SARAH GUO

The Button Is Two Pixels Off

Sarah Guo noticed it. The button. Two pixels off. While everyone else scrolled past, she stopped. That's where taste begins, she says - in loneliness. One person who can't unsee what's wrong. Where it ends is different: leadership.

By October 2022, when Guo launched Conviction with a thesis on AI-native companies, she'd already logged nearly a decade at Greylock Partners. She'd been promoted to general partner at an age when most people are still figuring out their LinkedIn headlines. She was the youngest. She was the first female GP in the firm's 53-year run. She worked on 40+ investments. Then she walked.

Conviction's portfolio now reads like a who's-who of companies that didn't exist when you last updated your resume: Harvey, the legal AI startup at $3 billion. Mistral, the French open-source phenomenon at $6 billion. Sierra, Bret Taylor's conversational AI play at $4.5 billion. Baseten at $825 million. Eight unicorns total. Forty-four companies in less than four years.

Goldman to Greylock: The Bell Labs Kid

Guo's parents worked at Bell Labs before founding Casa Systems, a cloud networking enterprise. She grew up in Wisconsin, New Jersey, and New England - startup environments baked into childhood. By the time she landed at Goldman Sachs in 2011, she'd already studied literature, economics, statistics, and architecture at Penn's Wharton and Lauder Institute. Four degrees. One person.

At Goldman, she helped take Workday public. Advised Twitter, Netflix, Zynga, Nvidia. Met people like Aneel Bhusri and Reid Hoffman. Two years later, in 2013, she joined Greylock. Five years after that, she made partner.

The math is simple but the execution isn't: become one of the youngest general partners in venture capital history at one of its most storied firms while still in your twenties. The kind of thing that makes people ask how you did it, as if there's a recipe. There isn't. There's only what Guo calls intellectual honesty - the willingness to pursue truth through uncomfortable tension.

Software 3.0 and the Conviction Bet

When Guo talks about "Software 3.0," she means intelligent software. AI-native applications. The stuff that doesn't just automate tasks but understands context, learns patterns, makes decisions. Conviction is purpose-built for this - $1 to $25 million checks for companies at the inflection point between idea and inevitability.

In late 2024, Conviction raised a $230 million second fund. Mike Vernal, formerly of Sequoia, joined as co-GP. The firm stayed small - eight people total. Small teams, big bets. That's the thesis.

No Priors: Asking the Questions That Matter

Every week, Guo and Elad Gil - tech founder, super angel, fellow traveler in the AI wilderness - host No Priors. The podcast asks the questions everyone's thinking but few are equipped to answer: How far away is AGI? What markets are at risk for disruption? How will commerce, culture, and society change?

They talk to the world's leading AI engineers, researchers, and founders. Recent guests discussed autonomous agents, debt structures for GPU clusters, power distribution bottlenecks, how Notion is transitioning "from a tool where humans do the work, to one where humans manage a swarm of agents."

The podcast is popular because Guo and Gil don't perform. They ask. They probe. They let uncomfortable silences breathe. Intellectual honesty, again - rare and decisive.

The @saranormous Origin Story

Her Twitter handle is @saranormous. The pun is intentional - enormous ambition wrapped in self-aware humor. She uses the ⚰️ emoji to mean "I'm dead, let me share something funny." She carries her phone naked, no case. During COVID, she tweeted: "need an intern? hire my sister, she's great."

She wrote a Medium post called "Correcting a Misperception: The Founder-Asshole," pushing back against Silicon Valley's myth that you need to be a jerk to be visionary. You don't. You can walk through walls while remaining transparent and empathetic. You can care about pixels and people.

When people say San Francisco is dead, Guo corrects them: "anyone who thinks SF is dead is being willfully blind to how important being an ecosystem 'insider' still is." Sure, if you were a Stripe engineering leader, you can move to Miami and recruit and raise money. But the magic of Silicon Valley is how quickly former outsiders can become relevant.

Forbes Midas Seed List and What Comes Next

In 2025, Guo debuted at #21 on Forbes' Midas Seed List, the ranking of top early-stage VCs. She'd already made Forbes 30 Under 30 years earlier, but this was different. This wasn't potential. This was proof.

By April 2026, she was speaking at HumanX in San Francisco, discussing why serious AI applications are taking more control over their models and inference, the tradeoffs between specialized systems and monolithic models, what investors look for as signals of durable advantage.

She's reading The Making of the Atomic Bomb by Richard Rhodes. Actively seeking book recommendations. Living in Silicon Valley. Focused on building family and community alongside building Conviction.

The Philosophy: Taste, Truth, Tension

Guo's philosophy starts with taste. Not the algorithmic kind, but the human kind. The ability to see what's two pixels off and fix it. The loneliness of being the only one who notices. The leadership of making others see it too.

Then there's intellectual honesty - pursuing truth even when it's uncomfortable. Especially when it's uncomfortable. The tension between what you want to be true and what is actually true. That tension, she says, compounds better than talent.

Finally, there's the conviction (lowercase 'c') that the future isn't evenly distributed yet, but it's coming. Software 3.0. AI-native applications. Companies that don't just use AI as a feature but are built on AI as foundation. Harvey automating legal research. Mistral open-sourcing frontier models. Sierra turning customer service into conversation. Notion managing swarms of agents instead of humans managing tasks.

What the Numbers Don't Tell You

Eight unicorns. Forty-four portfolio companies. Two funds. One podcast. These are the metrics. But metrics don't capture the texture of a career built on noticing what others miss.

They don't capture the decision to leave Greylock at the height of success because the opportunity cost of not building Conviction was too high. They don't capture the vulnerability of tweeting about your sister needing an internship or admitting you carry your phone without a case. They don't capture the intellectual labor of hosting a podcast that asks genuinely hard questions about AGI timelines and societal disruption.

Guo isn't building a traditional venture firm. Conviction stayed small when it could have scaled. It focused on AI when diversification would have been safer. It partnered with founders tackling the hardest problems - legal AI, open-source models, conversational interfaces - instead of chasing consumer social or fintech.

The venture industry, Guo says, could use more curiosity, more punk, more diversity. She's providing all three. Curiosity in the form of No Priors, where no question is too big. Punk in the form of Conviction, where the thesis is singular and unapologetic. Diversity in the form of herself - one of the first female GPs at Greylock, now one of the most influential AI investors of her generation.

The Button, Revisited

It starts with the button. Two pixels off. Most people don't notice. Guo does. And she can't unsee it until it's fixed.

That's taste - the loneliness of noticing what's wrong, the leadership of making it right. That's intellectual honesty - the willingness to say "this is broken" even when everyone else says "it's fine." That's conviction - the belief that details matter, that truth compounds, that the future belongs to people who can't unsee what needs to change.

Sarah Guo saw the button. She fixed it. Then she built a firm that finds other people fixing buttons - Harvey fixing legal research, Mistral fixing AI accessibility, Sierra fixing customer service. Eight unicorns later, the pattern is clear: she doesn't just invest in AI. She invests in people who can't unsee what's two pixels off.

The rest of us are still scrolling.