Erin Frey

Was heads down in an M&A process (my fellow bankers know exactly what that is like 😉) when I had my first conversation with @brainstub . Everything about it felt different. A data engine built to scale. A robot demo that’s real (not stitched or hand-picked) A team that treats every detail like it matters. It all stood out immediately and I knew I had to jump on the opportunity and make the pivot. Working with this team has been incredible, and I’m beyond grateful for everything I’ve learned over these last 4 months. Special thanks to @perryzjia @alpercanbe Proud to share the team story video below. This group is something else. @sundayrobotics 🤍

In 2026 ultra-highly produced content, especially video, is starting to be an anti-signal. throwing money at the problem belies a lack of internal POV / grassroots connection with the audience. make something personal, raw, and a little unfinished if you want to make it human. VC funds (mea culpa) are the biggest perpetrators of these random acts of marketing. We have such large budgets that it's easy to throw 10,20K at a professional videographer and have them go wild for one video. The industry as a whole lags 2-3 years behind what startups are figuring out works in marketing (with the interesting exception of events, where VCs are way ahead of the curve). So you are now starting to see quality-maxxing video podcasts coming out from funds en masse, years after startups took advantage of a small tear in the timeline alpha to use this medium to their advantage. But nobody is watching them! I don't click on highly produced video because the signal to noise ratio is nonexistent. It's all production quality and cuts, no underlying point of view. The hosts barely prepare for these episodes and read through a list of pre-written questions, none of which meaningfully challenge the guests. These videos say, look, we have money and we don't know what to do with it. Why did more than half a million people watch Dwarkesh uncomfortably argue with Richard Sutton in a drab conference room in Edmonton (probably)? Because the conversation was fucking interesting. We would have watched it if it was recorded in a dimly lit basement with an iphone camera. We watch him because he's a damn good interviewer, not because "more megapixels gud." The GOAT will always be the blog post on the engineer's personal site, talking about how they've gone deeper on some dumb topic than you ever thought possible. All substance, all deep focus, all original, no production value at all. Or you've got neal.fun where the production value itself is wrapped in the core substance. This is a subversion of the medium for sure, but it's equally thoughtful. This is why I'm putting most of my budget into writing and events this year, very little video (if occasional interesting stuff). You either have something to say or you don't; you have either done your research or you haven't; you cannot use wads of cash to conceal surface level engagement. Time will tell if this works, but I can tell you for sure that I'd rather die penning sick blog posts with the homies than watch yet another surface-level sitting interview with $20K worth of "cinematic style" video editing that goes nowhere fast.

I wanted to share something I built over the last few weeks: isometric.nyc is a massive isometric pixel art map of NYC, built with nano banana and coding agents. I didn't write a single line of code.

Announcing Linear Reviews. A modern code review experience for humans and agents. Join the waitlist for early access: linear.app/reviews