My 861 Bookmarks

Very powerful prompt: "Explain it with gradually increasing complexity." https://t.co/1aCqrWWno5

Smart people suffer the most from thinking that they have to be as uniquely successful as they are uniquely gifted, so they get stuck in this “can’t settle for less but also don’t know exactly what it is that I’m trying to achieve” limbo

You have to make it. There's no other choice. You've got to make it. You have to become the best of the best of the best in your field. Aiming for anything lower is not worth it. Put in all the hard work it takes. Fear is not an option. Become good enough to trust your own self

If being smart is part of your identity, what’s your plan for when AGI commoditizes intelligence?

Best resume template! https://t.co/mjTgOBB4G5

My repeatable hack for acquiring inspiration is to read about the history of an impressive company in a different field. Often it gives a new vantage point to solve a tough problem that you couldn't have easily journeyed to through grit.

how do i speed run the basics for reinforcement learning? is there some toy example I ought to implement?

I’m obsessed with the startup ideation process. Eg. How do we create more hyper-growth companies & what were the top companies really like at the earliest days? This book might be the closest thing to a repeatable framework. https://t.co/6hgX5AybZf

The spatial computing app that I think intellectuals truly desire is war room software, collaborative spaces to do thinking and memory palacing within https://t.co/9NLYHsvk8Y

AI agent infra in 2023 vs 2024 from @MadronaVentures @jturow Summary of 6 key themes in 2024: https://t.co/lG0BuaVS11

I hate the discourse that tries to pit STEM and Humanities majors against each other, when they should be teaming up to bully the true enemies: Business Majors.

Once again thinking about this https://t.co/WeoxMRyDX1

Become AI polymath - Play with trained models, learn prompt engineering, connect them with eachother, give them vector database memory, put them into agent or multiagent frameworks such as Langchain, AutoGen or CrewAI, and so on, with for example Andrew Ng. - Implement and https://t.co/OAqKjkwSgl

The Rise of the “Software Creator” https://t.co/5q2k47NLJ9

elon musk on advice for young people https://t.co/RfoSIUGyHz

This would immediately be my favorite way to learn: Enter a product idea and get a curriculum that teaches you how to build it—by the end of the course you ship it

tips on how to make friends 1) do a sport (me: bjj) 2) have a creative outlet (me: pottery) 3) learn to cook a group meal (me: japanese curry katsu) Do sports and creative stuff, invite people over for food + a beer, or bring curry to the pottery studio (which i did) Make

Name a film, preferably newish, that, while widely considered great, has you itching to get on your soapbox and shouting "DO YOU ALL REALIZE HOW GREAT THIS IS? WE NEED TO BE TALKING ABOUT HOW GREAT THIS IS. ALL. THE. TIME." Yes, I just rewatched Phantom Thread.

I give this book to everyone. Tufte works through hundreds of syntax patterns and their effects on the reader, using real examples drawn from every type of writing imaginable. This book is a fucking joy. https://t.co/3t797GQsYw

I rebuilt @Replit — Sandbox is an open-source cloud code editing environment with an AI copilot and multiplayer collaboration, made with @Nextjs + @CloudflareDev Workers https://t.co/gLX0j9rY2L

https://t.co/41X300RWkv "Essays on programming I think about a lot" A nice collection :)

To all my followers: I really can’t tell you how excited I am by this book. It’s one of the most wonderful things I’ve read in a very, very long time.

a deep generalist is someone who discovers new niches by following their curiosity with extreme faith, and once the newly found niche is understood and its complexity exhausted, in a meta-systematic fashion, it’s time for them to move on

This paper introduces 26 guiding principles designed to streamline the process of querying and prompting large language models. Paper - 'Principled Instructions Are All You Need' Tested on Llama 1/2 (7B, 13B, and 70B) and GPT-3.5/4. https://t.co/IsF6fmoiuX

@AlexTheGoodman AI generated, human curated.

We are moving from a knowledge economy to an allocation economy. https://t.co/6h1PP2MZlw

If you're feeling really awful one day, try changing up something. Anything. Probably you'll feel better the next day because of regression to the mean, but you'll also have acquired a new superstition to practice for the rest of your life.

Name something that’s forever funny no matter how many times it’s repeated.

Clarity AI is 100% open-source and can be self-hosted while Magnific is closed: https://t.co/QmjeuKkRA2 Clarity AI is free to use on Replicate, while Magnific is paid only: https://t.co/hp1elJFooV Clarity AI has an API while Magnific does not, making it unable to integrate into https://t.co/B7Mi9mwhId

I still can't believe you can index your whole file system locally using BGE ( 100mb ) & vector db and then run llama 3 (4GB) on it what a time to be alive indexed my 20 pdf files in < 3 sec lol https://t.co/NSIhgD7t8Q