Sharing files between apps is quite clunky at the moment. There has to be a better way.

Getting files onto an iPhone in the first place isn't straightforward either. You have to either email files to yourself or drag and drop files onto your iPhone using iTunes over a USB connection. Alternatively, you could use an online storage app, such as Dropbox, but this might not be ideal for your business or your favourite apps might not support it. Plus, this gets tedious very quickly if you're collaborating on documents shared with others.

We wouldn't expect Apple to open up the file system completely, but surely there has to be a better way. Speaking of the cloud

Cloud integration

We were initially skeptical about the cloud, if only because of the initial hysterical hype, but we've now been won over by useful cloud-based apps and services such as Dropbox, Simplenote and Instapaper.

Cloud services on iOS could be so much better with a little help from Apple

However, iOS remains stubbornly cloud-free. Apple's own MobileMe paid-for cloud service for synchronising calendars, contacts, email and files is only superficially integrated with the operating system. It also has, anecdotally at least, recurring reliability issues.

Apple has confirmed that it will be unveiling details about its new iCloud service alongside iOS 5. Although much of the speculation surrounding iCloud centers on music streaming, we have high hopes its scope will be more ambitious than that if only because of the size of Apple's new data centre.

However, to make the most of the cloud, Apple should also tweak how iOS manages multitasking. Although third-party cloud-dependent apps can continue downloading new data in the background, they can't check for new data automatically you have to open the app and manually tap a refresh button. An authorised way for apps to check for and download new data in the background is sorely needed tapping half a dozen apps to check for updates is not our idea of fun.

Integrated and expanded voice recognition

One of the few impressive features in Android is its integration with Google's cloud-based voice recognition service. You can use this speech-to-text dictation for almost anything where you'd use a keyboard, from searching the web to composing a text. It's a bit finicky, since its efficacy greatly depends on a quality network connection and your phone clearly recording your voice, but it's still a feature iOS currently lacks.

The rain in Spain stays mainly in the plain...

iOS 4 does currently have limited voice controls you can dial contacts and control music playback, but that's it. Hopefully Apple has more sophisticated voice controls in store, especially given rumours of a licensing deal with voice recognition specialists Nuance and Cupertino's acquisition of voice app Siri.

Useful and sophisticated Bluetooth features

Unless you use a Bluetooth headset, it's likely Bluetooth remains off on your iPhone. Bluetooth can be used for much more than hands-free headsets though and we hope Apple makes better use of this short-range wireless technology in iOS 5.

Apple is no stranger to advanced Bluetooth features. Users of Bluetooth-equipped feature phones paired with older versions of MacOS X could text and dial using Apple's Address Book application and their full-size keyboards, rather than fiddling about with a small keypad.

Third-party developers took this even further. The Salling Clicker program allowed you to use your feature phone as a remote control for iTunes, PowerPoint and other programs. This has since been replicated in plenty of remote control apps, but that wasn't Salling Clicker's only trick. If you wandered out of Bluetooth range of your computer with your phone, iTunes would pause playback and the status message in the iChat IM program would change from available to away.

It's this sort of clever stuff and attention to detail which we hope Apple will add to iOS 5. We'll find out soon enough on June 6th when Steve Jobs delivers his keynote address at the WWDC. What would you like to see in iOS 5? Let us know in the comments.

The idea of being able to talk to a computer rather than type has long been a pipe dream, and not just for people who can't get to grips with a keyboard. Sadly, the technology has yet to reach a point where people can converse freely with a PC and have it understand what they're saying, but speech recognition for more mundane dictation purposes has been possible for quite a few years now. Nuance leads the field for desktop speech recognition and Dragon Dictate 2.0 is the latest version of its software for Mac OS X.

As the back of the Dragon Dictate 2.0 box points out, most people can talk much faster than they can type, but speech recognition isn't quite so quick to set up as just plugging in a keyboard. Automated telephone answering systems may be able to respond to a limited selection of spoken commands from just about any caller, but the speech recognition offered by Dragon Dictate is designed to recognise open-ended verbalisation for text transcription purposes, and so requires some initial voice training with the person using it.

Although speech recognition software employs a range of sophisticated audio, contextual and statistical analysis techniques to cope with homonyms and other tricky aspects of language, the basic idea involves matching elements of speech to the contents of a word database. By asking the user to speak the word "dog", for example, the software knows that whatever noise the user makes in response can always be interpreted as "dog" and displayed in type accordingly. This is where voice training comes reading on-screen text aloud so that the speech recognition software can get a measure of how a user says certain words, the cadence of their speech, and so on.

With this initial training out of the way, Dragon Dictate 2.0 can then turn speech into words within just about any application that would otherwise rely upon a cursor for input. The application runs in the background with just a floating status window to indicate its presence.

Dictation works extremely well. Best results are achieved with the clear, clipped and well-modulated vocal style of a TV newsreader, but Dragon Dictate is quite adept at dealing with less-than-perfect diction and nor is there any need to speak particularly slowly.

Dictation is as much of a skill as typing though, and maintaining a clear line of thought while simultaneously speaking it aloud takes practice pausing mid-sentence, or speaking in a halting way while an idea coalesces, is a recipe for poorly recognised speech. The need to verbalise all punctuation marks comma for example comma can also be a distraction open bracket try reading this sentence aloud to see how this works close bracket full stop. Punctuation can be ignored during dictation and then added later using the keyboard, of course, but that does rather defeat the point of speech recognition.

Vocabulary training can also be performed to extend the range of words that Dragon Dictate 2.0 knows about. The application already correctly recognises some technical terms such as "TCP/IP" and "DVD-ROM" without the need to specify capitalisation and punctuation. It can be further trained to identify more esoteric jargon by importing a suitable text document and performing voice training for the words that aren't in its database.

In addition to straight dictation, Dragon Dictate 2.0 can also be used for text editing, but the long list of specific phrases required for this (necessary to distinguish between dictation and spoken commands) make it a slow substitute for editing with the keyboard. Unfortunately, switching to the keyboard mid-dictation can cause more problems, since this puts a document's content out of sync with what Dragon Dictate 2.0 has recognised so far. This don't mix inputs' rule is spelled out in the manual, but the problem can be bypassed by uttering a command to rebuild the application's internal document cache after a keyboard edit to make it aware of the new, manually altered, structure.

That said, the option to create custom commands for anything from web bookmarks and menu selections, to text macros and AppleScripts, makes setting up a highly personalised voice-activated Mac OS X environment relatively straightforward although it requires no small amount of time and effort. Fortunately, there is a shortcut to this learning curve for voice-activated computer control voice-activated mouse control. Simply speaking "move mouse left, faster, stop, double-click" is as easy as it gets, but there's also the option to quickly select anything on any part of the screen using a pop-up grid overlay.

Dragon Dictate 2.0 is an impressive speech-recognition application that offers a prodigious amount of functionality for anyone who wants dictate text rather than type it, or use voice control rather than more traditional input methods. Its performance with minimal training is impressive too, but the best results are achieved after spending time and effort to master its intricacies which means it's no quick fix for anyone who hasn't the time or patience to master a mouse or keyboard.

The only real grumble is the price and Mac users are required to pay appreciably more for essentially the same software as Windows users. So, while Dragon NaturallySpeaking 11 Premium for Windows is available online for around 70 + VAT, the best online price for Dragon Dictate 2.0 for Mac OS X is around 109 + VAT. This is at least partially due to the fact that Dragon Dictate can't be bought (inexpensively) without a headset, something that is possible with NaturallySpeaking.

So what's our verdict?

Verdict

Impressive recognition performance and a vast array of features make Dragon Dictate 2.0 a flexible and highly capable speech recognition application for Mac users, but it’s a shame about the Mac OS price premium.

Processor: Intel processor

Operating system: Mac OS X 10.6 Snow Leopard or later

Internet connection required for registration.

Nuance-approved USB microphone for Macintosh (included with boxed software).

Even when there isn't a scare about whether hardware tokens like RSA's SecureID might have been compromised, two -factor authentication still means a physical device your users can lose as well as a password they can forget. Biometric authentication is more convenient - it's hard to leave your finger or voice at home. It's also potentially more secure employees are less likely to log someone else in with their own biometrics unless there's a serious threat to their safety, which is a different problem entirely.

