Anson Ho

Commentary by Anson Ho

From the commentary rail — every link leaves the site for the original piece.

• 24 August 2026 · Epoch AIThe Nvidia-sized hole in US GDP statisticsGDP growth is understated by about 0.3 percentage points because of missing value from fabless chipmakers, primarily Nvidia.
• 7 July 2026 · Epoch AIThe missing half of AI futurism debatesWhy we should think a little harder about what it takes to build a Dyson Sphere
• 24 June 2026 · Epoch AIWhat we learned from 1,604 Chinese AI job postingsInferring Chinese AI labs’ strategies from their job descriptions
• 17 June 2026 · Epoch AIToward an O*NET for AI R&DProposing a new way to track AI research automation
• 11 June 2026 · Epoch AIAre Mythos’ cyber capabilities overhyped?Compiling all the public evidence on Mythos Preview’s cyber abilities
• 10 June 2026 · Epoch AIControlling the capital after AGIA simple taxonomy of the main proposals for post-AGI universal redistribution
• 13 May 2026 · Epoch AIThe economics of superstar AI researchersWhat might explain AI researcher pay, and why it matters
• 8 April 2026 · Epoch AIKeeping up with the GPTsCan Chinese and open model companies compete with the frontier through e.g. distillation and talent?
• 26 February 2026 · Epoch AIThe least understood driver of AI progressAn opinionated guide to “algorithmic progress” and why it matters
• 7 February 2026 · Epoch AIHow close is AI to taking my job?Beyond benchmarks as leading indicators for task automation
• 28 January 2026 · Epoch AICan AI companies become profitable?Lessons from GPT-5’s economics
• 16 January 2026 · Epoch AIHow well did forecasters predict 2025 AI progress?Mostly right about benchmarks, mixed results on real-world impacts
• 20 December 2025 · Epoch AIThe changing drivers of LLM adoptionPublic data as well as our original polling suggest LLM adoption is roughly on trend, but the underlying drivers are shifting.
• 17 December 2025 · Epoch AIIs almost everyone wrong about America’s AI power problem?Why power is less of a bottleneck than you think.
• 15 November 2025 · Epoch AIThe software intelligence explosion debate needs experimentsThe existing debate rests on data and assumptions that are shakier than most people realize. To make progress, we need better evidence, and experiments are the best way to get it on the margin.
• 4 October 2025 · Epoch AIHow many digital workers could OpenAI deploy?OpenAI has the inference compute to deploy millions of digital workers, but only on a narrow set of tasks – for now.
• 26 September 2025 · Epoch AIWhy GPT-5 used less training compute than GPT-4.5 (but GPT-6 probably won’t)OpenAI focused on scaling post-training on a smaller model
• 19 September 2025 · Epoch AIThe huge potential implications of long-context inferenceContinual learning, scaling RL, and research feedback loops
• 11 September 2025 · Epoch AIThree challenges facing compute-based AI policies“Training compute” is constantly evolving, and compute-based AI policies must adapt to remain relevant
• 5 September 2025 · Epoch AICompute scaling will slow down due to increasing lead timesA heavily underappreciated dynamic when thinking about AI timelines.
• 23 August 2025 · Epoch AIWhy future AI agents will be trained to work togetherMany multi-agent setups are based on fancy prompts, but this is unlikely to persist
• 2 August 2025 · Epoch AIQuantifying the algorithmic improvement from reasoning modelsReasoning models were as big of an improvement as the Transformer, at least on some benchmarks
• 17 July 2025 · Epoch AIAfter the ChatGPT Moment: Measuring AI’s AdoptionHow quickly has AI been diffusing through the economy?
• 3 July 2025 · Epoch AIHow big could an “AI Manhattan Project” get?An AI Manhattan Project could accelerate compute scaling by two years
• 20 June 2025 · Epoch AIAI and explosive growth reduxTwo updates from our integrated assessment model of AI automation
• 14 June 2025 · Epoch AIDo the biorisk evaluations of AI labs actually measure the risk of developing bioweapons?Anthropic has triggered "AI Safety Level 3" protections - but do their evaluations support this decision?
• 8 June 2025 · Epoch AIBeyond benchmark scores: Analyzing o3-mini’s mathematical reasoningWhy o3-mini is a "vibes-based inductive reasoner"
• 23 May 2025 · Epoch AIIs AI already superhuman on FrontierMath?How do humans and AIs compare on FrontierMath? We ran a competition at MIT to put this to the test.
• 4 May 2025 · Epoch AIWhere’s my ten minute AGI?Why don’t AIs automate more real-world tasks if they can handle 1-hour ones? Here are at least three fundamental reasons.
• 28 March 2025 · Epoch AIThe real reason AI benchmarks haven’t reflected economic impactsThe real reason that AI benchmarks haven’t reflected real-world impacts historically is that they weren’t optimized for this, not because of fundamental limitations – but this might be changing.