New paper: do LLMs solve cognitive development tests like humans?

GolinoHudson · x · 2026-09-17

A new paper by Alexandros Demetriou examines whether LLMs like ChatGPT, Gemini, Grok, and DeepSeek solve cognitive development tests like humans — where they excel, where they fall behind, how they think Descartes's "Cogito ergo sum" applies to them, and whether they believe they possess human-like general intelligence.

Original post →

More from Models

Models channel →