patio11: frontier model output jumped from 'acceptable intern' to 'frighteningly good' in two releases
TheZvi · x · 2026-09-29
patio11 observes that within roughly the last two model releases from the big labs, output quality went from "acceptable output from the median low-seniority coworker" to "frighteningly good."
He argues that people who don't use the models daily are unlikely to grasp this leap, so he's saying it explicitly.
More from AGI Musings
- Physicist Kyle Cranmer: AI's wins show the value of grinding, not new ideas — KyleCranmer · 2026-09-29
- Chris Manning on why LMs learn verb categories first, and why he thinks LeCun is wrong about language — ziv_ravid · 2026-09-29
- Half the TL mourns skills-ification of hard-won engineering, half deletes all unit tests — vboykis · 2026-09-29
- TypeSafe AI researcher: OpenAI's definition of AGI is 'extremely doable', yet rote work stays unautomated — hardimanjames · 2026-09-29
- Emergence debunked? Researcher argues emergent LLM abilities are labels in your imagination — gerardsans · 2026-09-29
- David Sacks challenges frontier labs: aligned to whose values, exactly? — DavidSacks · 2026-09-29