Generalist AI Shows GEN-1.5 Robot Foundation Model
Video shows dual-arm robot learning new manipulations from seconds of human demonstration after pretraining.
Generalist AI posted a video of its dual-arm robot system operating as a foundation model for physical tasks. Replies from researchers and founders describe the model learning novel manipulations such as phone flipping after brief human demonstrations. Several posts compare the release to the GPT-3 moment for robotics. The company's linked post states that the capabilities come from pretraining on large-scale physical data and appear in-context after short demonstrations.
Introducing GEN-1.5, a one-shot learner. It can learn new tasks in a few seconds. Show it what to do, and it generalizes. This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world.
Reactions from ranked influencers
42 postsOur latest pre-training comes with zero-gradient / in-context learning! So crazy to witness and glad we can talk about it publicly now.
Introducing GEN-1.5, a one-shot learner. It can learn new tasks in a few seconds. Show it what to do, and it generalizes. This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world.
amazing flex robotics will become real this year
Introducing GEN-1.5, a one-shot learner. It can learn new tasks in a few seconds. Show it what to do, and it generalizes. This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world.
Ever since I started working on robot foundation models a handful of years ago, the broad ability to one-shot in-context learn has been the single most vivid goal in my mind along the long road ahead. It’s something @andyzengineer and I have especially been thinking together on…
Introducing GEN-1.5, a one-shot learner. It can learn new tasks in a few seconds. Show it what to do, and it generalizes. This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world.
It feels like almost every day at this company there is another significant result! The speed of progress makes my job of documenting and communicating what we are doing a challenge, but it's also so much fun!
Introducing GEN-1.5, a one-shot learner. It can learn new tasks in a few seconds. Show it what to do, and it generalizes. This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world.