Why measure AI by what it replaces, not by what it lets humans do?
Admirable_Wasabi_732 · reddit · 2026-09-10
The author challenges the dominant "replacement" frame for measuring AI progress and revives Douglas Engelbart's 1962 augmentation tradition:
- Current evaluation centers on machine capability: benchmark scores, tasks done without humans, jobs automated — not on new human capability unlocked.
- After Deep Blue beat Kasparov in 1997, Kasparov's Advanced Chess reframed the question from human vs. machine to human alone vs. human + machine.
- Calculators extend narrow skills, telescopes perception, writing memory and reasoning; AI is unusual in potentially extending reasoning, memory, learning, perception, creativity, communication and action simultaneously.
- The author proposes asking two questions per new capability: what can it replace, and what does a human become capable of because it exists?
They invite examples where AI genuinely expanded what a person can do, beyond merely accelerating existing skills.
Related event: Rethinking AI Metrics: Measure Human Augmentation, Not Just Replacement(2 posts)→
More from AGI Musings
- Benjamin Todd: AI progress beat superforecasters over 18 months — ben_j_todd · 2026-10-26
- AI researchers guessed robots wouldn't outrun humans until 2034 — it already happened — ben_j_todd · 2026-10-26
- AI folks, stop saying "biology" when you mean drug development — KevinKaichuang · 2026-10-26
- ESR on the Bugpocalypse: locking down frontier models only hands the game to Chinese open weights — colinmcnamara · 2026-10-26
- Anthropic: junior tech hires fell from 27% to 9% as agent hours hit 6x employee hours — thione · 2026-10-26
- Anthropic's Jascha: you may have only 1–5 years of intellectual contribution before AGI — GabGarrett · 2026-10-26