When Michael Seibel lost his position at the startup incubator Y Combinator, he didn’t find out in typical tech industry fashion, which might entail an email calling him to a Zoom meeting where the bad news would be delivered. He did it to himself. Today Seibel is announcing that he’s stepping down as YC’s managing director, a job that entailed running the heart of the business: selecting startup founders for the three-month program and running the boot-camp-style operation that hones the vision and execution of their ideas so they can raise money, release products, and attempt to become the next Airbnb or Stripe (both YC alumni).

Considering how important YC has been to the tech startup ecosystem, Seibel’s exit will have more resonance than your average corporate reshuffle. For one thing, the person who runs YC’s blue-chip accelerator has a significant hand in shaping the next generation of tech companies. And in recent months, YC has found itself in the crossfire of a war between tech and progressives. Whether intentional or not, Seibel, a well-liked entrepreneur and investor himself, is deftly stepping out of the line of fire.

Seibel explains the move as a more personal decision. Sometime last year he began to take stock, spurred in part by reading Strength to Strength, a book about career arcs, particularly pivots made late in life. He’s only 41, but precociousness is part of the founder mindset, and he’d been a startup CEO at 23. “I do everything early,” he says.

Michael SeibelCourtesy of Y Combinator

He realized that he had been running batches for as long as the person who first imagined YC into being, Paul Graham. After Covid waned, YC had returned to an in-person experience, and the software that it had developed to smooth the remote Covid-era program made an IRL operation easier to manage. Now the program works by splitting each batch of new startups into four groups, none larger than Dunbar’s Number of 150, estimated to be the maximum number of relationship’s a human brain can properly maintain. Each group has its own leader, so YC had less need for someone to oversee each cohort as a whole. And though Seibel enjoyed managing the overall program, he much preferred direct contact with company founders. So he will now become one of those four group leaders, who each mentor a quarter of the batch. It’s a particularly exciting time to do that, Seibel says, as many of the companies hinge on the AI boom.

Close observers of YC—and many in the startup ecosystem monitor the accelerator with the diligence of a behavior-tracking ad network—might wonder whether Seibel’s move might have something to do with his being passed over for the leadership of the entire operation. Forbes has reported that he was disappointed not to be tapped as CEO after the incubator’s president, Geoff Ralston, who had taken over when Sam Altman went full time leading OpenAI, left at the end of 2022. Ralston was replaced by YC’s former design guru, Garry Tan. Seibel tells me he did not feel dissed, though he would have accepted the job if offered. “If it was something that people thought was going to be the right thing, I was happy to do it. If not, I was more than happy to not,” he says. “My whole goal was to do whatever YC needed for me.”

Seibel’s self-demotion seems to be in keeping with a recent rethinking at Y Combinator: a refocusing toward a scrappy, boots-on-the-ground startup accelerator as it was under its initial leader and cofounder Graham. His successor, Altman, started a sprawling research operation that, among other things, launched OpenAI. Ralston had his own dreams, and YC started a continuity fund to enable it to make later-stage investments into maturing startups. Ralston was also enamored with scale. The Winter 2022 batch included 412 companies, each funded by the traditional seed investment from YC. Ralston boosted that initial slug of capital from $125,000 to $500,000 per company, for a 7 percent stake. When I last asked him whether there was a limit to how many startups YC could accommodate in each batch, Ralston said there wasn’t. It was possible, he believed, for a batch to number “thousands” of startups.

Under Tan, who took over in January 2023, there’s been a refocus on the founders themselves. Tan says YC had become kind of an umbrella company saying yes to a lot of things. “I asked, ‘How do we focus on what made YC awesome in the first place?’” The answer was mentoring cool founders, chosen through an exacting application process. The continuity fund was discontinued. YC had already separated itselffrom Altman’s research division, which is now called Open Research. The only remaining trace of Altman’s research operation within the company now is a financial stake in OpenAI. Most notably, batch sizes have been cut almost in half. Beginning Summer 2022, they numbered in the mid 200’s, with the current batch inching up to 260. This isn’t due to demand—27,000 companies applied for those slots.

That’s more intense competition than gaining entry to an elite university like Harvard or Stanford, which Y Combinator is often likened to. While Seibel resists the comparison, I suspect that, as with Harvard, many of those 27,000 applicants to YC are just as interested in the halo that comes from being accepted and in getting access to its network of alumni as they are in the education that they will get. When I ask Tan whether YC might now be part of the establishment, his pushback is tepid. “It was probably entering that realm,” he says. “I like to still think of this as entrepreneurs trying to figure it out—like Rage Against the Machine.”

Speaking of rage, it’s a good word to summarize how a lot of progressives feel about the tech industry. There’s an ongoing battle in San Francisco between techies demanding the city adopt tougher policies in the pandemic-blighted downtown area and liberals who cherish the city’s antiauthoritarian heritage. In May 2023, YC moved its headquarters from Mountain View to a large pier facility in San Francisco, steering itself into the teeth of this controversy. And one of the loudest voices in the conflict has been Tan, a participant of the movement to recall the city’s progressive district attorney and a source of intemperate tirades against the civic leaders.

In an era of tech scrutiny, the ethic of Y Combinator—which sees mentoring founders as akin to a social good to wider society—is being questioned. In the early days of YC, the fabled interviews that determine entry to the program had a vibe of American Idol to them. Founders were underdog talents reaching for the brass ring of glory. Now it’s more like being tapped for a secret society like Skull and Bones: Simply getting accepted means you will be not only well-funded, but cosseted by a network of 10,000 founders eager to help you out. Not exactly underdogs. Though Seibel disagrees: “When you’re a startup the whole ecosystem is the underdog,” he says. “Here in the Bay Area, we're surrounded by the big tech giants.”

