Billions of dollars are being wasted on AI, Stanford expert says

DAVOS, Switzerland – Many companies are getting stuck in the AI experimentation phase and have yet to see a substantial return on their investments, experts in the technology say.

Why it matters: Companies are investing hundreds of billions of dollars into AI at rapid speed as the business world increasingly signals that doing so is an economic imperative.

Axios' Ina Fried moderated conversations with Stanford Institute for Human-Centered AI professor and senior fellow Erik Brynjolfsson and Credo AI founder and CEO Navrina Singh. The Jan. 23 conversations were sponsored by C3 AI.

What they're saying: "There's an amazing hype here and billions of dollars are being wasted," Brynjolfsson said.

• Ultimately, the investments "only matter if they turn into business value, and we're not seeing that right now," Brynjolfsson said.

Brynjolfsson, an economist specializing in productivity, said the impact of AI on productivity has been minimal.

• "There's always this difficulty of translating even the most amazing, or maybe especially the most amazing, technologies, into productivity and business value," he said. "I call it the 'productivity J-curve,' because it sometimes even gets worse before it gets better."
• "We saw it with electricity, the steam engine, early computers, we're seeing it now. … The real challenge, the bottleneck, is figuring out how to identify business value."

Having rules in place for AI governance can help companies maximize the benefits of AI and minimize risk, Singh said.

• Singh's Credo AI helps companies quantify and manage AI pitfalls and rein in bias through defined rules.
• "[G]overnance is the way that most of these businesses, especially in generative AI, are getting the competitive advantage," Singh said.
• "The companies that we work with … understand it's not about just bringing in AI. It's about how they bring in AI by building trust not only with their customers but with their internal stakeholders, their employees."

What's next: The time for just experimenting with AI has passed, Brynjolfsson said. Now, companies "have to really connect it to a specific business goal."

• "I can bet on it. … All the companies who started to adopt the AI governance from the design phase, they are more competitive going into the future of especially agentic AI," Singh said.

Sponsored content:

In a View From the Top sponsored segment, C3 AI chairman and CEO Tom Siebel said most companies are still in the very early stages of adopting AI to its full potential.

• "As it relates to the application of AI to business and social processes, it is the first half of the first inning and the first batter is on his way to the plate," Siebel said. "We're just getting started."