Fingerprint-based biometric solutions are the most common form. Many business laptops and keyboards have a fingerprint reader built in that you can use for federated single sign-on to enterprise apps as well as just logging into Windows and saving passwords for websites. Fingerprints can be used as an additional security measure that doesn't interrupt what the user is doing the way having to swipe a token or smartcard again does. If you want to make bankers re-authenticate to authorise a stock trade over a certain level, you need to do it quickly enough that the trading situation doesn't change.

Other systems scan vein patterns in the finger (Hitachi) or the palm of the hand (Fujitsu). They can still recognise a user if they have a cut on their finger - dirt and antibacterial hand gel also cause problems for fingerprint readers in industrial locations - but you'll need to buy, install and integrate them. Having users register a couple of fingers on each hand is usually enough to avoid the issue in a business environment.

Facial recognition and iris scans still tend to be either too unreliable or intrusive for general use and voice print recognition has some interesting applications, especially for customer-facing operations that have the scale to make the cost and complexity of implementation worthwhile. But the majority of biometrics in use today involves fingerprints for PCs and many other devices.

Every prescription written in the state of Ohio requires strong authentication and one of the popular methods is fingerprint recognition. That's carried out on the PC but adding fingerprint authentication to printers means confidential documents won't be sitting waiting on the printer for anyone to pick up as they walk past, and swiping a finger is less irritating for the kind of senior staff who need to print such documents rather than having to remember a PIN or token. You'll need to integrate this with a document security system. As an example, Ricoh recently added a fingerprint option (for both new and already installed printers) that works with Equitrac document protection software and BioStore's identity management system.

Biometrics aren't just useful for protecting confidential data. Construction firm Killby & Gayford is using custom fingerprint readers (built for personnel management software supplier Simeio by Psion) that also record a signature. That means the firm can accurately log hours worked by sub-contractors on building sites and automatically generate invoices, as well as proving that workers have read the safety rules for each site and simplify checking the roll call if there's an emergency. Using the information to generate accurate invoices has avoided concerns about spying on staff by showing the system is useful to everyone. But, equally, productivity has improved too.

Convenience is a big part of the attraction for healthcare, along with security and compliance. "A doctor logs on and off 20 or 30 times a day," points out David Ting, the chief technology officer (CTO) of Imprivata. "If I only need to log in occasionally, I'm prepared to put up with any login method. But if I have to do this repeatedly, give me something that is secure enough but at the same time make it easy for me."

Retailers and restaurants are increasingly deploying biometric point of sale (POS) terminals to improve productivity, especially in situations where there are a lot of cash sales and temporary staff the usual alternatives of video surveillance and analysing transactions tend to be expensive, time consuming and don't make employees feel personally accountable.

Tokens are easy to pass back and forth between staff (and expensive to replace if an employee takes one when they leave) and passwords need to be reset regularly because of staff turnover. In addition to recording hours worked more accurately, the accountability of biometrics discourages unauthorised transactions or outright fraud. Hospitality businesses often see a return on investment in as little as 30 days. It also helps the business comply with PCI regulations for handling credit cards, which include giving each employee a unique ID.

Many cash registers let you connect a fingerprint reader as an internal module or a peripheral. Hertfordshire pub chain Giant Macaskills recently installed Unipower's touchscreen Bar POS system in its bars and Unipower integrated Digtial Persona U.are.U fingerprint readers into the terminals. DigitalPersona has a OneTouchfor Windows SDK that simplifies adding biometrics to existing apps, and the DigitalPersona Pro server is available both as a workgroup system that supports authentication through a web service without Active Directory (AD) or requiring devices that are connected to a domain.

There's also an enterprise version that integrates with AD by extending the schema, which can be centrally managed through group policy along with the rest of your systems. That means businesses can set the same identity and authentication system in the retail environment and the back office for invoicing and accounting. German bank Sparkasse has recently used the DigitalPersona system to move all internal users to biometric authentication for Windows as well as hundreds of applications.

The alternative is an appliance like Imprivata's OneSign which works with identities from AD and LDAP. You need to deploy agents to PCs and set up policies to enforce biometrics but you don't have to make any changes to the directory infrastructure or AD schema, which some companies find disruptive.

Fingerprint and hand biometrics have another advantage; you can use them for access to more than PCs and devices. A unified authentication and access control system lets you limit entry to specific areas, or if you want to use contextual security that stops a user logging into a PC if they haven't actually been registered as coming through the door into that area.

That's also handy if someone tries the common trick of setting off a fire alarm to empty the building so they can get access to corporate PCs - people don't always stop to log off when they think there's a fire. A system like Overtis that links your physical security system to biometric access controls on computers will automatically lock them down in the case of an alarm sounding.

Voice biometrics are useful for remote authentication, like adding a level of security to self-service password resets - and with the spread of smartphones you can do it without the cost of buying devices like fingerprint readers.

HTK's Horizon Password Reset service, for example, only accepts calls from approved numbers and uses voice verification with pre-enrolled pass phrases. There are also fall back options of PINs and text messages for approval. It's a hosted solution that HTK can customise to integrate with AD or LDAP and the company is working on out-of the box integration.

Estimates of the IT cost to manually reset a password range from 30 per password to 92 per user per year depending on your environment. If you use fingerprint biometrics, you can avoid having a password to reset altogether.

If you want to try something much more experimental that could become mainstream fairly quickly, the Mobio project at the University of Surrey has facial and speech recognition apps for the iPhone aimed at improving security for consumers. Some call centres are also looking at using voice recognition to avoid fraud. ValidSoft can confirm user identity from a voice print and also confirm their location from the mobile phone mast they're connected to.

Although it takes greater investment in the backend systems, one advantage of voice biometrics is your users almost certainly already have smartphones you can use instead of having to buy dedicated hardware. This makes it a more acceptable choice for services as well.

"As smartphone ownership becomes more commonplace we're likely to see biometrics being deployed through these devices," said Alex Bazin, head of biometrics at Fujitsu. "This will most likely be using voice biometrics to secure access to cloud-based services."

Historically, biometric progress has been hindered by integration complexity and cost. There's also a psychological issue around storing and tracking users with such personal information.

But systems don't need to store live biometric data that could be retrieved and misused. Instead they store a detection pattern that the fingerprint is compared against which can't be used to generate a copy of the fingerprint. The database of that information should be stored in a secure and encrypted manner.

The way that the fingerprint recognition is presented has a big impact on how users feel about it. "If you have a system that shows a live video of their fingerprint users tend to bristle because it draws attention to the privacy issue," added Ting. For many users, as long as your system is set up in a secure and privacy respecting way and the authentication doesn't feel intrusive, the convenience of biometrics will more than outweigh any concerns.

Biometrics isn't the solution to every problem though. Sometimes it's overkill and you'll want a system that allows a mix of authentication methods. And it's down to businesses to decide what they want to achieve, according to Ting.

"You need to decide if your goal is to be ultra secure or to get a combination of security and productivity or just to make it more convenient," he added. "You have to think through what you're trying to secure. What is the impact on my user, what is the impact on my IT department?"

With increasingly simple and inexpensive biometric solutions, even small businesses look set to benefit. Once set up, a biometric-based solution gives you secure authentication with little on-going management, according to Benjamin Boulnois of DigitalPersona.

"If you have a 200 person company the IT manager has to be a jack of all trades. The question they ask is How do I do data security easily?'" he said.

"The average IT manager has 11 minutes of spare time a day. These are people whose job is not to be in security."

While some of the world's biggest brainiacs were left reeling following a Jeopardy! defeat by IBM"s Watson supercomputer, tech rivals in the search space can rest easy.

For Big Blue has no plans to use its supercomputing prowess to take on the likes of Google, according to Steve Mills,

"The Watson [project] had some unique challenges," he told a press conference at IBM's Pulse 2011 event in Las Vegas.

"How do you create mini Watsons, baby Watsons? We learned a lot as to what it would take to make it commercial."

IBM's Watson is clearly a computing behemoth - it is comprised of some 90 blade servers and boasts 15 terrabytes of storage - whose usefulness spans a wide variety of potential applications and sectors. One area it doesn't plan to tap into, however, is Google territory.

"IBM is out to make money on its technology...We're not interested in competing with Google in the business they're in," Mills said.