That stance is pure Seibel. If nothing else, his conscious self-demotion puts a spotlight on an entrepreneur whose modest public profile belies a powerful impact on the startup world. I first met him in 2007, when he was a recent graduate of YC, a cohort then in low double digits. Seibel’s company was a wacky project called Justin.tv, devoted to livestreaming people’s lives 24/7, starting with that of eponymous cofounder Justin Kan. These looney videographers prowled a San Francisco highrise where so many YC founders rented space that it was nicknamed the “Y-scraper.” One might argue Justin.tv was a precursor to the influencer economy—after several pivots it morphed into streaming platform Twitch, purchased by Amazon for almost a billion dollars. Seibel had a talent for sharing the essentials of foundership. He was also African American, relatively rare in a field overly dominated by privileged, white Stanford grads. His upbeat demeanor and canny knowledge of the inside game of growing a company has benefited hundreds of startups. And since he is a personal investor in dozens of companies, and is on the boards of Reddit and Dropbox, the association with YC has been very good to him as well.

While becoming the CEO of Y Combinator might have been a capstone to his career, he says he’s a huge fan of Tan’s and is feeling great about his decision to move away from management to become more hands-on with future founders. “I always knew I wanted to kind of teach as my last job,” Seibel says. “Could you imagine teaching anywhere better than here?” Maybe … a university? I hear that there’s some teaching at those places as well.

I’ve embedded inside two different Y Combinator batches. The first was for Newsweek, when all of 12 startups were in the game. Four years later I returned to track the much larger Winter 2011 batch for WIRED. That cohort was transformational for YC, as midway through the process a surprise announcement dramatically changed the terms for every wannabe unicorn: Instead of an initial investment of around $20,000 for each company, every participant was staked $150,000. (These days, each startup gets $500,000.) Here’s a description of the moment that happened in late January 2011.

A couple of days after Prototype Day, the class convenes for a special Friday-night session at YC headquarters. Nobody knows why they are there; when Graham announced the mandatory meeting, he gave no clue as to what he had planned.

There has, however, been plenty of speculation. Is there going to be a big party for [YC cofounder] Jessica Livingston’s 40th birthday? Is Steve Jobs going to speak? Is Barack Obama?

Now Graham stands before the class. In an early sign of the evening’s significance, he is actually wearing long pants. He introduces the group to Ron Conway, the noted angel investor. Then Graham introduces a special guest, Yuri Milner. Milner has been conducting a relentless campaign to become a prominent Silicon Valley investor, putting hundreds of millions into Facebook, Groupon, and Zynga. Milner is not in Mountain View; he’s attending the World Economic Forum in Davos. But [YC partner Trevor Blackwell] has set him up with one of the Anybots, which Milner can control remotely from Europe. A tiny screen atop the wheeled robot’s saucer-shaped “head” carries a projection of Milner’s face, allowing him to talk to the group.

“So, the surprise,” Graham says, gesturing to Conway and the Milner-bot, “is that they want to invest in all of you.”

For a few seconds there is stunned silence as 99 founders try to process this news. It’s like a denial-of-service attack on their brains. Finally, there’s a collective exhale and a round of applause. This is good. Then Graham explains the terms that Milner is offering in collaboration with Conway’s firm, SV Angel: “$150,000 on a convertible note,” he says. “No cap.”

Translation: Instead of demanding a relatively large share for their ground-floor stake, Conway and Milner are agreeing to invest at whatever valuation the next round of investors sets. In other words, they get no advantage from taking such an early position. These are the most founder-friendly terms imaginable, with no downside. The room erupts into applause, hoots, and shouts. The budding entrepreneurs look like an Oprah audience after learning that everyone is getting a free Pontiac.

Lesley asks, “Why has Apple abandoned their not-so-secret car project?”

Thanks for the question, Lesley. I suspect Kevin Lynch, the brilliant engineer who moved from heading the Apple Watch franchise to Apple’s auto venture codenamed Titan, might be asking the same question. Then again, he might be among the few that really know the answer. I don’t!

Surely among the reasons are factors that outside pundits have cited. Cars have lower margins than digital products. Weakening government resolve to push for electric vehicles means that automakers would get fewer tax breaks and an insufficient battery-charging infrastructure. Fully autonomous driving, which was allegedly the original idea for an Apple Car, has proven to be incredibly difficult to realize. With those thoughts in mind, Tim Cook might well have balked at spending the hundreds of billions of dollars necessary to ramp up manufacturing for an auto company significant enough to make Tesla look like Tinkertoy.

But let me suggest another reason that Cook unplugged this decade-long project. I find it significant that many of the engineers on the project have been transferred to work on generative AI. Maybe Apple realized that the very survival of a major tech company depended on being a major player in this hot area. Instead of trying to lure outside talent—a difficult task in this ultra-competitive arena—it instead opted to reassign a handy internal stock of machine-learning acumen, the engineers futilely banging away on Titan.

You can submit questions tomail@wired.com. Write ASK LEVY in the subject line.

Driven by power-hungry data centers, power plants are overwhelmed. And new ones to replace them will be fossil-fuel burning climate killers.

Seibel is on the board of Reddit, part of YC’s first batch. Could its impending IPO be a threat to the platform’s survival?

The judge has spoken: Craig Wright is not Satoshi.

The horrifying story of 764, a network of child predators operating on Discord, Minecraft, Roblox, and other sites popular with young people and teens.

Here are our favorite pillows. Wake me when this year is over.

Don't miss future subscriber-only editions of this column.Subscribe to WIRED (50% off for Plaintext readers)today.