Aaditya Nalawade

@himanshustwts @smallest_AI LFG

Career update: Excited to share that I have joined the incredible team at @smallest_AI to work on Research x Devrel! The team is cooking incredible small + efficient multi-modal models and it feels like an exciting time to push the frontier on scale!

@asimovinc @orcahand would go hard on asimov

Early hand designs for Asimov. We're testing different approaches to see what works. Nothing final yet. What would you prioritize: grip strength, dexterity, or modularity?

@himanshustwts this is a cool one

s in stanford stands for state-of-the-art

@fazalmittu_ great read

recently been interested in Meta's JEPA, an architecture for predictive models to be able to extract/understand higher level details from any modality of data. walkthrough on my website: fazalmittu.com/reports/ijepa going to be doing more of these going forward!

@suhasasumukh @KechoHayashi @hardeep_gambhir @localhosthq @KunalKSavita wanna visit so bad

a new chapter begins. welcome to our new office in bengaluru.

@rvivek @albinder @letsblinkit super well done video!

10-minute delivery in India isn't magic. It's really good engineering. @albinder and @letsblinkit built tech that most people never see. We went deep on how it works.

@saviomartin7 LFG bro!

weekend side project: simpleclaw.com > ultra simple one-click deploy OpenClaw done under 1 min > built for all normal people

@Mokshit06 🔥🔥

feels very weird to say i’ve spent 0.5-1% of my life working on this… but it’s finally out! (and yes 5.2 is better than 4.5)

@shivanimatla @cursor_ai @claudeai this is sooo good

volume on 🔊 built a tiny focus music maker where you can doodle your own ambient soundtrack and time your work sessions. made using @cursor_ai & @claudeai give it a try - focus-music-maker.vercel.app also share your doodles and music w me

@shoagarwal 🔥🔥

I trained a diffusion world model to generate 2D games autoregressively! sagarwal.me/blog/morpheus

@suhasasumukh @localhosthq lfg

the multidisciplinary lab for india’s technologists. launch party, feb 7. rsvp now. @localhosthq bangalore lab.

@madiator I just think good search + reasoning models do the job

Early on I used to use a lot of Deep Research.. but it i can't remember the last time i used it! Many months ago. Is this the same thing for you guys? Was this a fad?

@bamitsmanas @KairosComputer sooo cool!!

AI Interns, finally. You tell them what you need. They figure out how. You go live your life. That's @KairosComputer

@saurabhtwq well written sir!

hey hi! finally i published my blog on llm agent ! go check it out ! (link in replies)

@masonwang025 this is sooo beautifully written, couldn’t have captured it better

was pretty sure i was never going to start at stanford. then i did. (new essay!)

@_sathvikr @ycombinator @Etched @getinstachip yes see you there!

@aadityabuilds @ycombinator @Etched @getinstachip thanks yo also u gonna be at treehacks?

2025 wrapped: > deferred @ycombinator > joined @Etched (fumbled acquisition into internship) > realized I was building @getinstachip to continue my failed chip startup, but was still primarily interested in AI compute (not EDA) > shut down startup, hired full-time at Etched > didn't want to work on transformers without understanding why (from first principles) > advice from @sama reinforced this: only you can know what's right for you > joined @trainloop_ai (3 people at the time) to understand AI from ground up > interest in ARC-AGI + DSL-based approaches > Breakup (#3) - leading to spiral of reading and understanding what I want (where do my goals come from?) > started @Caltech as a math undergrad > studying core subjects, while reading books, training models, and building intuition about both transformers and human intelligence

@BalintFerenczy crazy stuff

Made a fun CodePen with an interactive cloud. Move it around, resize, set the weather. ⛅️🌩️ Next up: flares

@trycua super cool

Today we're announcing cua-bench: a framework for benchmarking, training data, and RL environments for computer-use AI agents. Why? Current agents show 10x variance across minor UI changes. Here's how we're fixing it.

@RyanNeverWrong @xai @Stanford @grok insane stuff, congrats bro!

Super exciting news to share. I’m starting at @xai on the post training team and just wrapped up my first day today. I’m taking a leave from @Stanford and I’m super excited about this new chapter of my life and making @grok the best model. More to come.