“This is not dot-com”: Fiverr CEO says AI bubble will trigger a reset, not a crash
Micha Kaufman warns of too many look-alike startups, but says the strongest players show real economic traction.
Fiverr founder and CEO Micha Kaufman pushed back against the growing chorus comparing the generative-AI boom to the dot-com bubble. What’s happening now, he insisted, is something fundamentally different, a rapid, disorienting transformation that will sweep away hundreds of companies but will not drag the broader economy down with it.
“I don’t think it will end like the dot-com,” Kaufman said as part of Calcalist and Commit’s AI Week. “Many investors will find themselves with companies that are not going anywhere, but I don’t think this reset will drag the economy down.”
He argued that the defining feature of the current moment is not speculative mania or inflated promisesת but speed. “The way AI came into our lives, it was a simple mic drop,” he said of the release of ChatGPT. “The distribution was endless, the speed is beyond the human ability to understand it and respond to it quickly.”
Kaufman did not deny that speculative exuberance has taken hold. “Anything that ends with dot-AI, immediately gets funded,” he said. The result, he argued, is an overcrowded landscape filled with companies whose differences are “very, very minor,” competing for customers who have not grown in number or spending power.
“Humanity has not changed its size. Humanity has not changed its purchasing power in three months or three years,” he said. “Which means that in the end, there is no need for so many solutions.”
Dozens of companies in narrow niches will “naturally, disappear over time,” he predicted, not because the underlying technology is flawed but because they lack a meaningful business model. The winnowing will be painful, but limited.
“This is not going to bring about some kind of decade or five-year slump,” Kaufman said. “It will reset the investment strategy of investors, but it’s not going to get worse.”
Kaufman insisted the comparison to 2000 is misplaced because today’s strongest AI companies already show genuine economic traction.
“These are the companies that already have significant economic proof,” he said. “Their customer growth rate is phenomenal. They manage to create a brand in their field.”
The dot-com crash, he said, was driven by businesses that were “so theoretical” and lacked “a business basis behind them.” This time, he said, the reset will be more selective: money will move away from superficial experiments and toward “really significant worlds of value.”
Although confident about the long-term outlook, Kaufman described the present moment as one of unusual instability for corporate leaders.
“It’s a moment that creates huge unease,” he said. “You have to allow yourself to say: I don’t know. There are endless possible futures, and only one of them is going to come true.”
He likened today’s AI implementations to the early days of the steam engine, when the breakthrough was real but the applications were not yet transformative. “They were looking for an application,” he said. “The first implementations were to put a steam engine on a horse-drawn carriage. That’s what’s happening to AI today. Nobody’s building a car yet.”
Kaufman warned that companies must experiment aggressively, even without clarity. “Fear is something that is managed,” he said. “Those who are afraid don’t do it, and those who don’t do it don’t succeed.”
Execution speed, he argued, will determine who survives the reset. “The ability to move quickly, to do as many experiments as possible, overall you reduce the cost of the mistake.”
If the hype was once intoxicating, users are now encountering friction and frustration. “You talk to an AI and it doesn’t do what you ask it to do,” he said. “It keeps telling you, ‘You’re right, I’ll do it better,’ and it doesn’t do it better.”
For that reason, he said, the human element remains irreplaceable. “We still need the human talent, the sense of taste, the understanding of how humans react, the creativity, the nonlinear thinking - all the things technology today does not know how to do.”
