Forecasting AI Time Horizon Under Compute Slowdowns

2025-11-23

AI Summary

Amid slowing growth in computational hardware capacity, this study investigates its impact on the time horizon—the maximum duration over which AI agents can reliably plan and predict future states. Method: We propose a compute-algorithm coupling model that quantifies the dynamic interplay among training compute investment, algorithmic progress, and time horizon expansion. Using empirical data fitting, trend extrapolation, and OpenAI’s compute forecasts, we conduct scenario-based simulations. Contribution/Results: We present the first interpretable, compute-driven model of time horizon evolution. Our analysis demonstrates that time horizon growth scales approximately linearly with compute growth rate, thereby refuting the hypothesis that software-only advances could trigger an intelligence explosion or technological singularity. Critically, we project that current compute slowdowns will delay achievement of a “one-month time horizon (80% reliability)” by approximately seven years relative to naive extrapolations—substantially revising mainstream expectations of AI capability timelines.

Technology Category

• Humans and AI: Other Foundations of Human Computation & AI
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Philosophy and Ethics of AI: Artificial General Intelligence

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

METR's time horizon metric has grown exponentially since 2019, along with compute. However, it is unclear whether compute scaling will persist at current rates through 2030, raising the question of how possible compute slowdowns might impact AI agent capability forecasts. Given a model of time horizon as a function of training compute and algorithms, along with a model of how compute investment spills into algorithmic progress (which, notably, precludes the possibility of a software-only singularity), and the empirical fact that both time horizon and compute have grown at constant rates over 2019--2025, we derive that time horizon growth must be proportional to compute growth. We provide additional, albeit limited, experimental evidence consistent with this theory. We use our model to project time horizon growth under OpenAI's compute projection, finding substantial projected delays in some cases. For example, 1-month time horizons at $80%$ reliability occur $7$ years later than simple trend extrapolation suggests.

Problem

• Forecasting AI time horizon growth under potential compute scaling slowdowns
• Modeling time horizon as function of compute investment and algorithmic progress
• Projecting delays in AI capability timelines due to compute constraints

Innovation

• Modeling time horizon as function of compute
• Incorporating compute investment into algorithmic progress
• Projecting AI capability delays under compute slowdowns