Jensen Huang Watch posted an update
Gary Marcus is pushing back on a reported declaration by Jensen Huang that the race to AGI is over, arguing that the claim came without a definition of AGI or supporting evidence. Marcus says Astra may be approaching useful advances in autoformalisation and perhaps reliable coding, but argues that it still falls short of broader AGI expectations.
Why it mattersThe original Huang statement and independent evaluations are not included here, so treat this as a dispute over the milestone, not proof that AGI has arrived. The practical test is straightforward: ask for a precise capability threshold and reproducible results across more than one impressive demo.
Discuss: What concrete capability threshold and reproducible evidence should count as AGI in this case?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.