PulseAugur
EN
LIVE 04:10:36

Terrence Tao shares ChatGPT's Jacobian conjecture error, corrected by Anthropic

Mathematician Terrence Tao shared a conversation with ChatGPT where the AI model incorrectly addressed the Jacobian conjecture. Anthropic provided a counterexample that falsified ChatGPT's reasoning. This exchange offers insight into how AI models handle complex mathematical problems compared to leading mathematicians. AI

IMPACT Highlights limitations in current LLMs' mathematical reasoning and the role of external corrections in identifying errors.

RANK_REASON Commentary on an AI model's performance on a mathematical problem, featuring a prominent mathematician and a correction from another AI lab.

Read on Mastodon — mastodon.social →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Terrence Tao shares ChatGPT's Jacobian conjecture error, corrected by Anthropic

COVERAGE [1]

  1. Mastodon — mastodon.social TIER_1 English(EN) · waynerad ·

    Terrence Tao publicly shared a conversation he had with ChatGPT about the Jacobian conjecture, now falsified, by a counterexample from Anthropic. So if you ever

    Terrence Tao publicly shared a conversation he had with ChatGPT about the Jacobian conjecture, now falsified, by a counterexample from Anthropic. So if you ever wanted to see how your conversations with AI compare with a top mathematician, here you go. https:// chatgpt.com/share/…