“I think most people are underestimating how bad the risks could be”: Anthropic CEO Dario Amodei lays out his vision of powerful AI — and explains why he sounds a tad pessimistic

The rise of generative AI has brought great promise for enterprises, according to Anthropic CEO Dario Amodei, but this latest industry trend has also fostered a culture of pessimism around the potential dangers of the technology.

Anthropic was set up by former OpenAI staff who left - at least partially - due to concerns about the potential risks of artificial general intelligence (AGI) and the way OpenAI was operating. That has often pitted Anthropic and Amodei as being pessimists or 'doomers', he noted in a recent blog post.

But that's not the case. Instead, he argues that overcoming risks is the only way to ensure a positive future for the technology — and unsurprisingly, he clearly thinks it has a big future.

"I think that most people are underestimating just how radical the upside of AI could be, just as I think most people are underestimating how bad the risks could be," he wrote.

The lengthy blog post detailing Amodei's thoughts about powerful AI — he says he doesn't like the term AGI — and the future of AI follows a similar post by OpenAI's Sam Altman.

Intriguingly, and both essays were released at a time when the respective companies were seeking funding. This could be coincidence, of course, but if not it suggests AI companies are keen to set out their stall and shape the debate around these technologies.

Anthropic acknowledges the downsides

So why does Amodei sometimes come across as a bit pessimistic? He suggests it's because he naturally avoids "grandiosity", hoping to avoid the baggage of sci-fi with over the top claims that can be off-putting to some people,

Notably, he suggests this is also an attempt to avoid the perception of propaganda. A plethora of enterprises globally have spent the last two years promoting the benefits of AI, but these firms do have a vested interest in this.

"AI companies talking about all the amazing benefits of AI can come off like propagandists, or as if they’re attempting to distract from downsides," he said.

Despite those concerns, Amodei listed the various ways AI could truly benefit society, too.

"Fear is one kind of motivator, but it’s not enough: we need hope as well,” he said.

What should we hope AI can help with? Amodei said the potential list of positive applications for AI is "extremely long" but focused on five areas that could offer the most benefit to quality of life for humans:

• Biology/physical health
• Neuroscience/mental health
• Economic development and poverty
• Peace and governance
• Work and meaning

“My basic prediction is that AI-enabled biology and medicine will allow us to compress the progress that human biologists would have achieved over the next 50-100 years into 5-10 years."

To achieve the serious gains in those areas, we need "powerful AI" — not necessarily AGI that is "smarter than a Nobel Prize winner" across multiple fields, but one that can complete tasks and easily interface with humans, Amodei said.

"We could summarize this as a “country of geniuses in a data center," he suggests.

The Anthropic chief executive admitted that some people believe powerful AI may never be built, but argues it'll happen within the next five to 10 years.

"I’ve tried to lay out a vision of a world that is both plausible if everything goes right with AI, and much better than the world today."

But — back to the risks — he warns that companies and everyone else will need to "do their part both to prevent risks and to realize the benefits."

More from ITPro

• Anthropic could be the champion AWS and Google need
• $100 billion to build an AI model? Anthropic CEO Dario Amodei predicts soaring AI training costs
• Anthropic just launched a $100 million investment fund for AI startups