One thing I’ve learned about building a firm is that you can’t build a brand the way you build a product. A venture brand is earned over years. Every conversation is a chance to strengthen your brand or weaken it. You show up, do the work & over time, that becomes the brand.

One question I get asked occasionally: "Are you still having fun as a VC?" I've never had more fun. You'd have to be the most boring investor alive to not be enjoying this moment. When we started the firm, there were a handful of interesting platform shifts. Mobile. Cloud. Fintech. Crypto. Consumer. Today there are 10 across robotics, finance, defense, aerospace, manufacturing, AI, healthcare, and more. Abundance is an understatement.

5. Many venture firms are out for the Fourth of July week, either this week, next week, or both. We expect Investment Committee next week to be much less active in terms of new pitches. Founders know this isn't the best week to be pitching for the most part. That said, hope everyone has a happy Fourth of July. Go America🇺🇸

4. We keep meeting brilliant founders building the wrong shape of company. One genre: a researcher with a real insight on how to make models better inside a vertical like science or medicine. The insight is usually genuine. But you have to assume the labs see the same opportunity, with more compute and more bandwidth to chase it. Anthropic, OpenAI, Google, Meta, xAI. Betting your vertical stays beneath their notice is a bet on their distraction, not your moat. Another genre: a quant who found a real signal and now wants to sell it. But if the signal is real, every buyer you add trades against it and the edge decays. And if it doesn't decay, why sell it. Trade it yourself. You can love the founder and still pass on the company. Most weeks, telling those two apart is the whole job.

3. We keep returning to @OpenRouter as the middleware/marketplace case study. We keep returning to OpenRouter as the case study. Model providers list their models on one side. Developers sit on the other, routing across all of them and actually putting them to work. Every model added gives developers more reason to be there. Every developer added gives providers more reason to list. The two sides recruit each other, and the middle becomes the market.

2. The most interesting founders we met this quarter are leaving software behind. Ex-lab, ex-SpaceX, ex-Tesla, who are in many cases walking away from the digital world as fast as AGI is arriving in it. The pattern is hard to miss. The people closest to the models are the ones betting on atoms.

Notes from Chapter One Investment Committee (June 29, 2026): 1. We’ve been debating pre-seed vs. seed more than usual. The gap in commercial progress between the two stages is often minimal, but the price difference is significant. We backed one company at a $7.5M post, and six months later they raised at $60M Same risk profile, very different entry. For a smaller fund, that ownership math matters a lot.

We went through 100 pounds of crawfish in an El Segundo parking lot on a Friday night and honestly that’s the dream.

Making the same bet

Last year we made an "exception" as an early stage fund to invest in the seed round of @gen_intuition. GI leverages Medal’s dataset of 2 billion videos to teach AI agents how to navigate space and time. What a fascinating company.

A pass-along deal from another VC is a yellow flag, not a gift. If a VC send you a deal they're not investing in, always ask why they didn't do it themselves. The best VC friendships are reciprocal. When I first started my career, my entire calendar was filled with recurring meetings with other investors. That list has become much smaller over the years.

Seed pricing has stopped being about the company. One early check from a known name sets the floor and the whole round prices off it. With the huge number of firms out there today, easy to get an early bid at a high price that sets off the entire round. The common fundraising strategy we're seeing is for a founder to set a valuation & get a high net worth investor to commit on a SAFE. Voila! You have your valuation. This isn't always the right strategy, to be clear, as leads or co-leads can always reset the price.

Most underrated category right now: marketplaces. The AI-native version especially where bot does intake, human expert does the work. The bot's the front door & the human's the reason you stay. VCs have largely forgotten about marketplaces as an investable category with the exception of data labeling co's like Mercor.

Notes from Chapter One Investment Committee (June 23th, 2026): We're testing a format where we share some of the key topics from our IC. A few things we got into today: An AI seed fund might be the hardest strategy in venture today When I started Chapter One in 2019, a well-known seed investor gave us the data. The best seed fund in the world had an 8% hit rate on fund-returners. The second-best, 5% - one in twenty bets returns the entire fund. Which means even they needed 20 companies just to return the fund once. So when LPs ask for concentration and a fund answers with 20 names, all AI, that's not concentration. That's the minimum number of shots required. And those numbers come from simpler markets before AI. The picking is now harder and portfolios are more correlated at the same time.

Last week one of our early-stage defense companies in Hawthorne raised a heavily oversubscribed seed round, co-led by two Bay Area firms. We were the only institutional check in their pre-seed, and we doubled down. Another LA win, to be announced soon.

One of the most successful subscription products of all time. Our team woke up the day after the release & realized we’d just built a product that increased our market cap by well over $10B. Wild times.

Another great day for Los Angeles & the United States. Congrats @isaiah_p_taylor & the entire @valaratomics team. An inspiring moment for our industry.

When people ask if I’m moving back to the Bay Area, I’ll send this map. We’re just getting started.