Why do AI projects fail?

It seems that everywhere you look these days, businesses are implementing AI features, tools, chatbots, and pilots. But researchers keep coming to the same conclusion about the benefits of enterprise AI adoption – that return on investment is slim to none.

While this isn’t the case for every business, it’s certainly a worry that hangs over discussions of the technology. The secret to making AI projects succeed is knowledge that every leader is after right now – and knowing what not to do is just as important.

Where are businesses going wrong with AI adoption? And how can they apply the learnings of the past few years to ramp up return on investment?Today, we’re joined by Alan Trefler, founder and CEO at Pegasystems, to unpack the main hurdles businesses face when it comes to AI adoption – and why AI failure is becoming such a widespread concern.

In this episode, Jane and Rory speak to Alan Trefler, founder and CEO at Pegasystems, to unpack the main hurdles businesses face when it comes to AI adoption – and why AI failure is becoming such a widespread concern.

Highlights

"You know, the thing about the AI capabilities is it's enormously powerful in putting together a demo. But when you want to do something that needs to be reliable, you need to have made sure that you've made the right decisions about what AI to use in what places."

"Well, I would say it's almost entirely strategic error in terms of people not actually having thought through what the different limitations of the AI are and how applicable they are to the particular problem they're trying to solve. Also, there is a completely mind-numbing amount of vendor hype in the market, I think, unduly covering and coloring the normal thought processes that people go through – and because it's beautiful and easy to get something to work a little, people assume that if they keep doing it the same way it'll work reliably."

"I think that not only is there not a good understanding at the board level, I would say in many technical organizations there's not a good understanding, which is why you have your articles about 95% of projects not going to production. It wasn't that those projects couldn't have gone into production, it's that they were doing the wrong things at design time and as a result, they did the wrong things at runtime."

Footnotes

• Why are so many AI projects destined for failure? Inexperienced staff, poor planning, and a shoehorned approach to agile development are all stifling innovation
• AI-generated code is now the cause of one-in-five breaches – but developers and security leaders alike are convinced the technology will come good eventually
• Nearly half of all digital initiatives still fail – here’s how you can learn from the ‘digital vanguard’ and deliver success

Subscribe 

• Subscribe to The IT Pro Podcast on Apple Podcasts
• Subscribe to The IT Pro Podcast on Spotify
• Subscribe to the IT Pro newsletter
• Join us on LinkedIn