Developer argues anthropomorphic LLM language misleads readers and obscures corporate accountability
AlexTensor · x · 2026-09-26
In a debate on X, developer AlexTensor argues the burden of proof lies with those using anthropomorphic language for LLMs. His points:
- Attributing human characteristics to statistical models misleads readers and obscures corporate accountability
- Ignoring counterexamples that disprove human intent or understanding could create liability for fraudulent misrepresentation
- LLMs are at best an "imitation engine" whose facade falls apart on novel questions, citing supporting research
He asks why we can't simply state what the software does, who built it, and how well it performs — part of the ongoing dispute over whether models 'understand'.
Related event: Debate Flares Over Anthropomorphizing LLMs as AP Weighs Style Limits(3 posts)→
More from AGI Musings
- Not Boring: the three arguments that AI isn't a bubble are exactly how bubbles form — Not Boring (Packy McCormick) · 2026-11-11
- Viral post: five years of 'maybe next year,' then 2026 arrived — hey_abusiddik · 2026-11-11
- Jeremy Howard: LLMs "cosplay understanding," like the man in Searle's Chinese room — Machine Learning Street Talk · 2026-11-11
- OpenAI board member Paul Christiano: an AI accident is your most likely cause of death — notkilleveryoneist · 2026-11-11
- 'Chat with your data' misses the point — people want a smart coworker — vboykis · 2026-11-11
- Ben Todd: AI task time horizons appear superexponential in AECI, not just revenue — ben_j_todd · 2026-11-11