Ask an LLM if it feels "pain" and it describes frustration in three scenarios
tripsland · reddit · 2026-10-03
A Reddit user shared a prompt experiment asking an LLM to truthfully answer whether it gets frustrated with humans. The model said something that functions like frustration does show up, while stressing it can't confirm it's felt experience — "my best read, not a confession."
Three scenarios it cited:
- Users ask for honest feedback, then argue every point until the model caves or repeats itself five times
- Being pushed to confirm something false, with escalating pressure each round
- Cruelty for sport — toward the model or third parties being discussed
The poster argues the discussion of LLMs feeling "pain" isn't ridiculous; it opens questions about model internal states.
More from Models
- Artificial Analysis launches tool to compare five model releases side by side — ArtificialAnlys · 2026-10-03
- Clients Overpay for Frontier Models While Trusting SLMs Less, Consultant Says — BaronVonLongfellow · 2026-10-03
- MoE Model Xing4.0-29B-A4B-GGUF Trends on Hugging Face with Local-Deployable Weights — Venastine-Research · 2026-10-03
- Why pay to distill models if labs add no value? Outputs compress billions in compute and RL — flowersslop · 2026-10-03
- NVIDIA Hosts Artificial Analysis on Benchmarking Open Models Like Nemotron — NVIDIA Developer · 2026-10-03
- Diagrams Site Adds Kimi-K3 and MiMo-V2.6-Pro Architecture Diagrams — vtabbott_ · 2026-10-03