AI is actually bad at math, ORCA shows

Thomas Claburn

2025-11-18 1 min read

<h4>ORCA benchmark trips up ChatGPT-5, Gemini 2.5 Flash, Claude Sonnet 4.5, Grok 4, and DeepSeek V3.2</h4> <p>In the world of George Orwell's 1984, two and two make five. And large language models are...

ORCA benchmark trips up ChatGPT-5, Gemini 2.5 Flash, Claude Sonnet 4.5, Grok 4, and DeepSeek V3.2

In the world of George Orwell's 1984, two and two make five. And large language models are not much better at math.…

Source: The Register - Software: AI + ML Word count: 193 words

Published on 2025-11-18 05:16