Founders can feel the major AI labs holding their breath. Sequoia partners say the answer is to pivot, not try to surpass the billions of dollars being spent.
“The reality is, if you sell a tool now, you’re putting that model in the spotlight and effectively competing with the next generation models that are coming out,” Julien Bek, a partner at Sequoia, said on Monday’s episode of TBPN.
Baek said the founders he works with are “wondering whether we’re just iterating on a model that replaces what we’re doing.”
Sequoia’s @JulienBek Many of the company’s founders now say they believe AI labs are “one step away” from disrupting their businesses.
He says the most defensible companies, and potentially the next trillion-dollar companies, will be those “masquerading as software businesses… https://t.co/HucxSDUNQl.” pic.twitter.com/QajMgm1fAi
— TBPN (@tbpn) March 9, 2026
That’s why Baek said the best bet is to focus on service. In a recent article for X, Baek explained why he thinks the next trillion-dollar company will be a software company masquerading as a services company.
Beck’s comments come as concerns about AI disruption continue to drive financial markets. Last month, Anthropic’s Claude’s advances sparked a stock market crash and a SaaSpocalypse scare that wiped out more than $1 trillion in technology valuations before some companies began to recover. Other industries are also concerned that leading AI models will soon disrupt their business models.
Baek said that by focusing on services, the founders can ride the wave of advances in AI, rather than trying to pioneer areas ripe for takeover.
“If you sell your work, you’re actually benefiting from what the models are doing and all the billions of dollars that are being put into AI,” Beck said. “So part of that prediction was that for every dollar spent on software, $6 would be spent on services. Previously, we could only really chase the dollar because we were building the tools.”
Baek said Sequioa is creating a map of the best “autopilot” companies by combining output from AI models with human judgment. He said the model is better at “everything related to intelligence,” but it still doesn’t address the natural niches that humans are good at.
“Judgment is something that humans really strive for, but it’s much harder to define,” Baek said. “It’s instinct, people call it taste, people call it experience. And these are things that we think humans will continue to be really good at for a very long time, until intelligence can absorb these skills into humans.”
In an interview with TBPN, Bek highlighted Sierra, an enterprise customer service startup co-founded by OpenAI Chairman Brett Taylor, as a great example of a company moving into “autopilot” mode. In “X,” Beck name-checked a long list of other newcomers, including Harper, Lilette, Anthelio, Harvey, and Juice Box.
Regarding industries ripe for disruption, Bek wrote in X magazine that insurance brokerage, accounting and auditing, and healthcare revenue cycles are all ripe for “autopilot” companies to dominate.
“For the foreseeable future, I think anything that requires more intelligence is actually going to be very interesting for these autopilots, because instead of having 10 humans and one AI, you can have one human and 10 AI, and that ratio will only change as the models improve,” he said in an interview with TBPN.
