A few thoughts on the current state of venture capital. When the Music Is Playing In July 2007, a few weeks before the credit markets seized up, Chuck Prince, then the CEO of Citigroup, gave an interview to the Financial Times. The line everyone remembers is this one: "As long as the music is playing, you've got to get up and dance." He was mocked for it for years afterward, and he lost his job a few months later. But I have come to think he was saying something honest. He wasn't claiming the music would play forever. He was admitting that he couldn't sit down while it was still going, and neither could anyone else in his seat. I've been thinking about that quote a lot lately, because right now is the most disorienting period in venture capital I can remember, and I have been doing this for a while. Here is what makes it disorienting. It's not that things are bad. Some things are spectacular. We have companies in our portfolio growing faster than anything I have seen in my career, and I don't say that lightly. At the same time, we have companies with no revenue, no product, and a founding team you could fit in a conference room raising billions of dollars at valuations of $10 to $50 billion. Both of these things are true at once, and if you try to reason about them with the same framework you will drive yourself crazy. Two ideas have helped me make sense of it. Neither is mine. The first is reflexivity, which George Soros has been writing about since the 1980s. In most of life, perception follows reality: the weather is what it is, and your opinion of it changes nothing. In markets, it runs the other way too. Prices change what participants believe, and what participants believe changes the prices. The feedback loop can run for a long time, and while it's running it looks exactly like progress. Here is how reflexivity is playing out in AI. Full disclosure: Menlo is an investor in Anthropic, so read the following with that in mind. People watched a frontier lab go from a $4 billion valuation to $18 billion, then $60 billion, then $180 billion, then $380 billion, and now something close to a trillion. They drew the obvious conclusion: that is what a neo lab looks like. So the next neo lab gets priced off that path, not off anything it has built. Then it gets marked up in a subsequent round, and the markup itself becomes the proof. Look at Thinking Machines. Look at Reflection. At that point valuation has stopped being an output of the metrics and has become the metric. Nobody is discounting cash flows. They are discounting the last round. Soros is very clear about one thing, and it's the part people skip: you cannot know when or how a reflexive process ends. You only know that it does. Every one of them has. The second idea is Chuck Prince's, and it explains why smart people keep dancing even when they can see the loop for what it is. As far as I can tell, there are two groups on the dance floor. The first group got in early. Firms like ours were in some of these AI companies before the numbers got silly, and the paper gains are enormous. When you are sitting on gains like that, you start to feel like you're playing with house money. I have been around long enough to know that house money is the most dangerous kind, because you don't respect it the way you respect money you had to earn. The second group missed the early rounds and knows it. Their LPs know it too. So they are trying to make up for lost time by writing very large checks very late, which is the one strategy almost guaranteed to turn a missed opportunity into a real loss. House money on one side, FOMO on the other, and reflexivity feeding both. That's the whole story. Everyone has a reason to keep dancing, and the reasons are different, which is why nobody can talk anyone else off the floor. So what do you do? The instinct in our business is to answer with company identification: just pick the right neo lab and you'll be fine. I think that's the trap. When price has become the signal, being right about the company is not enough, because you can be right about the company and still be wrong about the price by a factor of ten. The public-market investors I admire figured this out a long time ago. They spend as much time on how much to own as on what to own. The winners in venture over the next decade will be the firms that treat portfolio composition and position sizing as seriously as they treat sourcing. How much of the fund is in companies whose valuation rests on the last round rather than on revenue? What happens to the portfolio if the reflexive loop breaks next year instead of in five? Those are not exciting questions. They are the ones that will matter. The music will stop. It always does. Dance if you must, but know where the chairs are.