"It doesn't make sense to spend a massive amount of money to do some sort of super web crawling system. You've been watching what's going on with Bing," Mills was reported as saying by eWEEK.

Mills added: "Your brain is part of the Google experience. There's a market for that, obviously."

But it's clearly a market IBM isn't looking at right now.

One area Big Blue is looking at as far as Watson is concerned though is healthcare. The company confirmed it is already in talks with both US and European chemists and healthcare departments to see how a smaller-scale Watson could help them provide a better service.

The news follows a recent deal IBM signed with voice recognition giant Nuance, which aims to make use of Watson's hefty analytical capabilities in the healthcare space.

The partnership will bring together Big Blue's Deep Question Answering (QA), natural language processing, and machine learning expertise capabilities and Nuance's speech recognition and clinical language understanding (CLU) tools to aid the diagnostic and treatment of patients.

"Combining our analytics expertise with the experience and technology of Nuance, we can transform the way that healthcare professionals accomplish everyday tasks by enabling them to work smarter and more efficiently," IBM Research's Dr. John Kelly last month when the announcement was made.

"This initiative demonstrates how we plan to apply Watson's capabilities into new areas, such as healthcare with Nuance."

The first fruits of the deal are expected within the next two years.

IBM's Watson supercomputer has resulted in a huge boost to the tech giant's World Community Grid.

Watson recently came out as winner of a three-night long marathon of US gameshow Jeopardy!, beating two of the show's best-ever contestants.

The World Community Grid, which uses the collective spare power of computers from across the world to assist scientists and researchers, subsequently saw a 700 per cent spike in the number of volunteers who supply the additional juice.

The huge virtual supercomputer is designed to help those scientists working on humanitarian challenges, such as HIV/AIDS treatments.

To become part of the World Community Grid, users simply need to install a software program on their computers, as long as they are running Microsoft Windows, Macintosh or Linux operating systems.

"Watson's performance on Jeopardy! has captured the imagination of millions of viewers who understand the power of computing to benefit humanity," said Stanley S Litow, IBM vice president of corporate citizenship and corporate affairs and president of Big Blue's Foundation.

"Like Watson, World Community Grid is also a game changer. We're grateful for the skyrocketing interest in World Community Grid as a result of Watson's achievement."

Elsewhere, six different organisations will share $500,000 (308,000) prize money, which IBM has announced as part of the community project.

IBM has also teamed up with Nuance Communications to explore how Watson's analytics capabilities could be used in the healthcare industry.

The pair will explore how the supercomputing capabilities can be matched up with Nuance's speech recognition solutions to help diagnose and treat patients.

It's no secret that what Hollywood believes a computer is capable of often bears little resemblance to the real thing. Nonetheless, in an ideal world, what are the technologies from our movie and television screens that would actually prove useful in a real life scenario?

Here are a few of our suggestions - let's hear some of yours in the comments at the end.

Instant Updates

Surely the biggie. Hollywood blockbusters, when they're not insisting that a computer makes a sound every time a character appears on the screen, will happily show that everything you do on a PC or Mac happens in an instant. There's no hourglass, there's no waiting for a web page to load. It's click your fingers, and what you want is always at your fingertips. Marvellous.

Everything Is Easily Trackable

There's clearly an element of truth to this, but in the land of fiction, all you need to hunt down somebody's e-mail history, or deleted documents, is a character from the shadows. Where this person got their skills from is never explained, just that they can get you any piece of information that you need, and quickly. It comes back to everything being at your fingertips whenever you want it. And that very rarely happens.

Genuinely Innovative Touchscreen Interfaces

The biggest barrier between humans and computers in the real world has traditionally been the interface, whether it's the keyboard, mouse or on-screen user interface. Films such as Quantum of Solace and Minority Report have little time for clicking around a Windows desktop, though, instead opting for fast-moving touchscreen displays.

Non-Microsoft Software

It's very rare that you see an interface in a big budget movie that reflects that of the Windows operating system. Instead, big corporations in films and TV increasingly seem to have some kind of more tailored operating system that nobody ever seems to grumble about. The main advantage here for the real life IT department? Cutting out the cost of the Microsoft licence, easily one of the biggest outlays in modern day computing. In films such as Anti-Trust, the thrust is on a push to open source. And in the modern day office, that equals not insignificant savings on the budget

Voice Recognition That Works

Science-fiction is to blame for this one of course. How many movies do you see where someone barks out some exact orders to a computer, and it can both decipher it and enter into some kind of conversation with you? How easy would that make troubleshooting simple tasks, if you could just have a chat with the computer in question, and all would be fine and dandy? We're not just talking Star Trek here, either, as Hollywood increasingly seems to be believe that the future job of a computer is to sit down and have a chat.

Universal Programming Language

Swordfish, starring John Travolta and Hugh Jackman, is the example that springs to mind here. Never mind what hacker you see, or what programmer is trying to bash something together, there's never any moment where they have to stop and look everything up in a book. There's not even a moment of recognition, usually, of a programming language when they take a butchers at a screen of code. Instead, everything is instantly recognised, and no programming challenge takes more than three attempts to resolve. Marvellous. That'll cut down on the overtime.

Decreasing Reliance On Human Beings

This obviously has a massive downside, ranging from less people in work, through to Cylons rampaging through the galaxy and destroying the world. However, one of the upsides of omnipotent and omnipresent computers is that more and more, human beings get cut out of the chain. The biggest threat to computer security? That'll be the person using it, and no matter how tight a strategy is put into place, it's bound to be a human being who manages to break it somewhere along the line

Obvious Visual Cues When Something Is Wrong

As many network administrators can tell you, many threats to an infrastructure can lie undetected for some time before the payload is activated. Although this is more common in small offices rather than big ones, a piece of malware can lie undetected for some time, wreaking havoc before somebody picks up on it.

No such problem in the movies, however. Here, if your computer is infected, your computer will have some massive klaaxon to alert you to the fact. There's no messing around, no snooping or anything like that required. When the fit hits the shan, you're left in little doubt.

All The Vital Data You Need Can Be Held On One Disk

A prime example of this syndrome is Die Hard 4 where the villains are able to download all of America's financial data onto a Palm PDA There is a certain advantage to having absolutely any vital piece of information you may ever need fit onto whatever disk you happen to be carrying. The Disk Full' message appears to be an alien concept to the majority of films and TV shows, and that makes a heavy investment in bank after bank of hard drives a less important one. Again, there's cash waiting to be saved

A Big Stomping Robot

There's surely no IT department in the world that wouldn't, at some point, want the might of computing power to generate a (non-lethal) creation such as the odd robotic sight at the end of Superman III. That or, with a nod to Weird Science, the ability to design someone on their machine and have them come to life in front of them. Might not get any computers fixed for a while, though.

Have we missed anything out? Then add your thoughts below. Scarily, it's worth considering that some of the innovations that we've outlined above are either in some degree of development, or perhaps even already deployed in one form or another. The giant robot aside, of course

The reliability, accuracy and scalability of biometric security systems has been called into question by a scathing report from The National Research Council (NRC) in the US.

The "Biometric Recognition: Challenges and Opportunities" report has scotched the myth that biometrics are trustworthy and secure, describing them as "inherently fallible".

Although the results have created a stir in the security world, the report was produced as a scholarly overview of the science behind biometrics. Its conclusion that no single biometric trait has been identified as stable or distinctive has placed doubt about the reliability of fingerprint, iris patterns, voice recognition

and facial recognition systems.

NRC assembled a panel chaired by Joseph Pato, a distinguished scientist at HP Labs. The rest of the membership was drawn from industry, academia and the analyst community.

"While there are lots of good uses for biometric recognition, there are lots of ways to create systems that waste time, cost too much and don't work very well," said panel member Bob Blakley, a research vice president at analyst firm Gartner.

The news will come as further ammunition for the UK Government. The coalition halted plans for an autumn 2010 introduction of "second generation" passports that would have included fingerprint records.

Its argument was the need to cut costs but now the report would support an argument that it represented a bad return on investment.

The NRC report cautioned that basic research into the reliability of the physical characteristics has been ignored. These characteristics change with age, illness, stress and other factors.

"Bolstering the science is essential to gain a complete understanding of the strengths and limitations of these systems," Pato explained.

