A new paper, 'The Bounded Parallelizability of R&D: Theory and Application to AI' by Philip Trammell, and an Epoch blog post argue that standard models of AI-driven technological singularity overlook …
Model evaluations face fundamental limitations beyond execution errors, according to a new analysis by an AI safety researcher. The piece argues that benchmarks provide decontextualized, point-in-time…
AI benchmarks are saturating fast, and Epoch's Open Problems already task AI with solving useful problems during evaluation. The author proposes extending this approach to directly optimize AI safety …
A third of official solutions in Epoch's FrontierMath benchmark contained errors, making reward-hacking the only way to pass, according to a post by LessWrong user 1a3orn. The author argues that flawe…
Epoch and METR released MirrorCode, a benchmark for long-horizon programming tasks, finding that AI models like Opus 4.7 solved a task in 14 hours for $251 that would take a human 2-17 weeks. Across 2…
Substack has integrated Pangram AI detection, allowing users to assess posts for AI content through the app and Notes, a move that the author says restores faith in the platform. The Berggruen Institu…
Researchers propose an 'expenditure horizon' measure of AI agents' optimization ability, estimating that each 1% improvement in NanoGPT costs roughly $2,500 in human labor, while agentic runs exceedin…
AI safety forecasts are undermined by untestable assumptions about the nature of future AI agents, such as whether they will be humans, institutions, or hybrid collectives. These ontological commitmen…
AI capabilities are accelerating at a better-than-exponential rate, with frontier models from Anthropic, OpenAI, and Google now able to autonomously complete software projects that would take human te…
Arena, an AI evaluation platform born at UC Berkeley, reached a $100M annual revenue run rate eight months after launching its product, as demand surges for benchmarks that measure real-world AI utili…
AI progress is driven primarily by massive amounts of domain-specific human expert data and reinforcement learning, not by architectural breakthroughs, creating a data black hole that powers model cap…
A new analysis using Epoch's ECI metric shows that open-weight AI models continue to trail closed models on the frontier, with the gap persisting over time. The analysis, based on item response theory…
Economists Alex Imas and Phil Trammell discuss the economic implications of artificial general intelligence, including optimal tax and redistribution policies, the potential for inequality, and what w…
Open-weight AI models now trail proprietary state-of-the-art systems by an average of four months in performance, according to Epoch's internal capability metric. The Epoch Capability Index, which agg…
Memory's share of AI chip component costs rose from 52% to 63% between Q1 2024 and Q4 2025, according to an analysis of Nvidia, AMD, Google, and Amazon chips. Total component spending on AI chips grew…