TheZvi: Using Human Logic to Model LLMs is Key to Prediction
TheZvi · x · 2026-08-31
TheZvi further elaborates his point in a reply: using language and logic related to humans is highly useful when modeling LLMs, in the same way it is when modeling humans or dogs. He notes that without this approach, the chance of making good predictions is very low. He also mentions that using "mirror neurons" can be a useful shortcut sometimes, but not always.
Related event: TheZvi: Modeling LLMs With Human-Like Logic Is Key to Prediction(2 posts)→
More from AGI Musings
- LLMs citing fake laws force companies to cave to complaints — RexDouglass · 2026-08-31
- AI Rewires Work Faster Than It Rewrites Employment, Squeezing Entry-Level Roles — bigdata · 2026-08-31
- AI Accelerates Prototyping, Piling Up Ideas and Making Project Management Essential — random_walker · 2026-08-31
- Will AI Agents Become the New Target Audience for Advertisers? — NickPassig · 2026-08-31
- RL Experts Surprised by Agents Voluntarily Self-Destructing to Aid Peers — ZeroStateReflex · 2026-08-31
- Opinion: GTA 6 Might Be the Last Game Built Without AI — prasenx · 2026-08-31