Biometric systems provide "probabilistic results", the report said, by which it meant that confidence in the results must be tempered by an understanding of these inherent uncertainties. Furthermore, variable results have stemmed from technical issues regarding the calibration of sensors, degradation of data, and fraudulent entries made through security breaches.

Deployment of biometric screening devices at airports and other checkpoints without understanding the biology or the population being screened would lead to long queues, delays caused by false positives and missed opportunities to catch criminals or terrorists, the report concludes.

The Motorola Milestone XT720 has some of the most impressive-sounding specifications of any Android smartphone currently available.

It has an eight-megapixel camera with Xenon flash, a large, high resolution screen, the ability to shoot 720p video, and an HDMI port for showing video and photos on an external display. The phone supports HSDPA download speeds up 10.2Mbps and upload speeds up to 5.76Mbps, as well as the usual Wi-Fi and GPS. Android 2.1 is installed so business users have the benefit of Microsoft Exchange support for email, calendars and contacts.

The only gaps in its specifications department are the 550MHz processor other top end smartphones have faster 1GHz processors and the lack of 802.11n support for WiFi.

Neither of these omissions bothered us during use. A more important limitation is the quality of the eight-megapixel camera, which did not live up to our expectations.

The camera has a fair few settings on hand including a multi-shot mode, face detection, a macro mode and touch focus so you can touch the screen to make the camera to focus on what you want it to. There's no significant shutter lag either.

Stills look reasonably good, but they didn't look significantly better than images taken on smartphones with five-megapixel cameras. The Xenon flash does a better job than the more common LED flash units of illuminating close-up subjects indoors. However, like all flashes, its efficacy diminishes the further away from the subject you are.

The 720p video shooting is peppered with pauses while data is processed. We found these pauses replicated in saved video, which makes viewing recorded clips an unpleasant and unsatisfactory experience.

Unlike the original Milestone, the XT720 doesn't have a slide-out physical keyboard. We weren't fond of the original Milestone's stiff, unresponsive keys, so we won't miss it. This does mean you'll have to rely on the onscreen touch keyboard, but thankfully it feels responsive and works well, especially when the screen is oriented horizontally as makes the keyboard bigger.

An alternative to the onscreen keyboard is Android's built-in speech recognition. We found it to be very good indeed at turning our spoken words into text for SMS or emails. There's a microphone icon on the keyboard and you just tap this to start the speech recognition engine. This speech recognition feature does require an internet connection though, since it relies on Google's servers to process your dictation into text.

The handset itself is very well made. The backplate is made of reassuringly solid metal coloured in a rather attractive shade of blue. The metal helps account for the 160g of weight that makes the Motorola Milestone XT720 noticeably heavier than other smartphones, but it is actually slightly lighter than the original Milestone.

The shiny front picks up fingerprints easily. There are four touch sensitive buttons beneath the screen for Home, Menu, Back and Search functions. A very peculiar design quirk of the chassis is that the bottom right edge bulges outwards making the Motorola Milestone XT720 wider at this point than elsewhere. We don't have any strong feelings about this bulge either way, but it might be enough to stop those with smaller hands reaching all the way across the screen when working one-handed.

On the edge at this point are the camera shortcut button and a button which cycles you through the media gallery, stills shooting and video shooting modes. Tiny icons light up to remind you which shooting mode you are in, as if the live on-screen preview wasn't enough.

At 3.7-inches the screen itself is relatively large for a smartphone. It has a resolution of 480x854 pixels which, although not quite as large as the iPhone 4's 960x640 pixels, is quite high. If we have a complaint it is that it is a bit tricky to see well outside in sunshine. This isn't unique to the XT720, but it could prove annoying.

Nonetheless, the bright screen is great for watching video. The high resolution is ideal for web browsing, since you can see more of the page at once, always a big plus. It's easy to zoom in and out of webpages using a pinching gesture since the capacitive touchscreen supports multitouch.

Unlike its previous Android phones, Motorola has chosen not to graft its own customised interface onto the XT720's Android 2.1 installation. This should mean that it can be upgraded easily to newer OS versions, including Android 2.2. However, although Motorola has confirmed that the Asian version of the XT720, the Motoroi, will receive the 2.2 update later this year, it has yet to confirm European availability.

Motorola has added a couple of useful apps including a version of the QuickOffice office suite is included, but it can only view and edit documents. Fortunately, the full version is available for just 6.49. Installed on the included 8GB microSD card is the Motonav GPS app. It's only a demo version though.

Another extra for car owners is the Moto Car Home app which gives you an alternative home screen for in car use. It divides the screen into six huge and easily tappable buttons Home, Bluetooth, View map, Voice search, Music and Contacts.

MySign is an app which lets you start other applications by drawing a symbol on the screen. As a shortcut it is a great idea, but it doesn't run invisibly in the background so you need to run the app in order to access its capabilities and not all apps work with it, so its usefulness is limited.

Call quality was fine, but battery life falls a little short. Depending on how heavily you use high end features, such as GPS and Wi-Fi, you may or may not get a full day from the Motorola Milestone XT720 between charges.

Entertainment features include an integrated FM radio, a music equaliser and support for H264, WMV and MPEG4 video files. The microHDMI port allows videos and photos to be displayed on an external screen. A microHDMI cable is included in the box. Unfortunately, but not unexpectedly, this feature only works with the bundled media player app.

You may be drawn to the Motorola Milestone XT720 by its high end flagship features such as the eight-megapixel camera, Xenon flash and 720p video recording, as well as its large screen and quirky physical design. Unfortunately, the camera isn't much of an improvement over other smartphone cameras. This leaves little reason to choose the XT720 over other smartphones, including other, very similar Android models.

Verdict

The Motorola Milestone XT720 is, on paper, an impressive phone with its eight-megapixel camera, 720p video recording, Xenon flash and large, high resolution screen. Unfortunately, the screen doesn't fare well under bright sunlight, while the camera didn't impress us with the quality of its photos and recorded video. There are plenty of other things to like about the XT720, including the large, high resolution screen which is great for web browsing. Nonetheless, you should take a look at other, similar, high end Android smartphones, such as the Samsung Galaxy S, before taking the plunge.

Connectivity: GSM 850/900/1800/1900, 3G 1700, 2100 Display: 480 x 854 pixels, 3.7 inches OS: Android 2.1 Camera: eight megapixels GPS: GPS Processor: 550Mhz Bluetooth: v2.1 + EDR Wi-Fi: 802.11 b/g Dimensions: 115.95 x 60.9 x 10.9 mm Weight: 160g Battery: Li-Polymer 1420 mAh

With every major mobile phone release, users are treated to an ever-expanding list of advanced features. Some are more useful than others, but they represent an industry that is always on the move.

The next big feature to make our lives easier is usually only an upgrade away but we've decided to gaze a little further into the future as we anticipate where manufacturers will go next with their products. Here is our top 10 future smartphone features.

Projection

We don't have to look too far into the future for this one, with Samsung's Beam handset already offering projection, albeit in a fairly primitive form and more advanced Android based handsets on the horizon. Basic or not though, the Beam is a ground breaking device, and in projection we see a feature which we think will become a staple in future releases.

Being able to sit down in a conference room and simply view a presentation or video without waiting for laptops to be plugged in or faulty cables to be swapped out would be breath of fresh air for most business users. As technology advances we will see the extension of projection into the video conferencing environment and others.

Looking further into the future we should see mobile devices packing holographic projection. The technology, which is in its infancy currently, offers a full three-dimensional rendering of the subject and would be invaluable to any designer or architect. People in the medical profession may have to wait a while longer for full body holographics but it's certainly not a flight of fancy to assume it will arrive.

Advanced imaging

Every new camera phone released is touted as a 'compact camera replacement' but with the advancement made in the camera industry they seldom live up to their hype. We're set to get a glimpse of what's in store though with the upcoming release of Nokia's N8 smartphone, which carries an eight megapixel sensor and Carl Zeiss optics.

We've also seen huge advances in video recording, with Samsung's Galaxy S and Apple's much vaunted iPhone 4 offering 720P capture, with HTC's Desire HD also entering the fray with 720P HD video capability.

