Vivan Amin

10 Spicy Takes After Talking to 50 Friends at the Frontier of Robotics 🤖🔥 1/ Astra IS robotics' GPT-3 moment. Ironically, now that it's finally here, some in robotics seem reluctant to acknowledge it. Perhaps because it (sadly) didn't come from a robotics company? 2/ Robotics is becoming just another vertical for frontier labs. Not so different from coding in some respects. Why reinvent general intelligence for every industry? 3/ Robotics was too hard to underwrite until now. No convincing pretraining scaling law, no convergence on models, data, or hardware. No wonder some critics dismissed it as a fragmented industry playing an expensive "demo game". Astra may finally offer a credible path to scale. 4/ Robotics foundation-model startups have a case. But how strong is it? The argument: Astra handles System 2 (reasoning), while specialized models retain System 1 (skills) and System 0 (motor control). Fair. But isn't that already a narrower value proposition? And who says the boundary stops there? 5/ Follow the talent. Word is OpenAI reorganized its 30–40-person robotics team around the core Astra model. Hearing similar moves toward general models at @Tesla_Optimus. Org charts may tell us more than any demo. 6/ The robotics data gold rush is real. Million-hour procurement programs, some potentially reaching tens of millions of hours. Talk of budgets approaching $1B per frontier lab. 7/ Bullish on robotics hardware volumes. As intelligence becomes more accessible, the barriers to building useful robots could fall dramatically. The moat shifts toward reliability, dexterity, integration, and deployment. 8/ Robotics could fuel the next semiconductor and memory supercycle. A 100B-parameter model running on the edge requires ~50GB just for 4-bit weights. Now multiply that by millions of robots. Why do you think @JensenHuang is so obsessed with robotics? @elonmusk saw it coming: xitter.gq/elonmusk/status/211477… 9/ Is self-driving immune to the general-model threat? After coding and robotics, why assume self-driving remains the exclusive domain of proprietary, vertical models? Robotaxi increasingly feels like a solved problem to me. But if and when frontier labs move into self-driving, we could see an Apple vs. Android dynamic. Think about the implications for $TSLA, Waymo, and other OEMs. 10/ The next bottleneck may not be intelligence. It's deployment. Models are improving faster than our ability to deploy robots in the messy real world. Integration, hardware variance, failure recovery, and economics remain brutally hard. The real moat may be turning intelligence into reliable, profitable physical work. -- I realize some of these takes may rub people the wrong way, especially those who have spent years doing the hard work of advancing robotics. That's not my intention. I'm sharing them to spark debate, not to declare winners and losers. Astra has only been out for a month or so, and it's far too early to draw definitive conclusions. More importantly, when the TAM is this enormous - potentially the largest in tech - there should be plenty of room for different approaches and many winners. ++ Full diligence report: robonomics.substack.com/p/ro…

2027 will be the year of Physical AI.

🤖 I’m happy to announce rho-alpha 🤖, @Microsoft's first physical AI model for bimanual robotic manipulation 🦾🦾 built on Microsoft’s own phi-series VLM: microsoft.com/en-us/research… Rho-alpha is unusual among VLAs in a few ways. You can think of it as a VLA+, as we are experimenting with including tactile and force feedback into the set of rho-alpha’s “native” input modalities – in part, using simulation data we generate in #NVIDIA #IsaacSim - while also enabling rho-alpha to continually learn from real-time human feedback. Perhaps most importantly, we envision Rho-alpha and its training pipelines as a foundation – a foundation to help @Microsoft's partners in the robotics space to create powerful physical AI models for their hardware using their data. Currently, the predominant paradigm for this is VLA finetuning. Rho-alpha’s tooling affords much richer approaches by injecting the partner’s data into virtually any rho-alpha training stage to produce custom Rho-alpha versions for the partner’s robots and use cases. For the next few months, we will be refining rho-alpha and will report the results in due time. Why are we announcing rho-alpha already now? Because we want to invite you 🫵- robot OEMs, integrators, and operators - to work with us on the refinement process, and encourage you to apply for the Rho-alpha Research Early Access Program here: tinyurl.com/38p6bn4u Giant thanks and congratulations to @MSFTResearch SPEAR team (@t_hellebrekers, Michael Murray, Galen Mullins, @dean4ta__, Simran Bagaria, Harshavardhan Gajarla), our internal (Reuben Tan) and external research collaborators (@patrickhyin, @abhishekunique7), and TPM Vivan Amin for making rho-alpha happen. 👏 We have a lot more motors to fry and tactile sensors to burn out – that’s how progress in physical AI is made. Stay tuned! 🚀 #Robotics #physicalai @NVIDIARobotics