Google and DeepMind analyze 15 million AI interactions in ATLAS 1.0 white paper
minsuk_chang · x · 2026-07-24
Google and Google DeepMind released the ATLAS 1.0 white paper on AI use and the economy, based on 15 million de-identified interactions across the Gemini app, API, and Google AI Mode.
- The dataset spans both work and non-work usage, giving a broad view of how people actually use AI.
- The team spent significant effort classifying interactions into task categories for work and home.
- The report claims it surfaces many interesting patterns, including the overrepresentation of white-collar work in work-related use.
Related event: Google's ATLAS Report: 15M Interactions Reveal Real-World AI Usage(17 posts)→
More from Research
- Open-source decision models sold their own base in Red Alert 2 — then learned until no human could beat them — heathercmiller · 2026-10-27
- DeepMind paper: LLMs can't sample randomly on their own — pass random numbers from code instead — rohanpaul_ai · 2026-10-27
- World Model Assigns Weight and Friction to Objects, Then Simulates with a Physics Engine — jonstephens85 · 2026-10-27
- Token Collapse: Study Finds Graph Soft Prompting Tokens in LLMs Fail to Encode Structure — Sapienza · 2026-10-27
- Mathathon 2.0 launches with Terence Tao's SAIR, $1.5M in compute for teams — _sathvikr · 2026-10-27
- The 'sweet lesson' is real: two kinds of ML problems need opposite strategies — mike64_t · 2026-10-27