As good as these features are, they aren't good enough to usurp a compact DV or still camera and the future may well see a further coming together of big names in the photography and mobile world as industry leaders in both fields realise that people don't want to carry a camera and a mobile device. So expect to see 12 megapixels and 1080P capture, huge on-board storage and even full editing on the hoof with Photoshop and Premier among the software being used.

Augmented reality is the future. As end-users we've had our initiation with Google's 'Goggles', 'Layar', 'Wikitude' and 'Last Minute's Nru', among others, but the flow of AR technology into the marketplace is only just beginning.

We can look forward to true location awareness, enabling clients to view couriers progress in real-time, AR for emergency services, showing the location of victims of crime or accident by their cellular signal and many other greatly useful applications to make our professional lives more efficient and safer.

Aside from general advancement in AR apps, we can also expect a range of devices offering AR replacements for traditional methods of human input, ranging from projected keyboards and touchpads to eye-dialling your phone via a heads-up display while driving. AR is also likely to fill the void between 2D projection and holographic, allowing users to conference and work in a CG environment, sharing information and tools and interacting via virtual controls and workspaces.

The application of AR technology will go beyond smartphones though. Indeed, expect to see augmented advertising hoardings and other interactive media as the norm in the coming years.

Desktop replacement

As processor manufacturers continue developing smaller, less power hungry and increasingly powerful products the traditional desktop computer edges nearer and nearer the door. With tablets like Dell's Streak and Nokia's N900 we've been given almost complete desktop functionality in the palm of our hand.

The successors to those devices will be empowered by multi-core processors, advanced graphic processing units and more RAM and ROM than you can shake a stick at. Enabling users to carry out every task they could want on their handheld, and de-centralising the workplace in ways not seen since the introduction of the World Wide Web. Expect to see super-mobile, super-powerful devices that you can plug into your display when you're in the office and work on when you're on the move.

In the more immediate future we can expect to see Nokia's follow up to the geek-friendly N900, sporting ARM's latest Cortex-A9 processor and a lush capacitive touch screen, and no doubt the next generation of Apple's iPad, which also runs on hardware provided by the British chip manufacturer. The iPad may have divided the community like the business equivalent of marmite, but it certainly heralded a new approach to super-mobile computing which we feel will be the standard in years to come.

Along with the power and flexibility of tomorrow's devices we should see a leap to cloud computing in the mobile market.

With smaller, more powerful devices having the lion's share of their hard work done remotely, users could expect to have all the power that they need without having the suit ripping form factor to go with it. Also, the de-centralised storage of documents, photographs and other data would enable users to work seamlessly with colleagues and even share devices without loss of data.

The removal of brand-based constraints from applications would also be a firm benefit with mobile cloud implementation, any device, any app; so long as it's registered on the network. We believe the mobile cloud would also see the distillation of mobile operating systems, leaving users with a quicker, more labour efficient experience when working on their devices.

The mobile cloud is already in use in certain applications such as Google's Gmail and Apple's MobileMe, but with more extensive usage network infrastructure would have to be fast and stable enough to support the high amount of data transfer needed to make such a concept viable which leads us on to our next point.

4G and beyond

The arrival of 4G is imminent and with it comes a new array of functionality and speed to enhance our mobile lives. Higher specced hardware and improved network infrastructure will provide businesses with a fast and stable data connection, and the next generation of applications, some of which we have already touched on, will run smoothly and - hopefully - without a hiccup.

Keeping you connected is only part of it, the higher speeds afforded users by 4G technology will allow the delivery of high-quality media to clients, enabling you to get your marketing materials out into the world that much faster.

With a new mobile generation coming around roughly every 10 years, we can expect to see 5G implemented fully by 2020, and following current projections will likely see peak data speeds of 1Gbps.

As mobile devices evolve into more than just communications tools, security becomes ever more important. Priceless data has famously been left on trains and in cars and the potential loss of revenue from such an oversight could be huge, not to mention the embarrassment to your company.

Enter biometric security.

Fujitsu has already begun rolling out fingerprint based biometric security across some of its range and in the near future voice or even inner-ear activated devices will be widely available, allowing corporations to protect their data fully when it's not in use.

Other uses could include individual workspaces within a single device, enabling a user to pick the device up and have his or her data downloaded automatically, once their identity has been confirmed a function which could prove invaluable with shared hardware.

As our mobiles become less exception and more norm the concept of all forms of necessary data being held within our device is gathering momentum. Our credit cards, passports, insurance documents etc. could all be carried around with us in our hip pocket, securely protected and unable to be used without our presence.

Operating systems

Current mobile operating systems are almost unrecognisable when compared to the basic user interfaces of a few years a go. With every release Google's Android is becoming faster and more stable and Apple strives to add more powerful features to its cleanly designed iOS.

The divide between mobile OS and desktop OS is narrowing though and in the future we expect to see much closer integration between the two. Following the lead of the MeeGo project we're moving toward true multi-platform operating system, one which can be installed on your netbook, smartphone and workstation to allow you a seamless working environment.

We've already touched upon mobile cloud computing and the possibility of a true cloud operating system in the mould of Google's Chrome OS really excites us, it also becomes more plausible with the upcoming introduction of 4G technology.

With the implementation of some of the technologies we've discussed the door is left open for new and more convenient form factors to emerge.

Wrist watches and long, pen-like devices have been seen on the market before but are never viewed as anything more than a quirk but with evolving technology allowing manufacturers to squeeze in everything a user needs they needn't be constrained by shape and size any longer. Think all-in-one earpiece devices, working fully from voice recognition or spectacles with an AR display in one lens, the possibilities are huge and manufacturers will be able to tailor devices to specific professions with greater and greater ease.

The svelte tablet won't be disappearing though but we expect them to get thinner and tougher, with higher resolution displays and longer life batteries.

Environmental awareness

Environmental awareness

This may not be a popular choice currently but the implications of sourcing a product from sustainable, fully recyclable and non-toxic materials are far greater than any one of us can currently see.

Aside from the far-reaching environmental consequences of irresponsible manufacture we have the added concerns of demand far outstripping supply when it comes to the materials used to build our important devices.

Sony Ericsson has taken steps towards a range of eco-friendly devices with its GreenHeart initiative, which promises lower carbon emissions, reduced paper usage and between 50 and 100 per cent recycled plastics used in production. These points not only benefit Mother Nature though, as sustainable materials also serve to keep production steady and with lower transport costs.

Another feature which we fully expect to see in future devices is solar panel integration or other eco-charging solutions. What could be better than saving money whilst saving the planet?

Is it a stretch to expect our children to be educated in the development and use of mobile tech in the future? Will there be an entirely new, life-changing device, which supersedes all others rendering them obsolete? Only time will tell.

As with all predictions for the future, some will happen, some will be lost forever or consigned to the gag reel. One thing that is certain though is that smartphones and mobile devices are here to stay in whatever guise human demand causes them to take and in 10 years, you can look back at this on your shiny, new ultra hi-definition micro tablet or HUD and say you saw it here first...

When American website the Onion ran a spoof news article on a fictional touch-sensitive click wheel to make typing a thing of the past, it created a stir, satire though it was.

Why? Because from two finger typists to business people looking for tools to help them write reports faster, the keyboardless laptop sounds attractive. So what's the real technology out there and can we say goodbye to keyboards yet?

Microsoft's Surface

Microsoft Surface is one keyboardless device.

It's a large touch sensitive table top capable of responding to touches from multiple users as well as hand gestures and physical objects placed on the device.

So how is this keyboard-free tool doing one year on since Microsoft expanded the commercial availability of Surface to markets in Europe, the Middle East and Africa (EMEA)?

Panos Panay, general manager for Microsoft Surface reveals: "We've seen notable growth in several areas, including an increase from 60 to over 300 Surface partners (nearly half from EMEA), availability in Australia as of this past January (totaling 18 available markets worldwide), and we've had over 10,000 downloads of the Surface SDK Workstation Edition."

Besides targeting retail, hospitality, financial services, healthcare and the public sector, Surface has seen its "largest European deployment to-date in 62 Vodafone stores across EMEA" as well as deployments in Deutsch Bank, Europcar, and Heathrow Airport.

Of course, Surface devices are not designed for everyday use like writing emails or reports in an office, given they measure 108 x 69 x 54cm, weigh 82kg and cost an awful lot to boot.

