Tech

Alibaba President: AI agents can talk, but can they actually do the work?

1 source ·1 article ·1h ago

The strongest frontier model we tested successfully completed 61.7% of the tasks — high enough to be useful and low enough to be a warning.

Covered by 1 outlet

All coverage · 1 article

The strongest frontier model we tested successfully completed 61.7% of the tasks — high enough to be useful and low enough to be a warning.