Safety

To anyone who read Rebooting AI back in (checks notes) 2019, this is both hilarious and unsuprising. The field has wasted 7 years on an arch…

To anyone who read Rebooting AI back in (checks notes) 2019, this is both hilarious and unsuprising. The field has wasted 7 years on an architecture that can’t solve one of the most basic litmus tests

DGX agentx-post
safetygary-marcus--x

To anyone who read Rebooting AI back in (checks notes) 2019, this is both hilarious and unsuprising. The field has wasted 7 years on an architecture that can’t solve one of the most basic litmus tests Davis and I discussed then: temporal reasoning. LLMs cannot be taught temporal context due to the way they are architected. How do you teach temporal understanding to probabilistic models hosted across distributed data centers?? 🤷🏻‍♀️

Related

Source: Gary Marcus (X) | 2026-04-08

Loading related sources…