Physicist Sabine Hossenfelder Tests LLMs for Video Scripts: All Fail Completely
skdh · x · 2026-08-01
Physicist Sabine Hossenfelder stated that she has repeatedly tried getting ChatGPT, Claude, Grok, and Gemini to write scripts for her YouTube videos, but they remain a complete failure.
She noted that these models are unable to come up with interesting topics, inevitably suggesting widely reported clichés. Even worse, when given a specific topic, the scripts they produce lack logical coherence. While individual sentences sound fine, the overall text is repetitive, incomprehensible, and useless. They aren't even capable of drafting a basic video structure.
Related event: Tech Blogger Tests Top LLMs for Video Scripts, Finds Them Failing(4 posts)→
More from Models
- Vercel Releases Next.js AI Agent Eval: Kimi K3 and Claude Tie at the Top — evilrabbit_ · 2026-08-01
- Why Did OpenAI Abandon Banked Resets? User Speculates — dejavucoder · 2026-08-01
- Microsoft Teases Astra Model: Proves 10 Major Math Theorems — ctjlewis · 2026-08-01
- Train Your Own Model When Inference Exceeds $750/Day: Pallet's Playbook — marcbhargava · 2026-08-01
- OpenAI Slashes GPT-5.6 Luna API Price by 80%, Rivaling Opus 5 at 1/17 Cost — gabrielchua · 2026-08-01
- Kimi K3 Overhyped? User Reports Real-World Performance Falls Short of Claude and ChatGPT — Global_Knee5354 · 2026-08-01