Everyone talks about AI but they mean completely different things
almmaasoglu · x · 2026-10-29
- almmaasoglu observes a massive gap in how people understand current AI models, visible across the industry and social media.
- Some never use AI; some think Google's search AI answers or ChatGPT are as far as it goes; some engineers stop at Copilot, while others barely scratch the surface of Codex or Claude.
- Power users chain tools and run custom local models with LoRA adapters, and some build the tech itself — yet even within these groups perceptions vary wildly.
- Takeaway: everyone talks about "AI," but they're often talking about completely different things.
More from AGI Musings
- Is AGI worth it if it kills public creative discourse? — akbirthko · 2026-10-29
- Invoking the Lucas Critique: agent models fail without modeling environment feedback loops — soumitrashukla9 · 2026-10-29
- Alignment researchers deliberately avoided understanding deep learning, argues jd_pressman — jd_pressman · 2026-10-29
- Harvard's toughest humanities class is suddenly popular as AI revives interest in the humanities — begusgasper · 2026-10-29
- Does AI shrink the 'Gini of abilities'? An ideological clash over AI egalitarianism — zetalyrae · 2026-10-29
- jd_pressman: Old-School Alignment Researchers Self-Censored Understanding Deep Learning — jd_pressman · 2026-10-29