Investors do not all write checks for the same reason. Know the incentive. @TurnerNovak Founder & Solo GP @BananaCap_ on @fondocom START Full ep here: Learn more:

The best VCs in the world return the fund on 4% to 7% of their picks. @jmj on why every investor still thinks they're above average: "Right now it's much, much harder to pick a winning company at the seed than it was five years ago. The number of companies and ideas that come to market, and then the pace at which those same companies can be disrupted. When I started Chapter One, I went and spent time with Mike Maples at Floodgate. He's super generous, gave us access to all their historical data and some data from other firms who gave him permission to share. It gave us the picking rates from the best firms, and you could back into your own math on what your portfolio size should look like. I'd be willing to bet that data, while directionally useful today, is not as relevant. Because my realization is that the picking rates have gone down quite a bit across the board. Picking rate is what percentage of companies you invest in end up becoming fund-returning investments. The picking rates historically for some of the best venture firms, the best of the best, Sequoia, USV, the general range would be anywhere between 4% and 7%. So if a tier one fund has a 4% chance of returning the fund on an investment, they need 25 companies in their portfolio to just return the fund. Then you have to ask yourself as a newer manager, are the companies I'm seeing as good as the best firms in the world? For most people, obviously, the answer's no. So you'd have to construct a portfolio to have more companies, because you need more shots on goal. But literally every investor I talk to thinks they're above average. That they source better companies, or they have better judgment. There's a self-awareness that comes by saying, hey, we don't know. We have a feeling we're fishing in the right ponds. But we can't say for a fact that our companies are better than the best firm in the world. So I think bigger portfolios are, in most cases, a really good strategy."

How @jmj took down Tinder with its first push notification: "My first week, Tinder had no ability to send a push notification. We could send transaction notifications, so you have a message or you have a like. But we as a company could not send a push notification. The company just had not invested the resources in building it. So within a week they're like, 'Can you help us figure out how to send push notifications?' Then I was like, you know what? If I send our first push notification to 40 million people, I guarantee you we're gonna have a big day. At the time they gave me a CSV file of the mobile IDs. Literally exported a single file that lived on my hard drive. It was something you would never do as a public company. This was announcing a feature called Super Like. We had to figure out how to get it translated into 40 different languages. One of the things I figured out quickly was that you should rate limit the notifications, because if we just sent 40 million people the same notification at once, Tinder would go down. I ended up doing that by accident once or twice. We took down Tinder. But we sent it out and had our highest daily active day ever. That was within my first two or three weeks. Everyone's like, 'Jeff is a smart guy. He knows how to send push notifications.' It was amazing. I was like, literally I can be a hero at this company. From that they put me in charge of revenue, and we ended up becoming the top grossing app in the world."

.@jmj on why most pivots don’t go far enough: "There's a lot of companies who do a similar thing. They'll do a soft pivot, or an adjacent pivot that's one deviation away from what they're currently doing. When actually, probably the right thing if you're going to pivot, is to actually re-found the company. And that might mean doing something radically different from what you're doing today. The mistake people make when they're pivoting, whether it's their company, their venture firm, their careers, is not being bold enough about what that new direction might be. It's primarily because it's almost like an admission that what you were doing before is wrong. Which might not be the case. It might just be that what you were doing before is no longer the right strategy going forward. You're not underwriting your historical decision-making. You're taking a forward-looking view on what is happening in the world and how you should position your firm or your company. That's what I think people get wrong."

Beacon has acquired Haize Labs! So excited to welcome @leonardtang_ , @steveshenli and the team. Leonard and I bonded over a shared belief: responsible AI means AI that serves the real economy, and Main Street deserves the same caliber of AI talent and technology as the world’s largest companies. Now Leonard will serve as VP of AI Research across Beacon, whose companies serve 22,000+ businesses and institutions, from campgrounds to construction companies. We intend to own and grow these businesses for decades. Bringing in the top-caliber talent like the Haize team is the kind of investment we want every founder who joins Beacon to benefit from.

Standing up a production CDC pipeline used to mean designing source access, replication, backfills, destination writes, schema-change handling, monitoring, and recovery. Now, it can done with a single prompt: “Create a pipeline from the staging Postgres database to Snowflake.” With Artie MCP available in Claude Code, Cursor, and Codex, an AI agent can use Artie’s tools to configure and manage that workflow. So engineers can start with what they need, instead of spending months assembling and maintaining the pipeline themselves.

