swyx

⚛️📝 New on Overreacted: JSX Over The Wire overreacted.io/jsx-over-the...

YouTube video by Latent Space S1: the $6 DeepSeek R1 Competitor (ft. Entropix)

i went on @latent.space and chatted with @swyx.io about s1, entropix, and reasoning models

does this make me a podcasting bro?
www.youtube.com/watch?v=MLxF...

Why MCP Won Learnings from Anthropic's extraordinarily successful Launch and Workshop

Interesting post about this from @swyx.io and crew: www.latent.space/p/why-mcp-won

It seems like the next step up from llmstxt.org

I think the usefulness really depends on if you want repeatable output from your LLM or if you use the LLM to build repeatable process (like terraform for config).

The Agent Reasoning Interface: o1/o3, Claude 3, ChatGPT Canvas, Tasks, and Operator — with Karina Nguyen of OpenAI Karina Nguyen from OpenAI (and previously Anthropic) discusses her work on Claude, ChatGPT Canvas and Tasks, and the new AI interaction paradigms for human-computer collaboration.

Fascinating to hear evolution of behavioural engineering in the context of generative AI models. Always useful to hear what @swyx.io is up to and their views.
The Agent Reasoning Interface: o1/o3, Claude 3, ChatGPT Canvas, Tasks, and Operator —with Karina Nguyen of #OpenAI
www.latent.space/p/karina

YouTube video by Infer Vancouver Engineering AI Agents – Shawn "swyx" Wang (Infer, Dec 2024)

The @swyx.io talk from the NeurIPS Infer Vancouver is now live! He shares an overview of the state of AI engineering, gives a demo of realtime AI APIs, and answers folks’ burning questions. www.youtube.com/watch?v=Hx_0...

whenever you want :)

hey hi hello how are you holding up

Screenshot of a GitHub issue containing a screenshot of my personal site (with a bug on it that I just fixed today). This is what Swyx calls "learning exhaust". (As an aside: a screenshot of a screenshot is kinda like Inception!! 🙂)

github.com/r002/r002.gi...

PS. I'm also a huge fan of artifacts (Seth even blogged ab this today! (seths.blog/2025/01/the-...)

For me, I ❤ leaving a "learning exhaust" (h/t: @swyx.io) of screenshots whenever I build things. Later, you can then look back on them all to see how far you've traveled.

2 pieces of content on the topic of LLM powered agents that I'd recommend:
- The latest Latent Space episode by @swyx.io with
@simon.fedi.simonwillison.net.ap.brid.gy - highly recommend!
- The GoEX research paper. Late to the party on this one but the contents of it is more and more relevant today

React Distros Why we no longer have frontend framework wars, and musing on the state of React metaframeworks today and tomorrow

@swyx.io pitched the "React distro" thesis a few years ago:

www.swyx.io/react-distros

I think it's a reasonable take / mindset.

The issue is that the React team has effectively stopped supporting the way half the ecosystem uses React, docs/guidance-wise.

YouTube video by Latent Space 2024 Year in Review: The Big Scaling Debate, the Four Wars of AI, Top Themes and the Rise of Agents

If you want to start 2025 feeling up to date on all the AI trends (without any BS) watch this video (from @swyx.io ): www.youtube.com/watch?v=26PV...

2 hours packed with info and ideas. No fluff. Just watch it.

Getting Closure on React Hooks Learn React Hooks by building a clone of React - in 30 minutes!

React knows which component instance it's rendering based on the current Fiber instance. That gets saved as a module-scoped variable in the hooks impl file, and the hook functions iterate through the list of hooks in a field on that Fiber.

See @swyx.io 's "build hooks" talk:

www.swyx.io/hooks

YouTube video by ぶれねこ(BreNeko) グリオーク遠距離狙撃【ゼルダの伝説 ティアーズオブザキングダム】

youtube.com/shorts/Vf_qA...

we changed to @latent.space

Our picks of the top ~50 papers for AI Engineers to read from scratch to competent in 2025:

latent.space/p/2025-papers

Feedback welcome/enjoy!

yeah this time the core will probably be open sourced (but like not in a big splashy “we are making a framework” way. tired of that .)

YouTube video by Grant Sanderson Ain't No Twin Primes (live) | Bill Withers parody by 3Blue1Brown and Acapella Science

ok what the fuck all the talent is here but not evenly distributed

youtu.be/djzKCZHeVjY?...

i am rewriting it this week anyway. always changing process.

i have shared the whole process with the @latent.space #ai-in-action discord before but didnt want it to be super wide spread haha

hi friends, our handle is now @latent.space! so nice that @substack supports bluesky!

we'll be looking to publish more here going forward. automation tools and ideas for content u like to see welcome

smol.ai News and Hackathons for AI Engineers!

the smol.ai newsletter is truly a godsend (thanks @swyx.io)

with all of the model releases this week, neurips, and discord chats popping off, having a single place to start from really helps

highly recommend.

Weight streaming might work well on GPUs! I wonder why it isn't a thing...

Oh man, writing up the Cerebras paper on Weight Streaming was a rabbit hole filled with parallel algorithms.

learning-exhaust.hashnode.dev/one-thing-i-...

@picocreator.bsky.social @eugeneyan.com @swyx.io @latentspacepod.bsky.social

The Rise of the AI Engineer Emergent capabilities are creating an emerging job title beyond the Prompt Engineer.

The Rise of the AI Engineer. Written over a year know. I think @swyx.io nailed it and the rise is here. www.latent.space/p/ai-engineer

arguably can go to ls live without the lottery lol

no api though right

why is matillion here but not airbyte lmao

yeah it can still do it but it self interrupts a lot and is no longer fun

always well presented and polished work from addy. i aspire to speak and write like him someday.

I'm heading to #NeurIPS next week, I'll be in town Wednesday-Friday!

I'll be at a couple things:
- Wed 1-2pm: talking Transformer killers with
@picocreator.bsky.social at @swyx.io @latentspacepod.bsky.social live!
- Wed 11am: RedPajama poster (spotlight) with
Maurice Weber

1/2

if you only learn 1 thing from philosophy 101 i hope it is deontology vs consequentialism