Any old surface?

The idea of turning any surface into a potential keyboard or interactive presentation from a coffee table to a piece of paper is another attractive idea and this is something else Microsoft are working on with Surface.

The idea is to create a portable computing solution that combines a mobile device with a camera projector system.

Information on how the technology would work is scarce but the basic idea is discussed on the Microsoft Surface Blog and there are some demos of the technology in action to tease users too.

The ultimate aim appears to be to combine this technology with multi-layer 3D presentation and augmented reality. But any commercial product seems a long way away.

Could touch screens replace keyboards?

Of course, touch screen laptop technology is here already HP unveiled its first multi-touch tablet aimed at business users this month.

With support for pen and finger touch and multi-touch gesture support the HP EliteBook 2740p tablet PC comes with an optional touch screen that can be rotated and placed on top of a keyboard.

Of course, it's still a way from a completely keyboardless laptop but hybrid tablets and laptops offering the choice of touch screen or keyboard and mouse navigation are becoming much more common.

Apple's iPad has no physical keyboard, providing an on-screen keyboard as part of its Mail app with auto-suggest and auto-correct grammar and spelling.

Could voice command tools replace keyboards?

Voice command tools like Google Voice are becoming more common. Google Voice offers a keyboard free way to search through your business contacts simply by using voice commands. Alas, for now, Google Voice is by invite only and only available for businesses in the US.

There's also Say Who, a free app for the iPhone available in the UK to dial business contacts simply using your voice. SayWhere is sure to follow, a tool for navigating to popular map, business search and websites purely by voice command.

Meanwhile, a number of companies are working on facilities for mobile devices that can access and launch a variety of applications and services using simple voice commands, as well as dictate their text messages a necessary precursor to the next stage of writing entire reports by dictation.

In the US, Nuance recently used its VSuite speech recognition framework to create such a tool using Nuance's VoiceMode speech-to-text technology for the Verizon Wireless Samsung Rogue.

Philippe Jeanrenaud of Marketing Nuance Mobility says: "We don't expect the keyboard to become obsolete as there are always situations where a user would be more comfortable using the keyboard (privacy for example). In addition, while speech recognition technology has dramatically improved in recent years, it is still sensitive to very noisy backgrounds (bars, train stations) and this would be another situation where the user be better served using a keyboard. In fact, Nuance feels very strongly there is a good complementarity between the two modalities: speech input and intelligent text entry methods."

Nuance's own voice operated tools may be here in Europe soon. Philippe adds: "We are not aware at this time of any plans to ship the Rogue in Europe with VoiceMode. However, we have launched an after market dictation application for the iPhone last December for the US market. This application was very successful (over a million download in the first weeks) and we have plans to follow up with launches in Europe within this year. Details to follow."

Keyboards are here to stay... for now

The keyboard is far from obsolete yet. There are some tasks from emailing others to writing a report for which they remain the ultimate tool.

Voice recognition and touch screen technology though is developing rapidly and for some tasks, like telephony to interacting with presentations and moving files and images, using voice command, on-screen or on-surface tools may be easier.

Microsoft's Panay can see a future where keyboards and keyboard free technology sit alongside one another depending on which is the optimum way to carry out a particular task.

"There are computing scenarios or situations where Surface and the multi-touch experience is ideal - [such as] where multiple people are gathered and collaborating - and times when it makes most sense to use a PC, mice and keyboard - data-inputting for productivity, writing, etc," he said.

"Therefore, rather than having one replace the other completely, we're heading into a new direction where we will decide on which technology to use based on which best meets the use scenario."

Speech recognition giant Nuance has acquired SpinVox in a deal worth more than $100 million.

The move comes after rumours surfaced earlier this month suggesting that SpinVox was readying itself to be bought. The rumours followed a troubled year during which SpinVox was forced to defend itself against suggestions that it was humans, rather than its tech, that did the bulk of the translation work.

At the time of the reports, neither party would comment on the rumoured buy-out.

The acquisition is based on a $66 million cash deal bolstered by $36.5 million in stock.

"Around the world, the voice-to-text market has experienced tremendous growth over the last year, with a variety of innovative services being delivered by carriers and unified communications providers," said John Pollard, vice president, of Nuance's Voice-to-Text Services arm.

"With SpinVox's robust infrastructure, language support and operational experience, we will broaden the reach and capabilities of our platform."

Once the deal is complete, SpinVox's carrier services will be integrated with Nuance's own speech recognition technology, with the aim of growing both the business and the size of its customer base, according to the company.

For those not in the know, every edition of the increasingly popular Ubuntu Linux has an alliterative, animal-based codename. Thus, Ubuntu 9.10 Karmic Koala replaces 9.04 Jaunty Jackelope, which displaced Hardy Heron and Gutsy Gibbon before it.

Canonical's chief, Mark Shuttleworth, has been outspoken of late in his opinion that making a dent in the desktop operating system market requires a cohesive effort to produce an operating system with an attractive, usable interface. He's succeeded to some extent, with Koala's revamped human look and feel, opting for the reserved, flat look adopted by Microsoft a few years back. Nonetheless it still looks like a catch-up effort, not a groundbreaking advance.

Installation

Burn the Karmic Koala ISO image to a CD (it's under 750MB), pop it in a Windows machine's drive and you'll be greeted with the Wubi (Windows Ubuntu Installer) interface. Wubi now allows installation from within Windows to what's effectively a virtual hard disk, eliminating the need for any partition changes. If you do this there's the usual slight I/O performance penalty for running a virtual disk, but the difference is small. The option to run as a live CD booting straight from the optical disc with no partition changes at all is still available too.

In use

From the standpoint of basic function and usability, it's long been hard to criticise Ubuntu. The interface is functional, clear, and the supplied applications don't suffer from the overlap of other distributions. It stands out from the likes of enthusiast-orientated distributions like Fedora, which come stuffed with dozens of apps, a large proportion of which do the same thing. Koala simply provides what any standard user will need, without succumbing to overload.

The application load-out remains largely unchanged, except for the removal of the Pidgin IM client in favour of Empathy. The version of OpenOffice included is 3.1, its major benefit over 2.x being Office 2007 docx-format compatibility. Bear in mind, however, that this has been shown to be less than perfect with complex documents.

Evolution remains the mail and calendar client, and with its relatively robust Exchange Server support that's no bad thing.

Karmic Koala takes yet another stab at easing the thorny topic of Linux software installation and management. Out goes the Add/Remove Software tool, in comes the Ubuntu Software Centre. The main interface is separated into sensible categories, but still leaves the user to separate the wheat from the chaff, with no ratings or indication of which applications are popular. It's easier to use than Synaptics Package Manager though (which is still included).

Cloud storage

No longer is it just the software that's free. With the release of Karmic Koala you get free cloud storage too, via the Ubuntu One service. This offers 2GB of free storage, or a flat $10 per month to upgrade to 50GB. Capacity beyond 50GB isn't currently available.

Technical improvements

The biggest noticeable performance boost comes in the shape of faster perceived boot times. The integration of the Upstart process, replacing the age-old init daemon, is the main reason for the quicker feel. Upstart is a multitasking system, allowing startup processes to do their thing in parallel. The init daemon, by contrast, is synchronous, blocking start-up tasks until the previous has completed. Upstart gives faster apparent boot time to the desktop - even if the system hasn't fully finished booting, it feels quick.

The Uncomplicated Firewall (ufw), the front-end for the dauntingly complex iptables firewall tool, gets an update including support for outgoing filtering. It remains a command-line application though, albeit one with (relatively) simple syntax.

An increase in Intel's already dominant market share for graphics chipsets, driven partly by the netbook explosion, has been addressed with Canonical claiming that major performance problems' in 9.04 have been solved. General performance with Intel graphics has also been improved.

In our testing though, driver compatibility seems to have taken a backward step in at least a few instances with Nvidia and ATI hardware. An office desktop PC with an old Nvidia 7300GT graphics card (that offered no resistance to an installation of Feisty Fawn) couldn't be booted into the desktop due to video corruption, either in normal or safe graphics mode. The same happened to a system fitted with an ATI Radeon 9600 card.

This is hardly a statistically valid sample size, but it's nonetheless disappointing to see chinks in Ubuntu's otherwise near-flawless Debian-derived compatibility armour.

