Catching up with the Archetype AI dev team in Belgrade. It's incredible how easy it is to build a distributed team worldwide these days. Still, nothing beats face-to-face interaction.

Really enjoyed speaking at SRI on the 'AI in the Wild' panel organized by Post Industrial Institute Frode Odegard and moderated by Lawrence Lee, Head of Refinery at SRI and former Xerox Parc Head of Strategy. We discussed how to take AI into the physical world, including digital…

This is the team pioneering #PhysicalAI ! With backgrounds from Google Sony Disney NASA Samsung US MIT School of Engineering and Stanford University, the founders have a strong track record of building and shipping advanced AI for the physical world. Learn more about us: archetypeai.io/about

We are pioneering #physicalAI , a new form of AI capable of perceiving, understanding, and reasoning about the physical world. Think about it like Chat GPT but for physical world information.

We have a detailed blog about it. Check it out here: archetypeai.io/blog/what-is-p…

Looking forward to join the panel at Stanford Research Institute @SRI_Intlon 'AI in the Wild: Beyond the Cloud' to discuss future of large-scale AI-based systems in the physical world. 🌐🤖

🗓️ April 30, 2024
🕟 4:30 PM - 7:30 PM
📍 Open to everyone

More information, program…

I must say that the Tuk-Tuk drivers I rode with in Bangkok and Delhi years ago were pretty wild. They would fit right into this race

'An API for the real world to enable intelligence at every scale,' is how Tara Tan, Strange Ventures partner and very early investor in our company, explains the potential of Newton, the Large Behavior Models (LBM) that we are building at Archetype AI. In her blog post, she…

Interesting interaction metaphor for physical AI from Archetype AI..

Introducing 'Semantic lenses'—our interaction metaphor for a dynamic #PhysicalAI . Unlike traditional systems, it doesn't wait for questions. It decodes sensor data into real-time insights, predictions, and suggestions. Ready to see it in action? Check out our demo!
#ArchetypeAI

And that is why having a good theory is so critical.

Article from MIT Sloan School of Management talks about Large Sensor Models — foundation models that understand sensor data. It argues for the emergence of L[X]M models where [X] is the training data category. This is quite close to what we do in Archetype AI. Source: mitsloan.mit.edu/ideas-made-to-…

In our multi-year strategic partnership with @infineon , we're reshaping the vision of the 'smarter' home, tackling real-world challenges by integrating AI into its sensor portfolio, leveraging Archetype's AI tech.
Read more: archetypeai.io/blog/infineon-…

This is really good one. This is also very common in Japan business culture.

Bold vision on physical AI from Google's Demis Hassabis at last
TED Talks: 'AI will help us understand the world [...] my dream is to understand reality at the Planck scale — the resolution of reality.'

We are doing it, but at #Newton scale.

What are these two wires on Boston Dynamics' latest robot? Our theory is that they are patch antennas for Wi-Fi, used to control the robot.

🐄 Check out the cow crossing the road in the video! Newton cannot be fooled, and it turns the heat map off. This is from our customers in Delhi 🇮🇳

This is our Newton model responding to the prompt 'Follow the car in front' by generating a bounding box around the car it 'thinks' is in front. The 'car in front' is actually an ambiguous semantic concept, because all these cars are in front 😎 and it constantly changes. Can…