Rivian CEO RJ Scaringe founded Mind Robotics in 2025 and raised $1 billion on a single thesis: the robotics field is advancing faster than almost anyone outside it understands. His robots are not humanoids. They have two arms, two hands, and fingers with wider-than-human splay and functional fingernails for manipulating thin objects. They are built for factory floors, not living rooms.
The technical argument Scaringe makes is specific. Transformer-based models, the same architecture driving LLMs, are now enabling robots to learn generalizable motion skills rather than executing prescribed, hardcoded movements. Tell the robot to clean a surface, and it figures out the motions. Show it a credit card, and it uses its fingernails. Show it a phone, and it uses its fingers. That shift from scripted to generalized behavior is what Scaringe calls the inflection point.
The economic logic is the most readable part of the original piece. Scaringe draws a direct contrast with Elon Musk's Optimus: a general-purpose home robot charging $10,000 per year addresses a narrow market, while a factory robot earning back $50,000 per year across three shifts is an obvious industrial buy. Read the full interview for Scaringe's specific framing of why industrial deployment, not consumer deployment, is where the robotics ChatGPT moment actually lands first.
[READ ORIGINAL →]