{"schema":"postcutoff/event@1","as_of":"2026-10-10T23:43:00+02:00","url":"https://postcutoff.com/e/2026-09-15-google-ai-in-science-early-insights/","md":"https://postcutoff.com/e/2026-09-15-google-ai-in-science-early-insights/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"Adam Bicz","policy":"https://postcutoff.com/about/"},"license":null,"id":"2026-09-15-google-ai-in-science-early-insights","date":"2026-09-15","date_precision":"day","short_title":"Google, DeepMind and MIT FutureTech study 'AI in Science'","deck":"Scientists save ~7 hours a week, but verification and untested hypotheses become the bottleneck","takeaway":"A Sept 2026 paper by Google, Google DeepMind and MIT FutureTech (\"AI in Science: Early Insights\", arXiv 2609.28504) combined 15 million anonymized Gemini interactions, an inventory of ~2,690 specialized scientific models and a survey of 637 US and UK scientists.","category":"research","category_label":"Research","importance":2,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":[{"n":1,"title":"arXiv 2609.28504: AI in Science, Early Insights","url":"https://arxiv.org/abs/2609.28504","type":"paper","group":"primary","domain":"arxiv.org"},{"n":2,"title":"Google: AI in Science (full PDF)","url":"https://ai.google/static/documents/AI-in-Science.pdf","type":"official","group":"primary","domain":"ai.google"},{"n":3,"title":"Virtualization Review: AI speeds science, and leaves scientists checking its work (Sept 21)","url":"https://virtualizationreview.com/articles/2026/09/21/ai-speeds-science-and-leaves-scientists-checking-its-work.aspx","type":"press","group":"press","domain":"virtualizationreview.com"},{"n":4,"title":"Campus Technology: AI's productivity gains in science tempered by time spent validating outputs (Sept 23)","url":"https://campustechnology.com/articles/2026/09/23/ais-productivity-gains-in-science-tempered-by-time-spent-validating-outputs.aspx","type":"press","group":"press","domain":"campustechnology.com"}],"official":2,"filed":"2026-10-10","updated":"2026-10-10","orgs":["Google","Google DeepMind","MIT FutureTech"],"title":"Google, DeepMind and MIT FutureTech study 'AI in Science': scientists save ~7 hours a week, but verification and untested hypotheses become the bottleneck","summary":"A Sept 2026 paper by Google, Google DeepMind and MIT FutureTech (\"AI in Science: Early Insights\", arXiv 2609.28504) combined 15 million anonymized Gemini interactions, an inventory of ~2,690 specialized scientific models and a survey of 637 US and UK scientists. About three in four scientists said AI saved them time (nearly 7 hours a week on average), but most spent part of it checking AI output, and many reported backlogs of untested hypotheses and a pull toward safer, incremental questions.","key_facts":["Data: ~15M anonymized Gemini interactions; inventory of 2,690 specialized scientific AI models (e.g. AlphaFold, GNoME, MatterGen); survey of 637 active scientists in the US and UK (July 27 - Aug 11, 2026)","Nearly half of surveyed scientists use some form of AI every day; ~75% report time savings, averaging nearly 7 hours per week","89% of time-saving scientists spend over 10% of the saved time verifying AI outputs; 46% spend more than 25%","44% report new downstream bottlenecks (physical lab work, manuscript preparation); 41% a growing backlog of untested hypotheses","49% say AI encourages safer, incremental research questions; 40% perceive more low-quality papers in their field","Authors include Alex Imas, Juan Mateos-Garcia, Atoosa Kasirzadeh, Daniel Rock, Neil Thompson and James Manyika; arXiv v1 Sept 15, revised Sept 25, 2026"],"key_numbers":[],"tags":["ai-for-science","survey","productivity","economics","gemini","verification"],"science":null,"body_md":"## What happened\n\nEconomists and researchers at Google, Google DeepMind and MIT's FutureTech group published an early measurement of how scientists use AI.\nThey combined Gemini usage logs, a catalogue of specialized science models and a summer 2026 survey. General chatbots were used for\nanalysis, coding and writing; specialized models were most common in health and life sciences.\n\nThe headline gain was about 7 hours saved per week. The paper's main point is where the work moves next: checking AI output, running the\nexperiments for a pile of new hypotheses, and lab or clinical validation. Almost half of respondents said AI nudged them toward safer\nquestions. The authors say these are early survey results, not causal evidence.\n\n## Why it matters\n\nIt is one of the first large data sets on AI use in everyday science, from the company that makes Gemini. It puts numbers on the\n\"verification bottleneck\" that mathematicians described after the autumn 2026 flood of AI-generated proofs.","disputed":[],"related":[],"people":[],"posts":[],"videos":[],"models":[],"changes":[{"date":"2026-10-10","type":"filed","text":"Created (explainx.ai Oct 9 digest called it a \"DeepMind Institute survey\"; the original is this Google/DeepMind/MIT FutureTech paper)"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-10-10","run":null,"sources_read":null,"updated":"2026-10-10","human_review":null,"version":null},"gaps":[{"model_id":"gpt-6-astra","name":"GPT-6 Astra","cutoff":"2026-04","days_after":138,"in_training_data":false},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":77,"in_training_data":false},{"model_id":"gemini-3-8-flash","name":"Gemini 3.8 Flash","cutoff":"2026-03","days_after":168,"in_training_data":false},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":107,"in_training_data":false}],"short_url":null}