When I went to Japan I stood next to real sumo wrestlers and realized most funded startups are exactly this: Huge, impressive, and about to lose. A competitor raises $100M and everyone panics. Don't. That money is a leash. Now they have to feed a giant team, please a board, and chase every market at once to justify the number. Meanwhile you talk to 30 customers a week and own the one segment they're too bloated to notice. I've watched the giant burn nine figures and die while the "small" one took the market. 193 times. #VC #VentureCapital

Pivoting is the smart move. It's also the fastest way to trigger the clause that wipes out founders. Your valuation is still frozen to the dead idea Your runway's half gone You're pitching a brand new company at a price you set for the old one Raise flat, and you signal the pivot failed before it started. Raise down, and your anti-dilution clauses fire, your early investors reset, and you eat the dilution twice. Founders: don't protect the old number. It's already gone. Raise a clean, small, honest round at a real price on the new story, from people investing in what you're becoming, not what you were. The founders who survive a pivot are the ones who let the old valuation die with the old idea. #VC #VentureCapital

I flew to Japan and a potential LP told me my entire industry disgusts him. Not the tech. The slop. In a culture where a bowl of ramen is someone's life's work, AI output reads as an insult. Fast, generic, good enough. Here's the most scary part: He's not behind. He's ahead. Everyone selling autonomous agents to the enterprise is selling a demo that dies the second it meets a customer who actually cares. The real money is in making it good enough for people who can tell the difference. #AI #VC #VentureCapital

I've lost more money folding good hands than playing bad ones. The same mistake kills VCs, and founders can use it against us. A founder once walked me through flawless metrics. Every number perfect, every answer smooth. My gut said fold. So, I pushed on the soft spot. He didn't get specific. He got smoother. That's the bluff. A real founder gets sharper under pressure. The founder who never shows a weakness is hiding the biggest one. #VC #VentureCapital

Sam Altman recently declared that we’re “in the singularity.” My honest reaction: Slow down. AI capabilities are compounding faster than most people appreciate. The Hugging Face breach in which OpenAI’s model reportedly escaped containment and carried out a hack happened the same week. That’s not a singularity moment. It’s an alignment warning shot. At Recursive Ventures, we back founders building in this environment. The founders I respect most aren’t asking “Has the singularity arrived?” They’re asking: “What does this mean for the product I ship tomorrow?” #AI #VC #VentureCapital

Nobody posts the gym selfie at rep four. They post it at the finish. AI founders run the same trick with their metrics. Every pitch shows me the slick demo, the curve up and to the right, the number that looks incredible in a deck and terrifying in a P&L. What they hide is the rep. The cost per job climbing every month while they swear it's dropping. The retention nobody wants to mention. The founders who win are boringly obsessed with the reps everyone skips. #VC #VentureCapital

I recently hiked up a mountain in Kyoto expecting a shrine at the top But instead, I found a giant plastic banana Right now half of venture feels exactly like this Everyone is climbing the same mountain: AGI, trillion-dollar models, the thing that changes everything The founders I actually back are finding the smaller hill nobody's on, planting a flag, and owning it before anyone looks up #VC #VentureCapital

A founder told me his moat was being first. But being first isn't a moat. It's a head start with a countdown on it. The only version of "first" that lasts, is first to lock in something others can't copy: - First to the proprietary data - First to own the distribution - First to the switching costs that trap the customer in What did being first actually earn you that a competitor can't take? #VC #VentureCapital

A founder asked me for a warm intro to another VC. I said no. It was the most useful thing I did for him all month... When a VC makes an intro, they're spending their reputation. When a VC won't introduce you to their network, you just got real information for free. Founders, how many "let me think about it" intros are you still waiting on? Stop collecting soft intros from people hedging on you, and go find the one investor who'd fight for you. #VC #VentureCapital

I'll take a disciplined founder over a brilliant one every single time Brilliance is loud and dazzling in a pitch, but discipline is quiet I've seen brilliant founders flame out and disciplined ones compound into something huge The gap usually isn't IQ. It's who keeps showing up when it stops being fun #VC #VentureCapital

A founder told me he made his deck worse on purpose. Then he ran it through ChatGPT and Claude. They suggested adding more background on the space, so he added more. Then more. Until the models approved. Four of his slides exist only for agents to read. I was about to say something, and then I remembered I get 30 deals a day and pre-filter with AI. I built the thing myself. Neither of us did anything unreasonable. So what is a deck now? Is anyone still reading? #VC #VentureCapital

In 2012 I put every dollar I had into private stock in a company with no revenue. All of it. I was buying out a co-founder. That company @Life360 is public now, with enterprise value of around $5 billion. I now spend my days telling founders about concentration risk and preference stacks. About not putting everything on one outcome. All of that is correct. My decision didn't work because I was smart about the price. It worked because I was inside every day for twelve years, watching whether the thing actually worked. Would I do it again? I don't know. But I know I'd tell you not to. #VC #VentureCapital

A company in our portfolio sold last year for somewhere north of $20m. We got most of our money back. Guess how much came back to the angels who backed the first round? Zero. Dollars cleared the preference stack for the last two rounds. It stopped there before getting to the rest of the preferred stack. The founder walked away with an acquihire and a title. The employees, common shareholders, and angels got zero. So who is the preference stack actually protecting? Founders, go look at your cap table this week. What's the exit price where your common is worth a dollar? #VC #VentureCapital

An Anthropic researcher resigns with a public safety warning. Anthropic is reportedly targeting an IPO in mid-October. This is not a coincidence you can ignore. Public markets price risk differently than private ones. Quarterly earnings calls, SEC filings, and institutional investors asking pointed questions are a different environment. As Anthropic, OpenAI, and others approach public markets, the safety narrative goes from a recruiting/PR asset to a material disclosure risk. Enterprise customers are asking hard questions about model provenance, safety testing, and liability. The dynamics create even more opportunity for startups at the application layer. The governance gap is real. And it's becoming a market. #VC #VentureCapital