Investors were reminded this week as tech stocks stumbled that we are in the early innings of AI. Long-term, there’s a lot to gain, but in the short term there could be a lot to lose.
The first bit of bad news came from Sam Altman, the CEO of OpenAI, who admitted at a recent dinner that we are likely witnessing a dotcom-like AI bubble.
The second came from an MIT study, which poured cold water on the idea that companies are creating revenue-boosting efficiencies from AI. Instead, it found that 95% of AI pilot programs are failing to generate any meaningful return for their companies.
Fortune’s Sheryl Estrada spoke with the lead MIT study author, Aditya Challapally, who said the highest-returning projects focused on solving a single, very specific problem. Additionally, the most successful projects utilized external AI solutions versus home-grown ones, which were often costly and ineffective.
As Altman pointed out, bubbles aren’t baseless. They stem from “a kernel of truth” about the potential of a new kind of business or technology.
“Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes,” Altman told the room. “Is AI the most important thing to happen in a very long time? My opinion is also yes.”
But how long it will take to realize AI’s potential for true corporate gains is anyone’s guess.
For more on Sheryl’s interview with the author of the MIT AI pilot study, read here.