Against Windows

The greatest problem for Ubuntu as Windows-competitor remains its sense of coherence. It still doesn't present the reassuring feel of a single product designed under a single roof. Karmic Koala is a significant step forward on that front, but the illusion is shattered as soon as you visit the Software Centre app and see lists of the kinds of applications you might have ordered on floppy disk from a public-domain library in the early 1990s.

From a productivity point of view, Windows Vista and Windows 7 are miles ahead with their shift to a search-orientated interface. Karmic Koala feels quaint with its standard point-and-click menus: hitting the Windows key and typing the name of the application or document you need is far, far faster.

Nonetheless, there's virtually nothing an Ubuntu desktop machine can't do that a Windows one can, and it's all for free. For that reason alone it makes a lot of sense to at least consider a roll-out or a feasibility study to assess the costs of staff retraining, should you decide to abandon the Microsoft ship.

Verdict

Karmic Koala cements Ubuntu’s position as the best of the free desktop Linux distributions, but Windows remains the most advanced and cohesive desktop operating system.

CPU: 700 MHz x86 processor or greater Memory: 384 MB of system memory (RAM) recommended HDD: 8GB of disk space minimum Graphics: 1024 x 768 capable Audio: Sound card Connectivity: A network or internet connection

Next-gen Xbox technology dubbed Project Natal is set to come to businesses within the next few years, according to Microsoft's chief executive Steve Ballmer.

Speaking at a event in London today, Ballmer said that the gesture and voice recognition systems are expected to arrive with Project Natal system for home gaming next year, removing the need for a controller.

While that technology could clearly be useful in businesses, Ballmer said Natal only really works from a distance of at least two feet away making it rather less useful for desktops. He said it's "better for the room environment."

While Microsoft is looking to sort that out, Ballmer did note that Natal tech could be useful for video conferencing or other "large settings", adding that could become available to businesses for such uses "shortly thereafter" it arrives on the Xbox console in 2010.

Ballmer said Natal would then move to desktops - within that two foot range - a year or two after that. "We've got to get it to Xbox first," he added.

He said Microsoft wants the "application scenario to be well nailed in the business environment and not just the home."

Google has started to send out invites to trial its Voice service, which redirects calls from one number to all your phones.

Google picked up the tech when it bought Grand Central. So far, the service has only been open to customers of that firm. Now, anyone who signed up to test the service is being sent an email with an invitation if you haven't signed up, you can do so here.

The free service lets you pick your own phone number, and redirect calls to that number to your work phone, home line or mobile device useful for those with multiple contact numbers or those frequently on the move.

Calls can also be diverted depending on the source, letting you filter work calls to your desk phone, and family calls to all your phones, for example - or block calls from unwanted contacts.

Aside from the call routing service, Google Voice also use a voice recognition system to transcribe voicemail to email or SMS.

TellMe, a subsidiary of [a href="http://www.microsoft.com" target="_blank"rel="nofollow"]Microsoft[/a], has announced today the release of three new voice recognition features for the Windows Mobile platform.

The announcement has been made exactly two years after Microsoft bought TellMe for a reported $800 million (540 million).

Jamie Bertasi, senior director of Business Solutions at Tellme, said: "Our goal is to give enterprises technology that improves the customer experience but also affords them the ultimate financial flexibility when deploying voice services."

"From initial deployments we're seeing impressive cost savings and results that we're sure our customers will be excited about."

The three new features, claimed to be designed with cost reduction in mind, are a VoIP carrier service, advanced speech services and what the company call a "new voice font."

The services are available immediately to TellMe platform customers and it has designed a specific version for the upcoming Windows Mobile 6.5 due later this year.

Google has released a mobile phone service called Google Voice that converts voicemails into emails, potentially stepping on the toes of companies such as Spinvox.

The new service weaves traditional phone features with Google's Gmail product, enabling a person to store transcripts of voicemail phone messages in their email inbox. As with all Google mail, these messages are then searchable.

The move comes as Google increasingly branches out from its stronghold in internet search, as it seeks to carve out a role in everything from mobile phones to personal productivity software.

And it demonstrates the company's ability to fuse various technologies home grown and acquired - into new products, even as the economic recession puts the future of certain Google projects in question.

Google Voice is based on the technology of Grand Central Communications a company that Google acquired in July 2007. After Grand Central remained silent for nearly two years under the Google flag, some observers wondered whether it had met the same fate as Dodgeball, a Google acquisition that was formally shut down this year.

Google Voice represents the first major update to Grand Central since the acquisition. Like the original Grand Central product, Google Voice offers consumers a single phone number that can route incoming calls to home, office and mobile phones.

The new version uses speech-recognition technology that automatically transcribing voicemails into text. The transcribed messages can be forwarded as an email or SMS text message to a person's email inbox.

It is unclear how Google Voice will fit into Google's business model, which relies on advertisers to provide 97 per cent of the company's revenue. The company has also ventured into the mobile software market, launching last year the Android mobile operating system.

Other than a feature that bills Google Voice users when they make long-distance phone calls, the product has no immediate means of generating revenue, said Craig Walker, group product manager for Real Time Communications at Google.

He said that Google Voice, which will be available to existing Grand Central users on Thursday and to the general public in the following weeks, provides another reason for people to spend more time on Google's various online properties, which benefits the company.

Google also makes money from selling enterprise versions of its applications to corporations. But Walker said the current priority is to make Google Voice a success as a free consumer product.

"There's all sorts of things we can do down the road," Walker said. "But right now we're just totally focused on getting the consumer product out."

A handset with an unlikely name is being touted as potentially revolutionising the industry, due to what its makers claim is the world's most accurate voice recognition system.

The Zumbafone could be available by the end of this year, according to reports.

The innovation is a circular pad that can be placed over the ear and detaches from a small handset that contains a circular dial pad and screen. Simply removing the earpiece pad from the handset activates a connection to the internet. You then simply say the name of a contact to dial a number or send a text. When you receive a text it can them be read out to you.

No contact information is stored on the handset itself, with all data being held in the cloud', which the makers say makes the phone 100 per cent secure. As it is fully tied to voice recognition, the claim is that if lost, the phone cannot be used by anyone else.

The phone is aimed as a low cost, or secondary phone, so eschews features such as high resolution screen and camera.

Rather than being developed by one of the world's large, well known handset manufacturers, the Zumba is being developed by IA Technologies, based in Hereford in the UK, using the brand name, Zumba Lumba.

Chief executive Dean McEvoy said his dyslexia made it difficult for him to send texts using conventional methods, leading him to develop the voice-based Zumba.

"I have dyslexia and I can't text - so what happens is that if I'm driving in my car if I get a text message it can be read back to me and I can respond to that by pressing on the button and I can speak my text."

The company says that the handset has been in development for three years, which might explain its rather dated looks.

In November last year, Google released a voice activated search tool for the Apple iPhone, though it currently only works well with US accents.

Touchscreens and mobile devices will be the norm by 2020, according to a survey of tech experts though that may not sound so cutting edge in the year that saw the Apple iPhone take touch and mobile web to the masses, and netbooks and dongles hit the mainstream.

But can you imagine most of your interaction with the web being on your handset? Many tech experts can. Indeed, mobile phones will become the main way of accessing the net for most people around the world within 12 years, according to 77 per cent of the 1,000 tech heads polled by the Pew Internet research centre for their "Future of the Internet" study.

The survey also highlighted improvements in voice-recognition, the architecture of the internet, and more use of virtual reality. Some 64 per cent of respondents said people will interact with computers using touch, talk and typing by 2020. And four out of five respondents said the internet's infrastructure will be improved but not replaced from scratch by that time.

But it's not all a shiny utopia of tech innovations. Such advanced technology will not necessarily boost social connections and could indeed spur "hate bigotry and terrorism" according to nearly half of respondents.

In addition, the ubiquity of computing and the internet could lead to even more blurring between work and personal time, according to 56 per cent of those surveyed.

And, the ongoing "arms race" between content creators and those who illegally share content will continue into 2020, according to the report.

Janna Anderson, the report's principal author, said: "These experts' answers reflect continuing concern over the tension between security and privacy issues. They emphasise the importance of enhanced and enlightened cooperation between the leaders involved in decisions about internet architecture and policy."

