About & Methodology
How we measure how cooked humanity is.
Are We Cooked? is a daily-updated dashboard that tracks the pace and impact of AI progress across multiple dimensions. It aggregates public data across multiple dimensions — from hard data like ArXiv papers and benchmark scores, to vibe checks like the Doomer Ratio and Normie Awareness Index.
It aggregates data from 12 sources, feeds it to Claude for analysis, and generates composite scores.
Each sub-index is scored 0-100 by Claude based on that day's data. Scores map to threat levels:
Capability Cooked
Tracks AI benchmark scores (MMLU, SWE-bench, HumanEval), model releases, and capability breakthroughs.
Jobs Cooked
Monitors AI job displacement signals, automation trends, hiring shifts, and labor market data.
Investment Cooked
Tracks AI funding levels, compute growth, enterprise adoption rates, and GPU demand.
Content Cooked
Measures AI-generated content prevalence, code assistant usage, AI paper volumes, and creative tool adoption.
Overall Cooked
A weighted composite of all sub-indices, reflecting the general state of AI disruption.
Hard Data
HuggingFace
Trending models, downloads, and community activity. Feeds Capability Cooked.
GitHub Trending
Trending AI repos, star counts, and growth. Feeds Capability Cooked.
ArXiv
Recent AI paper submissions from cs.AI, cs.CL, and cs.LG. Feeds Capability Cooked.
News RSS
AI headlines from TechCrunch, VentureBeat, The Verge, and Ars Technica. Feeds Content Cooked.
Benchmarks
Latest scores from public leaderboards and model evaluations. Feeds Capability Cooked.
Metaculus
AGI timeline community predictions and forecaster counts. Feeds Capability Cooked.
Inference Cost
Current pricing ($/1M tokens) from OpenAI, Anthropic, and Google. Feeds Investment + Jobs Cooked.
Vibe Check
Hacker News AI Density
Percentage of AI stories on the Hacker News front page. Feeds Content Cooked.
Karpathy Tweets
Recent tweets from Andrej Karpathy — expert sentiment on AI progress. Feeds Capability Cooked.
Disclaimer: This dashboard is for entertainment and informational purposes only. Scores are AI-generated analysis of public data and should not be used for investment, career, or policy decisions. The "cooked" framing is intentionally tongue-in-cheek.