Founder, CEO at @PhysicalAI. Tech leader and executive, interaction designer, scientist. Google, Disney, Sony before. 2019 National Design Award. TED speaker.
San Francisco & Bay Area, USA
Joined August 2011
- Tweets3.2K
- Following331
- Followers5.4K
- Likes2.2K
I have been spending this week at the NTT Data Open Innovation Week in Malaga, Spain. Together with our lead creative engineer, we have been working with NTT customers and the NTT team in rapid development sessions to solve hard customer problems and move towards production. It is super intense, but it is great to go deep on the customer use cases and figure out how Newton physical world models can solve hard challenges for customers.
Ivan Poupyrev, PhD retweeted
Newton’s embeddings turn a window of raw sensor readings into a single point, placed so that moments that look alike in the physical world land near each other.
This public dataset, PAMAP2, shows data recorded when a group of nine people wore three motion sensors, on the wrist, the chest and the ankle, while going through twelve everyday activities.
This visualization tool helps our team see the data in more detail and isolate different states: Rest, walk, run, cycle, stairs and housework each settle into their own region.
Seeing why a model groups two moments together is the first step toward understanding what it tells you about the physical world.
Built with Newton on the PAMAP2 physical activity dataset (Reiss and Stricker, 2012).
Ivan Poupyrev, PhD retweeted
When people hear Physical AI, they picture a humanoid robot. Our CTO, Nick Gillian, makes the case that most of the real opportunity lies in existing equipment rather than in deploying a humanoid workforce.
Every factory is streaming radar, video, vibration, and current. Newton is one model that reads all of it, catches what’s off in real time, and adapts to machines it’s never seen, right at the edge.
Full conversation with @AI_in_the_AM ↓
“The washing machine has changed the world more than the internet has.”
— Ha-Joon Chang, 23 Things They Don’t Tell You About Capitalism
Imagine what Physical AI could do when it automates everything.
Physical AI > Digital AI.
Newton now understands radio frequency (RF) data!
We just released results comparing a new Newton RF model against specialized RF models. Newton RF was trained on a large volume of RF data that looks nothing like other physical measurements, and we improved the model architecture to capture the spatial signal characteristics inherent to radio frequency.
The results: Newton RF beat all five specialized RF models in 18 of 20 radio tasks, including localization, channel modeling, and blind demodulation.
This is remarkable. In RF, models are highly specialized, each carefully designed for a single task: one model to decode, one to localize, one to identify the transmitter, and so on. Jonah Casebeer and @_jaimelien_ have proved that our Newton world model can rule them all!
Even more unexpectedly, Newton RF also performed strongly on non-RF tasks like vibration, audio, EEG, and seismic data. That means we should be able to expand Newton into the one model that understands any physical phenomenon, reads any sensor, and works for any use case. One world model that encodes the entire physical world.
Full research: archetypeai.io/blog/newton-r…
Say “AI in the physical world,” and everyone pictures a robot. The robot is the least interesting part of the story.
The opportunity goes far beyond robots: models need to learn from real sensor data, adapt as physical systems change, and run on the edge and offline.
Thanks to @TechCrunch for having me and giving me the stage to discuss what it will take to bring AI into the physical world with @ethanjb and @IanRountree.
I'll be speaking at the TechCrunch Disrupt Breakout Stage in 15 minutes alongside @ethanjb and @IanRountree.
See you at 11:30 a.m.
Ivan Poupyrev, PhD retweeted
Our Chief Scientist @_jaimelien_ joins the panel at Long Horizon in San Francisco on Oct. 20 to take on one question: “Will World Models Eat Physical AI?”
“World model” is a broad term; some are built to generate or simulate the physical world. Newton is a physical world model that learns from real-world sensor data, so it’s built to understand what is actually happening on a machine or a factory floor.
That’s the view Jaime brings to the debate.
➡️ In SF? Join us: archetypeai.io/resources/eve…
Ivan Poupyrev, PhD retweeted
This advice is great for all Founders! I received it from my daughter on Father’s Day.
Roughly 85% of the global economy is tied to the physical world. Manufacturing, construction, and logistics are where AI can have a real impact.
I joined Inside the Silicon Mind to talk about Physical AI and how one model can generalize across an entire fleet of machines without a custom model for every use case.
Thanks for having me. Full episode: youtube.com/watch?v=ZpOOpAOp…
Imagine asking a motor how it’s doing, and an AI agent giving you a response in natural language.
That’s how @ipoupyrev described the type of world model we’re building on Inside the Silicon Mind. Sensors already measure things like temperature, rotation speed, vibration, current and flow. Physical AI is capable of reading those signals and tells you what’s going on inside the machine, in words you can act on.
Ivan and the hosts get into how it all works. Have a listen: youtube.com/watch?v=ZpOOpAOp…
Physical AI systems are already pretty good at describing the world. The next logical step is to act on it: perceive, decide, act, verify.
Next week, I’ll be on a panel with @ethanjb, and @IanRountree at TechCrunch Disrupt.
Our panel is called “Physical AI isn’t a Model Problem, it’s an Execution Problem.” Join us to understand what separates intelligence that can generalize to the variability of the physical world from intelligence that doesn’t.
Our mission is to help humanity make sense of the real world. Making sense of it only matters if it leads to action.
More than half of industrial downtime still traces back to manual handoffs. The next step for Physical AI is closing that loop, from perception to action to verification.
On Oct 15, our CEO @ipoupyrev takes the stage at TechCrunch Disrupt in San Francisco. He'll break down what it takes for AI to handle the messiness of the real world.
🔗 techcrunch.com/events/techcr…
Ivan Poupyrev, PhD retweeted
Across 63 wind turbines, Newton found nine failure modes no one had seen before.
Anomaly Discovery Agent, powered by our world model, Newton, reads patterns across sensor streams and surfaces the issues no single signal stream can identify. Newton did this out of the box: no failure history to label first, and no model built turbine by turbine.
archetypeai.io/agents
Proud of our team! @FastCompany named Archetype’s world model, Newton, one of its Next Big Things in Tech 2026 for foundational AI. Our ambition: encode the entire physical world in one AI model, extending our understanding beyond the limits of human perception.
Honored to see @PhysicalAI alongside @cerebras @liquidai @GoodfireAI, pushing the boundaries of AI together. Onward!
Video is overrated. Nature’s vital signs—sound, vibration, electromagnetic signals, motion—reveal faults, strengths, and what’s happening inside an oil well, furnace, or nuclear reactor. No camera. Like a doctor saying “cut carbs, exercise more” just by reading your blood sugar and weight alone, without ever seeing you.
Ivan Poupyrev, PhD retweeted
"Large language models may know a lot, but because they have been trained on text and images, that limits their understanding of our world. By contrast, Archetype AI’s Newton foundation model can interpret data from cameras, lidar, radar, inertial measurement units, and other sensors used in enterprise domains such as construction, logistics, manufacturing, and urban planning."
- @harrymccracken, @FastCompany
fastcompany.com/91612044/fou…
“Companies that are still in denial about Physical AI may not be around to see it play out.”
100% agree with David Brousell: AI moves gradually, then suddenly, as Hemingway put it. Thanks to Shahid Ahmed and @nttdata_inc for hosting the Physical AI Summit.
