Is the 2x AI R&D speedup compounding? One year of data can't yet tell a level shift from an RSI loop
austinc3301 · x · 2026-11-10
Does the 2x acceleration compound? A one-off 2x is huge but isn't an intelligence explosion — it depends on each better model further speeding progress. METR's redacted evidence hints the speedup "could be above 2x and increasing," though they'd be surprised if far higher. Author's verdict: signs of compounding exist but are very weak; one year of data can't distinguish a level shift from the start of an RSI loop.
More from AGI Musings
- Patrick Collison: status lags by a generation or more — _AustinCalvert_ · 2026-11-10
- MIRI launches 'Do Not Accept' campaign challenging AI companies' decisions for humanity — DavidSKrueger · 2026-11-10
- FDT Debate: jessi_cata Argues FDT Is Logical, Policy Dependence Is a Different Concept — jessi_cata · 2026-11-10
- Top capabilities researchers forming alignment reading groups, says it's now P0 — tszzl · 2026-11-10
- Frogdie exec: compiling code once needed a datacenter — few saw desktop power coming — draginol · 2026-11-10
- Remote work dying the same year everyone got an agent is 'the dumbest timing' — StewartalsopIII · 2026-11-10