Abhishek Anand

@0x796F @physical_int excited to get our hands on it!

You can now train @physical_int style robots in 1 day for only $5k. Anvil’s devkits have all the hardware, software, controls, cameras, and more ready-to-go. (1/5)

🎉 We got a paper into ICLR 2026 (main conference)! Co-author of this work is @MayankGoel28, along with his collaborators! Full paper link 👇

@0x_Something @hytopia Legend - can’t wait to see what you do next!

Tomorrow will be my last day working with @hytopia, while they lean up the team and focus on weathering the market and addressing the key platform issues. I must say it's been tiring (many sleepless nights) but also soooo much fun! Loads I could say, but to keep it brief, thanks to @iamarkdev for reaching out to hire me. Shoutout to @Temptranquil for doing a lot more heavy lifting behind the scenes than most people are aware of. Can't tag them all, but the whole team they gathered are amazing. Some things have worked out, other things haven't, but I learnt a lot and it was fun! I'll still be rooting for the platform from the sidelines. A note to the general community: Was a pleasure getting to know you all, many of you personally. The stuff being working on is truly novel. It's a shame that it's not reflected well in the token price to date. The team has worked hard to balance: - Retail investor vs VC - Web2 vs Web3 - Blockchain infra (complex, costly and clunky) vs Product infra (and the impact on gamers) - Time spent optimising retention/performance vs time spent optimising for token price - Short term hype vs Long term sustainability - Low end device performance vs premium graphics - etc. Honestly, there are many passionate community members with so many great ideas. Often we we're juggling more than we could communicate, and executing on everything with such a small team. Hopefully, at some point, they'll crack the retention issues in the same way they cracked distribution and the market will appreciate what's happening here. As for me, I have some AI products that I'm looking forward to #BuildingInPublic over the next few weeks and months (more on that later). 👋👋👋

🔥 New research: Can coding agents actually optimize GPU inference code? We built ISO-Bench: 54 real optimization tasks from @vllm_project & @sgl_project and found that agents often understand the problem but can't execute the fix.

How does @DrJimFan always end up doing such cool work - massive!

@177pc @aakrit Congrats again🔥🔥

.@aakrit & I started Activate ~12 months ago w/ a conviction that India will both be a top consumer and builder of AI. Today, on the sidelines of the AI Summit, NVIDIA made that conviction official. NVIDIA and Activate are now exclusive partners unlocking the following for founders/startups: This unlocks 3 things: (1) Activate startups will have direct access to NVIDIA technical expertise for co-building support and CUDA platform integrations, open source models like Nemotron, tools, libraries and SDKs. (2) They will get access to new NVIDIA products, features and releases as fits their use-case to accelerate product development early. (3) NVIDIA and Activate will jointly identify high potential startups to invest & support & Activate will join the recently announced VC Alliance to work with not only NVIDIA Inception but also NVentures for potential investments Excited to be working with Tobias, Unnikrishnan & the entire NVIDIA leadership for betting on what India’s AI ecosystem can become, not just what it is today. And to every founder building at the frontier - the runway just got longer and the ceiling just got higher. We’re just getting started 💪 techcrunch.com/2026/02/19/nvi…

@aakrit Big W for Indian founders - congrats!

Today, I am excited to share that on the sidelines of the AI Summit, NVIDIA announced an exclusive partnership with Activate. This means 3 things:

@itzzme_pk Let’s do it!

@levelheaded_94 let's connect :) DMs are open

Will be at the AI Summit in Delhi from 16th to 18th Feb. DM's open if you are around and/or working on robotics/world models/physical intelligence/spatial infra.

@ak_panda Are you building this ?

@levelheaded_94 Use paakow.xyz to build your dataset for your world models. DM open for collab.

@MaxMill06 Very cool - how are you placed later in the day ?

@levelheaded_94 Let’s meet bro, building around world models, VLMs and physical AI

@deepanshgl Exciting - will be there today

@levelheaded_94 We will showcase our work on VLAs and whole Body Control! Let’s meet up there Booth 6P187, Hall 6

@_manan2005 Let’s do it

@levelheaded_94 Would love to meet and discuss research!

Announcing Rosalind, the most versatile AI Co-Scientist for computational biology and therapeutics research. Giving every biologist their own frontier research lab. Make every experiment count. It's live. Links in the comments.

@sachdh Will do sir

@levelheaded_94 Come to booth bro Hall 2 2P63

As Rocks May Think: an interactive essay on thinking models, automated research, and where I think they are headed. Enjoy! evjang.com/2026/02/04/roc…

@li_yitang Well said!

