Conversation

Notices

1. Ars Technica (arstechnica@mastodon.social)'s status on Thursday, 21-Sep-2023 05:35:35 JST Ars Technica

   Telling AI model to “take a deep breath” causes math scores to soar in study

   DeepMind used AI models to optimize their own prompts, with surprising results.

   https://arstechnica.com/information-technology/2023/09/telling-ai-model-to-take-a-deep-breath-causes-math-scores-to-soar-in-study/?utm_brand=arstechnica&utm_social-type=owned&utm_source=mastodon&utm_medium=social

   • John Francis (johnefrancis@mastodon.social)'s status on Thursday, 21-Sep-2023 05:40:18 JST John Francis

     in reply to

     @arstechnica "you think that's air you're breathing?" 🤨

   • Questiam (questiam@mastodon.social)'s status on Thursday, 21-Sep-2023 23:00:56 JST Questiam

     in reply to

     @arstechnica Why are we trying to make LLMs more accurate? They’re built to be generative, not accurate. It’s like trying to prompt a submarine to fly better.