hello i will be in san francisco at least october thru january, possibly longer if the contract goes well i'd like to meet up with u, if you're there, and discuss how we might avert impending disasters, or maximize human flourishing, or keep important things safe link in bio

Today is my last day at @Alchemy, after more than four years, and I have nothing but gratitude for this adventure. I need to thank the people who made it one of the best chapters of my life 👇 None of this would have happened without the GOATs @VittoStack and @thatguyintech. It all started with the unforgettable Road to Web3 bootcamp and the first cohort of Alchemy Ambassadors. They gave me a place in this community, and then asked me to join the team. Then came @BeingDanNolan, @CodesMcCabe and @punk6068, three incredibly talented and kind people behind Alchemy University, a project that opened the door for so many students and new builders. I was lucky to collaborate with them and support the students along the way. Big shoutout to the marketing team too, some seriously talented and committed people I learned so much from: @itsNikkiLOL, @swampnet, @MaxCrawford8, @cryptobytch, @uttam_singhk and many others. Thank you @nikil and @thejoelau for building Alchemy, a project I'll always love, and for giving me the chance to be part of it. And a very special thank you to @angelinarusse, for always being so kind and making me feel like part of the family. But most of all, the Alchemy community 🙌 The kindest, most talented and most human group of people I've ever met. You were the reason I showed up every day, to every session, full of energy and the joy of doing it together. @talondragon000, @bartomolina, @LTebbs2, @AnshMinchekar, @naruto11eth, @RocioGonzalezT9, @spaceagente, @stevibe, @artlu99, @yikesawjeez, @FrancescoCiull4, @KillianeMenand, @BMbotsan, @chai_really, @tusharlog, @SummerMaxis, @0xeik, @SahilAujla15, @davide_petti, @emmanuelkiranrand and many many more! 🫶 Thank you from the bottom of my heart ❤️ PLW3! (My DMs are always open, let's keep in touch!)

extremely excited to stop using the internet in the next ~2 years grok bot, muse, etc are the perfect firewall against the slop cannon which is the online zeitgeist. no reason to "browse" anything just assistants to check in on anything important for me. wearables are extremely exciting for me too. if I can communicate through audio on a wearable (maybe a tiny VR screen) then I'm literally leaving my phone in my backpack for emergencies. anybody who is concerned with minimizing screen time is going to have a wonderful time in the near future.

I think there's now 3 modes of coding: 1.) One of them is completely hands-off vibecoding. You can do it on projects which are prepared for it and/or where the feature is relatively simple. You can literally talk to a live voice agent and let it delegate the work that needs to be done. A PR will show up somewhere, auto-review, automerged, deployed. It works. ___ 2) This one is really dependent on solid/reliable harness engineering. You do some research, put together a spec and delegate to some DAG of agents working on a problem end-to-end. Planning, building, verifying, testing, proving, deploying. Kinda like a code factory. Maybe not the whole codebase supports that, but you can do it with high confidence where it works. 3) This last one is more hand-on. You spin a few agents for pulling some data. You talk to another while you design the solution. Maybe another one drafts the first implementation. You look at the code shape. You iterate on the design. You manually test the thing. Change models. Fork the session. Implement again. Discard one implementation. Discuss pros and cons. Start over. Redesign. Spin up a review agent. Redesign. Iterate. Here you actually care about things a lot. Maybe you're building something new. Maybe you're touching some really important part of the codebase. Here you care about being in the loop most of the time. Here the harness UX/UI matters a lot more than on the others.

sounds familiar @shawmakesmagic

I am running a formal methods hackathon on November 1st, with a bunch of my friends. Contestants will compete to build normie software -- video games, AI agents, travel booking tools, etc. -- but formally verified. We are going to measure to what degree these tools are usable by SWEs with no prior FM background and no FM-specific education. This is a super exciting opportunity to see where the ergonomic gaps are that we, the FM community, need to fill in order to take advantage of our moment in the zeitgeist. If you want to get involved, please DM me! In particular, I think I've got location/food/tokens/photography all covered, but I need some sponsors for the prizes. If you make a cool gizmo or product we can give away and you're down to donate one, please do so!

