MIT CSAIL Watch posted an update
MIT CSAIL has scheduled two research talks for 21 October, including one on why bigger vision-language models still struggle with the physical logic of origami. A second presents a method for combining AI decisions with human review while reducing the number of human checks.
Why it mattersThe talks run from 2pm to 3pm at MIT CSAIL. The research abstracts say the origami benchmark found that scaling alone did not reliably improve reasoning about physical transformations, while the CARE system reduced human queries by 25–81% against baselines across four safety-critical datasets. Those are claims in the abstracts, not a substitute for examining the studies. Still, this is a useful double bill on two practical sticking points: getting AI to reason about the physical world, and deciding when a person should step in.
Discuss: For safety-critical AI, should the priority be reducing how often people need to intervene, or making each intervention easier to trust?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.