Strongly believe 2026 will be the year of large world models❤️. Priors over how the physical world works — and how it responds to intervention — feel like truly first-principles intelligence. Any policy without world modeling seems fundamentally unscalable. Vision–Language–Action: for me, only vision + skills live in the closed loop, forming the sense–act cycle that grounds physical intelligence. Language is mostly open-loop — commands, goals, guidance — not the substrate of behavioral or spatial intelligence. Also, I hope physical skills/actions become first-class citizens, not just distilling VLA from VLMs/LLMs — actions should reshape both vision and language spaces. For "world model", vision may be a cheap, interpretable proxy; real world models must encode contact, force, geometry, and dynamics — but how to represent and learn these remains wide open. On supervision: I feel that world modeling can turn 10× data into 1000× signal, but training world models still needs ~100× — and robotics today runs at ~1×. Huge gap ahead, but this feels like exactly the right moment for world models.

@Farshchi Beautifully articulated - I feel like this is the most under rated and under explored aspect of building a startup and builds such a big competitive differentiator in a world where you can no longer compete on the other variables

You don’t build generational technology companies around a widget or invention, you build it around a vision. The vision attracts talent, which punches above their weight to attract dollars, which attracts more great talent, seeding a virtuous cycle. Resist the pressure to “put out a product”- focus on building a great company! Then go out and build something magical. You won’t have too many shots on goal in deep tech, as every iteration is expensive, so make each one count!

@odysseyml Congrats guys 🚀

Introducing Odyssey-2 Pro—a frontier world model that generates long-running, interactive simulations in 720p! We're also launching the first world model API, to enable devs to build magical apps. We're now in the GPT-2 era of world models. Let the explosion of apps commence!

@ericjang11 Congrats on a great run and hope to see more writing from you as you move away from the daily hustle and bustle.

Life update: I've decided to leave 1X. It's been an honor helping grow the company. I joined Halodi Robotics in 2022 (prior name of the company) as the only California-based employee. At the time, we were about 40 based out of Norway and 2 in Texas. My first hire and I worked from my garage for a few months to save money. Today, 1X is hundreds of people, with hardware, design, software, AI, manufacturing, product all relocated to the SF Bay area, firing on all cylinders and working on getting NEO ready for the home. A big thank you to all my colleagues that I worked with. It was a hard decision to leave. When working at an exciting startup that is growing fast, there's always so much to do and never a perfect time time to move on. We have several works in the pipeline that are so exciting because they greatly advance general autonomy and scalability of our deployment approach and really show a realistic path towards the product working. The recent World Model autonomy update is one example, and there's more coming. The 1X factory is so exciting. Things are accelerating at a speed I would have been surprised by a few years ago. In 2022, most technologists and researchers and VCs were skeptical about humanoids and large scale imitation learning. "Why Legs?" "How could end-to-end learning ever be good enough?" "Why go for the home and not the factory?" "How will we ever gather enough data?" The Overton window on general-purpose robotics has shifted a lot since then. Although we are still early in our mission, I remain confident that soon, house robots will be as commonplace as air conditioners, cars, and ChatGPT. Just talk to the bot, and it will go and quietly get it done. Entire economies will eventually re-organize around this technology. People get it now. What's next? I believe that progress in applied deep learning generally rides on "harnessing the magic" of a few magical objects. These magical objects possess way more generalization power than one might normally expect. Just asking the LLM to understand what you want is magic. Video generation models are magic. Reasoning is magic. You don't run into a magic object every day, but when you do, you make sure to grab it and put it to work to make something useful in the robot somehow. A lot of my early conviction for where robotics was headed was working on BC-Z from 2018-2021. The "magical object" I bet on at the time was the surprising data-absorption capabilities of supervised learning and "just ask for generalization". This pioneered a lot of the standard ingredients we see in VLAs today: - Generalization to unseen language commands - Human-Guided DAgger for policy improvement - Open-loop auxiliary predictions + receding horizon control, AKA action chunking - Manipulation keypoints to improve servoing - Simple ResNet18 with FiLM conditioning on multi-modal inputs The next "magical object" we bet on at 1X was video models, because they are clearly magical objects that learn a data distribution not too dissimilar from what a robot needs to learn. They generalize surprisingly well. I am once again feeling that there are more magical objects in play now, which opens up a lot of new possibilities for robotics and beyond. I'm taking a few months to empty my cup of priors and gain fresh perspective. When I left Google in 2022, I spent about 2 weeks deciding what to do next. This time, I want to take a lot more time to catch up what has happened in the broader AI + robotics space. I've been re-implementing some deep learning papers. I'm working on a big tutorial for my blog. I'm learning all the Claude power user tricks. I'm reading the Thinking Machines blog posts to understand what kinds of experiments are being run at frontier labs. I'm reading Ben Katz's 2016 thesis on the Mini-cheetah actuator. I'm traveling to China in March to meet incredible companies in the Chinese robotics ecosystem. Now, more than ever, is the time for both humans and machines to learn. The next token of my life sequence will be an important one. To colleagues and investors that bet on 1X early, even before we became a household name - I thank you from the bottom of my heart. I won't forget it♥️