Databricks CEO Ali Ghodsi thinks the AI boom (and bubble) are creating pockets of relentless sameness across tech.
“I’m a little bit worried,” he said. “There’s definitely too much of everybody doing the same thing in general…It’s kind of like watching kids play soccer—all of them run towards the same ball. Then, there’s a big collision in the middle, somebody kicks the ball somewhere else—and then all of them run that way.”
Ghodsi and Databricks, in many ways, stand near the top of the AI heap. The company was founded in 2013, and has since raised billions in venture backing as it becomes a stalwart data, analytics, and AI platform. A favorite among enterprises, the company’s footprint is astonishing: In Q3, Databricks said it surpassed a $4.8 billion annual revenue run-rate, marking year-over-year growth of 55%.
Databricks announced that revenue figure with another key piece of financial news on Tuesday: that it had raised north of $4 billion—at a $134 billion valuation—for its Series L. And yes, that’s pretty darn far into the alphabet, a testament to how much financial infrastructure now exists for high-flying unicorns to stay private (and, frankly, calling Databricks a unicorn seems exceedingly quaint).
Quite a few big-name usual suspect investors were in the mix for this round: Insight Partners, Fidelity, and J.P. Morgan Asset Management led the round, joined by a range of heavy hitters, from Andreessen Horowitz to Blackstone to Temasek. Even with all this money moving about, a $134 billion valuation on its face remains jaw-dropping.
However, taken in the context of the AI-addled marketplace, it’s far from the craziest number out there. Elon Musk’s SpaceX is
anticipated to go public at well over $1 trillion, while OpenAI stands at $500 billion (and mark my words, that number will go up soon).
“If I wanted to have a crazy, crazy valuation, we would have gone public in the last 12 months,” Ghodsi says. “I’m not saying it’s guaranteed, but there’s a decent chance we would have gotten retail investors excited. As an AI company, valuation could have gone ahead of itself very easily. But what goes up comes down. So, we wanted to avoid that scenario. We just want to have a fair valuation that we can continue growing into over the next few years, every time we fundraise.”
Of course, this begs the natural, inexorable question: When
will Databricks go public? The company’s widely expected to hit the public markets in 2026, a prediction I bring directly to Ghodsi.
“I would not rule it out, but I also wouldn’t take it to the bank,” said Ghodsi. “We’re ready. It’s not an ‘if,’ it’s just a ‘when.’ I would like to avoid the scenario: If there’s a major correction, and you suddenly have to start tightening everything and produce massive EBITDA, you can’t spend as much time on massive opportunities for the future. And I’d like to invest in those opportunities.”
So, the answer is not tomorrow, not never, and some version of “we’re thinking about it.” It’s a tough market though, in certain ways, one where everyone is waiting for the AI-bubble shoe to drop. And Ghodsi does think parts of AI are undoubtedly bubbly.
“There are companies with massive valuations and zero revenue,” he told
Fortune. “That’s obviously a bubble. And by massive valuations—I mean billions, tens of billions, twenties of billions, thirties of billions, with almost no revenue to show for it.”
(For the record, this is also the part of the market I most worry about, where many of these startups are valued richly, but far from too big to fail. An easy spot for capital and jobs to get torched in an AI spending pullback.)
But there are areas Ghodsi absolutely thinks will come out on the other side: AI coding tools, for example, which he believes engineers and vibe coders alike will cling to even amid a visceral bubble burst. He said that the rise of vibe coding has spurred demand for Databricks products (the logic: more software, produced more quickly, creates more demand for databases and, by extension, Databricks). The areas he’s concerned about are simple: the ones where founders and investors are pouring time into chasing the same soccer ball. One example of an area he’s concerned about:
“Gyms for reinforcement learning and forward-deployed engineers,” he said, referencing simulated environments for testing AI agents and the engineers who, well, deploy AI agents. “What’s your business model?…I’m not saying those things are bad. But I am saying this: If that’s all you’re doing, that’s what everyone is doing.”
See you tomorrow,
Allie Garfinkle
X: @agarfinksEmail: alexandra.garfinkle@fortune.comSubmit a deal for the Term Sheet newsletter
here.
Joey Abrams curated the deals section of today’s newsletter. Subscribe here.
