Generalist has announced GEN-1, a new physical AI system achieving a 99% success rate across a range of physical tasks. This includes tasks that previously required human skill and muscle memory.
Listen to the article
Hear the article with natural AI narration.
AI explained
What is Generalist's GEN-1 physical AI system?
GEN-1 is an autonomous physical AI system with a 99% success rate on various manual tasks. It improves on its predecessor by adapting and improvising in real time using extensive data from human movements.
- Summary: GEN-1 performs tasks like folding clothes and sorting parts with high precision, adapting after minimal training.
- Why it matters: Its ability to handle unexpected changes and learn quickly enables automation of tasks previously requiring human skill.
- Key point: GEN-1 uses over half a million hours of human movement data to achieve flexible, real-time physical task performance.
GEN-1: An Autonomous System with Improvisation Skills
GEN-1 is built on Generalist’s earlier model, GEN-0, and demonstrates significant performance improvements. The system can adapt and improvise when faced with unexpected situations, a major advantage compared to earlier robots limited to pre-programmed movements. Generalist has collected over half a million hours of “data hands,” capturing micro-movements and visual information from human manual tasks. This has enabled training the model with petabytes of physical interaction data.
GEN-1 can perform tasks such as folding clothes, sorting car parts, and even placing money into a wallet, all with high precision and speed. According to Generalist, the model can adapt to specific tasks after just one hour of training on relevant data. This allows it to handle tasks previously reserved for humans and to adjust to errors and unexpected events along the way. For example, it can modify its grip when small objects are moved, demonstrating its ability to learn and adapt in real time.
Implications for the U.S. Market
AIny brief assessment: GEN-1 offers American developers the opportunity to integrate advanced robotic solutions into manufacturing and service sectors. Its high success rate and improvisational capabilities could boost efficiency in industries reliant on repetitive tasks, making such robots a valuable addition to the U.S. workforce.
Source: Ars Technica
Read the full story in Norwegian
Les på norskRead also: OpenAI Alumni Launch New $100 Million Investment Fund
