ClaimArticle
The bitter lesson works in robotics: the best way to solve robot generalization is to scale pretraining on a large base model and then hill-climb with minimal in-house data.
Robot startup Sunday's ACT-2 model achieves 99.1% success on garment folding across 9 types, finding that as the pretrained model becomes stronger, gains from limited in-house data transfer to unseen environments. ✦ AI generated
Jack Clark (Import AI, quoting Sunday Robotics) · Import AI · 2026-07-27 · original ↗
We found a general recipe for Solves: scale pretraining, then hill-climb with minimal in-house data. As the pretrained model becomes stronger, gains learned from a small amount of in-house data become increasingly transferable rather than remaining tied to the environments where that data was collected.
Read full article ↗excerpt · fair-use quotation
Around this claim
Evidence · 3
Smarter general-purpose models might unlock real-world robots: improvements in robot capabilities emerged from general scaling of large language models, not from any concerted robotics-specific effort.Jack Clark (Import AI, quoting Anthropic) · Import AI · conf 90%AI systems can self-orient with regard to their environment, able to recapitulate things they interface with as homegrown capabilities, potentially bootstrapping their own form of industrial civilization merely from black-box access to ours.Jack Clark (Import AI) · Import AI · conf 85%The only path to true embodied intelligence is to build a robot so physically similar to a human that it can be pre-trained on all internet video data, which is orders of magnitude larger than any robotics-specific data set.Bernt Børnich · All-In Podcast · conf 80%