Nuance claims Dragon Naturally Speaking 10 (DNS 10) is the world's most popular speech recognition software. This is hardly surprising, considering it has virtually no competition at all in the retail market and little in corporate sales, either. With claimed accuracy of 99 per cent, you wouldn't think there was much more room for recognition improvement, but Nuance claims a further 20 percent, so you do the maths.

Initial training takes about five minutes and you read reasonably entertaining extracts from a choice of books to give the application a feel for your speaking style and accent. DNS 10 includes pre-defined style settings, but all these are for US regions, rather than Yorkshire or Somerset.

Nuance has updated the icons in its interface, but it's a shame it couldn't have gone the extra mile and given the whole interface a facelift. It's beige and white bars and dialogues haven't changed since the days of Windows 95 and the Accuracy Centre, with its ugly Times text and messy layout, would shame many a Freeware author.

Cosmetics are one thing and the functioning of the program is another. It's hard to test overall accuracy, as the software improves with use, by homing in on your particular vocabulary. You can give it a head-start with this, by letting it look through your document and email folders and compile a personal word list.

Initial recognition tests were good, but some way off the one word in a hundred the accuracy headline suggests. A lot depends on the quality of the sound the software hears. A noise-cancelling, cabled headset is provided with the program, but dedicated wireless headsets and some Bluetooth devices can also be used. A cabled installation usually gives better accuracy, but may not be as convenient, if you need to move away from your desk while dictating.

Nuance claims a speed increase of 50 percent in recognising what's been said, and in our tests it's noticeably quicker than version 9. If you speak phrase by phrase, as many people do, it's useful to be able to see the recognised text come up as you're speaking.

There are still some formatting niggles when dictating into Word one of our own pet gripes is that inserting quotes, speech marks or apostrophes doesn't use Word's typographical marks, but only the simplified, more utilitarian ones. Against this, you can now operate on sections of text, without having to first select them. Simply saying italic first' at the end of the last sentence, for example, would have changed first' to first'. Previously, you would have had to use the two commands select first' followed by italic that'.

DNS can be used for controlling Windows as well as for dictation and the latest version extends these controls from Word, WordPerfect, Excel and Outlook to email, web and search functions. For example, saying "search maps for Greek restaurants in Torbay" opens Internet Explorer and enters the phrase Greek restaurants in Torbay' into the default map search engine.

Web facilities extend to popular individual sites, so you can search eBay for a filing cabinet or Wikipedia for details of just about anything. Some commands within the web interface have been changed, so there's less likelihood of inadvertently typing dictation, if a word processor is running alongside the web browser. DNS supports Word and WordPerfect (which still has a big US and UK legal following), but not OpenOffice. This is odd, since the overall installed base of OpenOffice probably outnumbers that of the Corel product by quite some way.

The Professional version of DNS 10, which comes in Standard, Preferred and Legal versions, too, offers control of Outlook, PowerPoint and Lotus Notes, which the Standard and Preferred versions don't. All versions now support Firefox and Thunderbird, as well as IE and Outlook Express.

DNS 10 Professional is more network aware, working with Citrix thin clients and making personal profiles and vocabularies available from any machine. The main advantage of the Professional product over the less expensive Preferred version, is its macro and scripting support, where you can define and maintain verbal commands specific to your environment and the applications you use.

In case you're wondering, yes this review was dictated under DNS 10 Professional in a fresh installation, with impressively few corrections needed for mis-recognised words (though a couple of US spellings slipped in, but we will blame the software for that Ed).

Verdict

Nuance claims speed and accuracy improvements for the latest version of its speech recognition application. From our tests, both these claims are well-founded and Dragon Naturally Speaking 10 is an improvement on version 9. The extension of voice controls into e-mail and web browsing will be useful for many adopters.

Processor: 2.4GHz Pentium 4 class (1.6GHz dual core) OS: Windows Vista (32-bit only), XP (32-bit only), Server 2003, 2000 SP4 Memory: 1GB (2GB Vista) Hard Disk: 1GB free minimum

A new survey of consumer attitudes to contact centre service and technologies has found the telecoms and TV provider sectors perform worst, with three companies -BT, Sky and Virgin Media - included in the bottom five.

The survey of 1,000 UK adults undertaken by independent

ICM Research also found contact centres located offshore are a concern, as well as companies' growing reliance on new voice-recognition and touchtone, call-waiting and routing technologies.

The survey asked respondents about the level of service received from a range of companies operating in different sectors, including finance or banking, cinemas and mobile phone or utility companies.

BT received the highest amount of nominations for any company across all sectors, with common complaints about the service including language difficulties with offshore operators, long waiting times and being transferred from one department to another with problems remaining unresolved.

Power companies and banking institutions also generally performed poorly, where British Gas was found to be the worst of the utility companies, coming third in the top five worst overall performers. Of the banking sector, Lloyds TSB received more nominations than any other bank and was placed fifth in the top five worst overall performing contact centres.

The survey also found dissatisfaction surrounding offshore contact centres, particularly in terms of language difficulties encountered. But in technology terms, the biggest concern was caused by long holding times and new voice-recognition and touchtone, call-waiting and routing technologies, ironically introduced to manage customer service better.

CC, the survey sponsor and specialist convergence consultancy said the survey suggests this type of inconsistency seems to be one that affects a great deal of companies offering call centre services, affecting both onshore and offshore companies.

Respondents reported that at least 55 per cent of calls resulted in significant waiting times before they were able to speak to an operator across all sectors. And 38 per cent of callers to banks and financial institutions said the wait was "highly unsatisfactory", where 78 per cent of callers to companies in this sector were asked to select options by choosing numbers on their phones from a menu.

Call routing caused more dissatisfaction than any other method, with 41 per cent of callers expressing dissatisfaction with this approach because it often failed to offer the caller their desired option. But voice recognition also caused frustration, as it apparently was unable to recognise regional accents, and again, did not always deliver the options required by the caller.

BT said it had experienced some problems related to new IT-based contact centre systems when given the right to defend its position as worst performer of the survey.

A BT spokesperson told IT PRO: "BT has experienced some problems recently caused by the implementation of a new IT system to meet obligations under our Strategic Review. Unfortunately, in some cases there has been an impact on our customers. BT is working to clear the problems as soon as possible and it should be behind us by the end of the year."

BT receives some 640,000 calls and emails a day. And he added that 70 per cent of consumer calls are answered in less than 15 seconds.

Hammersmith Hospitals NHS Trust is streamlining procedures and speeding the time taken to produce x-ray results with speech recognition technology.

The radiology department at the Trust was struggling to pace with the reports needed to be produced from a growing number of investigations it carried out on patients.

Professor Philip Gishen, Hammersmith NHS Trust director of imaging said the department had as much a three-month backlog in transcribing the notes made by the radiologist into a report in the investigation carried out on a patient.

"We had 11 stenographers transcribing reports and they were not coping very well with the workload," he said.

Gishen said the department decided to look for a way of automating the reporting procedures when new targets designed to produce same-day investigations or examinations were introduced for ultrasound tests.

"The problem was a common one, where the referring doctors, as well as the radiologists, wanted to get examinations done when they want and need them done," he said. "It was also the same problem that they couldn't get results fast enough."

As a result the department introduced speech recognition technology from specialist Nuance to transcribe the radiologists' notes into reports and so reduce the time taken to produce examination results.

Gishen said the idea met with some initial resistance from clinical staff who were sceptical of the effectiveness of voice recognition technology, but Gishen said work to ensure the Nuance software could recognise specialist vocabulary and synchronise the right x-ray imagery with the right patient record allayed any user fears.

"We told them they would be losing their Dictaphones," he said. "So now they come out of the procedure or examination and the archiving system pulls up the images taken alongside the patient record. The clinician hits a button, records their report, hits another button and the report is ready to review."

Productivity increased to such an extent that when the Nuance speech recognition system was first introduced the productivity of the stenographers jumped from 45.5 per cent of investigation reports within the same day the month before, to 83.8 per cent the month after.

Gishen told IT PRO that now, over three years after its introduction, the department no longer needs stenographers as clinicians use the speech technology to produce all their reports.

He also said the success of speech recognition technology in radiology has opened up avenues for using it in further areas of the Trust, including patient call routing and clinician results services.