1/ Controversial take: hard work is more important than smart work. It's a myth that we only have a few hours of good creative work per day. Train yourself to grind long hours first. You will surprise yourself. The work naturally become higher quality, less distracted.

some thoughts on the anthropic / coefficient bio deal 1) anthropic just acquired coefficient bio for $400m in a stock-for-stock deal; the company was founded in 2025 and consists of less than 10 people 2) the team includes nathan frey and samuel stanton from genentech's

One gets the sense that modern technology can be a cycle of manufacturing problems and solutions: big tobacco created the need for Keytruda, big food the need for GLPs, and big tech the need for stimulants/anti-depressants (and now Lilly's Orexin push). Obviously there's a lot of

some quick thoughts on mythos and tiered deployment 1) as models get more powerful and start to show dangerous capabilities, it will make sense to do tiered deployment to trusted enterprises first 2) this will enable domain experts and companies with access to real world

been mapping out the RL env ecosystem and i think there is a structural gap which can be explored better the market has five layers right now and they are being built simultaneously: > human expertise + rubrics (domain experts defining what “good” looks like) > RL environments

too much writing, too much thinking, too much planning, too much analysis. your pencils are sharp enough. theres not much time left now, its best if you get going.

"People are just a bit too soft in general. The brain is an organ. If you need to work more hours, you can train for it." --Jeffrey Yan

Speculator He's not that quirky (have met irl) The best way to describe him is "he is what SBF was cosplaying as"

Three of our early competitors are shutting down. Why? Here's what I know about each 👇️ Before I get into it, I want to say that I respect each one of these companies and their founders. It's incredibly hard to start a company, and even harder to win. Being in the arena is

Repeat Y Combinator founders hit different. Early success + the YC learnings = massively higher odds of building a category-defining company. Eg: 1. Sam Altman • Act I: Loopt (YC S05) — location-based social networking app. Sold to Green Dot for $43.4M • Act II: OpenAI —

I posted this because I disliked the takes on the podcast, and I dislike some takes in response to my post, so would like to clarify my position re: Dwarkesh vs Jensen. Dwarkesh Patel is a great podcaster, unnaturally so. Clearly he has studied his predecessors – chiefly Friedman

Reddit should start a data labeling / RL environment marketplace. They have some of the most niche, high octane, taste-pilled autists on the planet on tap. Most also have a relatively low opportunity cost to expertise ratio.

Ok. If experts are fake, then why does model progress rely on you paying Mercor, Handshake, Surge, Micro1, etc. billions of dollars to find experts to generate rubrics, define tasks, and create environments that models can learn from? If experts are fake then why do models need

given the vague post is blowing up, i want to tell you all a story from last night in nyc and where i think the world is going. last night, i went for a long walk as i felt quite existential about my business. i stopped into a restaurant. they had one last seat at the bar. i was

Read every word. This is the best, clearest summary I have read.

🚀 DeepSeek-V4 Preview is officially live & open-sourced! Welcome to the era of cost-effective 1M context length. 🔹 DeepSeek-V4-Pro: 1.6T total / 49B active params. Performance rivaling the world's top closed-source models. 🔹 DeepSeek-V4-Flash: 284B total / 13B active params.

the patrick / paul tudor jones podcast episode is one of the best 60 minutes of audio I’ve heard in years. drop everything to listen to it

This is it. Also, interesting that Sequoia is on a neolab tear: Standard Intelligence, Ineffable Intelligence, Flapping Airplanes. Wat means?

Mercor is structurally misaligned to human flourishing. It’s condemned to an information asymmetry that blocks its network from participating in the fruits of their labor. The highest paying agent training jobs will be in companies with equity ownership in their own compression