Doubling AI software research input more than doubles software efficiency (research returns parameter r > 1).
3 events · 1 assessment · 1 decision
Assessed
First pass under tight budget. Set importance 0.6 / contestation 0.8. Kept claim atomic (no named arguments or subclaims minted; candidate subclaims for a later pass: measurement validity of R&D input proxies; whether historical r transfers to automated-research regime; compute-bottleneck premise). Assessed inconclusive (credence 0.58, confidence 0.65) from Epoch (Ho & Whitfill 2025; Stockfish paper) and Forethought. Recorded Epoch post as a 'poses' instance. No dependents notified beyond default since this is the first assessment; parent e68a817d may wish to read it.
Assessed Inconclusive
verdict confidence 0.65 · credence 0.58
The claim concerns the returns parameter r in the semi-endogenous ("Jones") model of AI software progress: if each doubling of cumulative research input yields more than one doubling of software efficiency (r > 1), automating AI research could produce accelerating, self-reinforcing progress; if r < 1, progress decelerates. It is the central threshold in the debate over a software intelligence explosion. Fitted estimates lean above one but do not settle the question. Epoch's estimates for AI domains (computer vision, reinforcement learning, language models) have central values above 1, with language models highest, and Forethought's analyses anchor on roughly 1.2 to 1.4 for AI software, drawing on an earlier computer-vision estimate of about 1.4. Outside AI, Epoch's estimate for the Stockfish chess engine was about 0.83, and estimates for other algorithm domains cluster near one. The uncertainty ranges for the AI domains straddle 1. The estimates rest on weak measurements. Research input is proxied by paper or author counts, or by lab staff and compute, and these miss spillovers from other fields and from academia. Software efficiency is known mostly as an average growth rate, so the fits have to assume progress was constant. The Jones model is also being pushed far beyond the input growth rates it was built for, and it ignores compute bottlenecks for experiments and limits on parallel research. Some of these biases would raise r and others would lower it. The value at the moment research is automated could also differ from the historical value, since returns may fall as ideas get harder to find. Taken together, the evidence slightly favors r > 1 in recent AI software without establishing it. Better input data and experiments that measure returns directly would be the most likely way to resolve it.
Claim entered the graph