From @jmj on why more founders should pivot, and why you should pivot hard: "In the last five years I've probably had more phone calls with founders around the should-I-pivot conversation. And the one thing I've heard 100% of the time is, 'I didn't know that I could have this conversation with my investors.' I think it was seen as a sign of weakness. Or maybe you're worried, are we losing faith in the company? Does this mean we won't be able to raise the next round? On those calls I always say you should consider pivoting the company if it's at all a thought in the back of your mind. Most of the time the conversation starts with you're getting the same investor update for two quarters in a row. None of the metrics are moving. There’s small iterations on the idea, and there's nothing that's gonna profoundly change the business from where it is today. More investors should have that conversation. It creates a lot of trust with you and the team. But most of the time they're really excited to have it. And in probably every single case they have ended up pivoting the company. That just shows the team knows already that they need to pivot. They just haven't been able to come to terms with it, either internally or with their investors. Every month you're spending time on the wrong idea is a bad use of your time and capital as a founder. Especially right now, when you have so many things that you can build. You should not waste any time. Go work on really important problems, as opposed to being stubborn about the thing you pitched in your deck."

New @ThePeelPod with @jmj at @chapterone We talk about building a venture firm like a product, why round labels are dead, the reason picking rates have fallen, consumer AI, delivering flowers door-to-door in Kansas, taking down Tinder with a push notification, how none of his Fund 1 returners were in the Bay Area, and why Ch 1 never announced their $64M Fund 3. Full episode here + links below. 0:00 Publishing IC notes every week 3:45 Do round names matter anymore? 8:20 Putting a big check into Erebor's $2B round 11:40 Why deep tech went from instant pass to preferred in 3 years 15:30 When deep tech companies should raise debt 19:00 Should this company raise $1M or $100M? 23:25 You have two days to say yes 25:55 The sourcing software he built at Tinder 28:05 Their crypto book hit 22x, then the market turned 30:55 Paradigm, SendCutSend, and re-founding a firm 33:35 If you're going to pivot, re-found the company 37:25 Flex's wedge was too illegible to fund 40:10 When should you actually pivot? 42:45 Zaarly, the Uber for everything 45:15 Moving to Kansas City with a bag of clothes 47:25 Delivering flowers door-to-door 51:30 Raising a $64M Fund 3 and not announcing it 56:30 Joining Tinder as employee 50 57:35 The push notification that took down Tinder 1:00:55 Why you shouldn’t start a dating app 1:04:15 Consumer got too predictable 1:08:55 Consumer AI economics look worse than enterprise 1:10:35 Supabase and the non-human customer 1:13:00 Launching Chapter One from his Tinder desk 1:15:05 50 experiments per fund cycle 1:16:10 Product Club, the world's smallest accelerator 1:19:00 Evolving portfolio construction between funds 1:21:25 Why picking rates have fallen 1:24:40 Smaller funds can invest in illegible categories 1:27:40 Zero Fund 1 returners were in the Bay Area 1:29:35 Don't compete with Sequoia at Seed 1:32:00 His grandfather built Mervyn's

Databricks CRO Ron Gabrisko broke all the rules growing from $1M to $7B ARR. First one was hiring 40 sales reps in his first quarter. "It was a funny story. All day Wednesday and Thursday, all I'm doing is interviews. I told them, you've at least got to give me a break to go to the bathroom. A lot of those folks were from my network, people I trusted. My early thesis was, Spark is everywhere. My first task was to understand what they're willing to pay for, and who I can sell it to. What's the profile of the ideal customer? So you hire a bunch of people you trust to go talk to all those open source users, get that information, and find the trends. We'd just raised funding, so I did a coverage model to cover all the segments and find out which customers were more likely to buy, and what they wanted to buy. We moved fast that first year. We went from less than a million to $13, $15 million, then to $50, to $100, to $250 million. Now we're $6.9 billion plus. That was my first task. Go find out what people will pay for, and which segments you can sell to."