Back
Research note: A simpler AI timelines model predicts 99% AI R&D automation in ~2032
webCredibility Rating
4/5
High(4)High quality. Established institution or organization with editorial oversight and accountability.
Rating inherited from publication venue: METR
This METR research note offers a simplified but concrete quantitative model for AI timelines, relevant to researchers and policymakers who need tractable frameworks for forecasting transformative AI and planning safety interventions accordingly.
Metadata
Importance: 72/100working paperanalysis
Summary
METR presents a simplified quantitative model for AI development timelines, forecasting that AI systems could automate approximately 99% of AI R&D tasks by around 2032. The model focuses on measurable proxies for AI capability growth and attempts to translate task automation benchmarks into concrete timeline predictions. It serves as a research note offering a tractable framework for reasoning about transformative AI arrival.
Key Points
- •Models AI progress using task automation as a measurable proxy, projecting ~99% AI R&D automation by approximately 2032.
- •Deliberately simplified approach aims to make timeline reasoning more tractable and transparent compared to complex multi-factor models.
- •Connects capability benchmarks to real-world R&D automation milestones, grounding abstract forecasts in concrete metrics.
- •Published by METR, an organization focused on evaluating frontier AI capabilities and risks.
- •Highlights the potential for recursive AI-driven acceleration of AI development itself once R&D automation reaches high levels.
Cited by 1 page
| Page | Type | Quality |
|---|---|---|
| AI Timelines | Concept | 95.0 |
Cached Content Preview
HTTP 200Fetched Apr 7, 202618 KB
A simpler AI timelines model predicts 99% AI R&D automation in ~2032 - METR
Research
Notes
Updates
About
Donate
Careers
Search
-->
Research
Notes
Updates
About
Donate
Careers
Menu
Research note: A simpler AI timelines model predicts 99% AI R&D automation in ~2032
CONTRIBUTORS
Thomas Kwa
DATE
February 10, 2026
SHARE
Copy Link
Citation
BibTeX Citation
×
@misc { a-simpler-ai-timelines-model-predicts-99-ai-r-d-automation-in-2032 ,
title = {A simpler AI timelines model predicts 99% AI R&D automation in ~2032} ,
author = {Thomas Kwa} ,
howpublished = {\url{https://metr.org/notes/2026-02-10-simpler-ai-timelines-model/}} ,
year = {2026} ,
month = {02} ,
}
Copy
Thomas Kwa
In this post, I describe a simple model for forecasting when AI will automate AI development. It is based on the AI Futures model , but more understandable and robust, and has deliberately conservative assumptions.
At current rates of compute growth and algorithmic progress, this model’s median prediction is >99% automation of AI R&D in late 2032. Most simulations result in a 1000x to 10,000,000x increase in AI efficiency and 300x-3000x research output by 2035. I therefore suspect that existing trends in compute growth and automation will still produce extremely powerful AI on “medium” timelines, even if the full coding automation and superhuman research taste that drive the AIFM’s “fast” timelines (superintelligence by ~mid-2031) don’t happen.
Why make this?
The AI Futures Model (AIFM) has 33 parameters; this has 8.
I previously summarized the AIFM on LessWrong and found it to be very complex. Its philosophy is to model AI takeoff in great detail, which I find admirable and somewhat necessary given the inherent complexity in the world. More complex models can be more accurate, but they can also be more sensitive to modeling assumptions, prone to overfitting, and harder to understand.
AIFM is extremely sensitive to time horizon in a way I wouldn’t endorse.
In particular, the “doubling difficulty growth factor”, which measures whether time horizon increases superexponentially, could change the date of automated coder from 2028 to 2049! I suspect that time horizon is too poorly defined to nail down this parameter, and rough estimates of more direct AI capability metrics like uplift can give much tighter confidence intervals.
Scope and limitations
First, this model doesn’t tre
... (truncated, 18 KB total)Resource ID:
038e3b06af2e1796 | Stable ID: sid_ZNuOMBs7di