Y Combinator CEO Garry Tan reported that about a quarter of the accelerator’s current startups create 95% of their code using artificial intelligence, which is radically changing the economics of early-stage technology entrepreneurship.
Startups at the earliest stages in Silicon Valley are gaining a significant advantage thanks to artificial intelligence. This was stated by Garry Tan, CEO of Y Combinator — a renowned startup accelerator that has supported companies such as Airbnb, Dropbox, and Stripe — in an interview with CNBC on Saturday, March 15, 2025.
At this week’s annual demo day in San Francisco, where founders presented their projects to an audience of potential venture investors, Tan noted that the current group of startups is developing significantly faster than previous cohorts and is already generating real revenue.
“Over the past nine months, the entire set of YC companies has collectively grown by 10% per week,” Tan said. “It’s not just one or two leading companies — the entire set is growing by 10% week over week. This has never happened before in early-stage venture funding.”
This growth surge has been made possible by breakthroughs in artificial intelligence. Application developers can now automate most repetitive tasks and generate new code using large language models. Tan called this “vibe coding” — an approach where models take over software generation. In some cases, AI can code entire applications.
“For about a quarter of current YC startups, 95% of their code is written by artificial intelligence,” Tan noted. “This sounds a bit scary, but on the other hand, it means founders don’t need a team of 50 or 100 engineers. Companies are reaching revenues of up to $10 million with teams of fewer than 10 people. You don’t need to raise as much capital. Capital works much longer.”
Garry Tan emphasized that the “growth at any cost” mentality that prevailed in Silicon Valley during the zero interest rate era “is a thing of the past,” giving way to increased attention to profitability. This focus on financial performance also applies to tech giants. Google, Meta, and Amazon have conducted several rounds of layoffs and limited hiring new employees.
While this has caused concern among some engineers, Tan describes the situation as a new opportunity. Creating a startup has become easier, and the best technology specialists no longer need to prove their value by working at large tech companies.
“There’s a lot of anxiety in the labor market, especially among young programmers,” Tan said. “Perhaps the very engineer who couldn’t get a job at Meta or Google can actually build an independent business with a turnover of $10 million or $100 million a year with ten employees — that’s such a powerful moment in the software sphere.”