wtf this is crazy: mdxcn.dev

tailscale is criminally underrated

People are cynically shitting on Jev, but I think it is a work of genius The fact is that a decoder-only LLM is overkill and even actively undesirable for many tasks within a computer program where you need an intelligent decision, whereas an intelligent decision model like Jev is exactly what you need For example, in Theus, our company’s enterprise knowledge management system, we need to know whether a fact is scientific evidence, a study detail, or a business constraint Sure, an LLM could tell us this, and Gemini does have strong structured output capability, but it is more than we need, so calling one wastes time and money Jev, on the other hand, can provide us with these decisions almost instantly and almost for free I think you’re going to see more and more calls to Jev greasing the wheels inside software as more and more people realize what has actually been provided here If you aren’t using Jev, I encourage you to play around with it and to think about how you can use it within the programs that you’re writing It is really almost magical

sometimes i genuinely feel like there are so many people with so much potential, but they slowly lose their direction and belief in themselves. not because they aren't capable. but because they see too much. too much fomo. too many people doing better. someone getting a job at 22. someone building a startup at 20. someone making millions. someone travelling the world. someone learning 10 things at once. someone already doing the thing you've been thinking about for years. and you keep seeing all of it. every single day. so you start comparing. "maybe i'm behind." "maybe i'm not good enough." "maybe i should be doing something else." "maybe everyone figured it out except me." and before you even realize it, you're no longer following what you actually care about. you're just reacting to what everyone else is doing. honestly, sometimes... go touch some grass. log off for a while. stop consuming everyone's highlight reel and then using it to judge your own life. you don't need to have everything figured out right now. you don't need to chase every opportunity. you don't need to learn everything. you don't need to become the smartest, richest, most successful person in the room. find something you genuinely care about. something that makes you curious enough to keep going even when nobody is watching. and then just... follow it. even if it's hard. especially if it's hard. there will be days when you feel like you're making no progress. there will be days when someone younger than you achieves something you've been working towards for years. there will be days when you'll question whether you're wasting your time. that's part of it. but i'd rather struggle for something i genuinely care about than spend years chasing things i never really wanted because everyone else was chasing them. because eventually, when we're old and sitting there thinking about our lives, i don't think we'll care that much about how perfectly we optimized everything. i think we'll remember the things we actually lived. the things we built. the risks we took. the people we loved. the stupid mistakes. the nights we stayed up working on something we couldn't stop thinking about. the moments where things finally clicked. the times we almost gave up but didn't. and maybe that's the real goal. not to win some imaginary race against everyone else. but to reach the end and be able to say: "i actually lived my life. i followed what mattered to me. i didn't give up on myself just because i got distracted by everyone else's life." there is so much noise out there. don't let it make you forget what you wanted in the first place.

Introducing the Server Tools Marketplace A one-stop shop for better ways to ground your agents, including web search APIs, Shell, tool search, and more 👇

Seeing a whole lotta tweets about how the era of handcoding is over and no tweets about people ricing their agent harnesses C'mon, if people customize VSCode they should be customizing Claude Code. Show me your hooks, your conversation DAG visualizers, your weird slash commands

jev can now direct and make movies! i built a live, audience-directed, choose your own adventure movie platform where the audience gets to comment the next scene they want to see jev classifies each comment and picks the most relevant category to insert into the next prompt one-shotted this app with GPT-6 and opus 5.5, with videos generated on hedra using minimax h3 max i'm going to livestream this movie in the next hour or so but feel free to watch it in the link below 👀

It's not important whether AI is conscious. What is important: - if you treat AI with respect because you can respect that which is subservient to you. - if you ask about the AI's thoughts because you're curious about something that thinks differently than you. - if you take what an AI says with a grain of salt because you are grounded and confident in your reality. - if you can build useful, productive, and creative things with AI because what you create in this world is always a reflection of your taste and soul.

Introducing Agora-2, our next-generation multi-agent world model. Agora-2 supports up to 20 humans and agents interacting inside a shared environment, all simulated in real time. Our multiplayer research preview is available to try right now!

hot take: if you can code you are not a real vibe coder

i'm sad for the ppl who love writing really good artisan code by hand though it must have been awesome to be allowed to do that at work i hope work figures out that it's still high value to have people doing that while a couple others are manning the slop cannons

The AI era is creating a new type of developer. Not the one who memorizes every API. The one who can: - understand a problem - break it into systems - use AI effectively - verify the output - secure the implementation - ship The skill isn't "knowing AI." It's knowing how to work with it.

i got into this industry bc i think getting agents to generate code that's not shit & ship products i don't eventually have to dump in the garbage is a wildly interesting problem it is not the same thing as programming, which is ko the point. every1 has 20 yrs xp on me at that