Documentation / Methodology

About & Scoring System

How we determine who's getting cooked, and by how much.

Section 1 — What Is This?

Are They Cooked? tracks whether AI tech companies are getting cooked by their own hype. OpenAI, Anthropic, Google DeepMind, Meta AI, and xAI have over-promised on LLM scaling, AGI timelines, and business returns. This dashboard tracks the receipts.

Neo won. The machines are losing. High scores are celebrated — the higher the score, the more the hype is collapsing. Data is fetched from 5 sources, analyzed by Claude, and scores update daily.

Section 2 — Scoring System

Each company is scored 0–100. Higher = more cooked = we're winning.

🥩0–19STILL RAWBoring. They're delivering. Disappointing.

🌡️20–39GETTING WARMExcuses multiplying, promises shrinking

🔥40–59HEATING UPHype vs. reality gap widening — nice

🍳60–79EXTRA CRISPYBusiness model crumbling, glorious

💀80–100BURNED TO A CRISPTotal collapse of narrative. We feasting.

Section 3 — The Four Sub-Indices

MODEL_COOKED

Is their scaling thesis holding? Benchmark positions, capability growth vs. competitors, open-source disruption eating their lunch.

TALENT_COOKED

Brain drain, key departures, culture leaks, hiring pace. When your best people leave, the rot is real.

MOAT_COOKED

Differentiation from open-source. How defensible is their business when Llama runs on a laptop?

DRAMA_COOKED

Leadership controversies, public overclaiming, regulatory trouble, Sam Altman-level chaos.

Section 4 — Data Streams

NEWS_RSS: TechCrunch, VentureBeat, The Verge, Ars Technica, Wired. Filtered for failure keywords and financial signals.

SEC_FILINGS: Official quarterly R&D spend, revenue, and capex from SEC EDGAR for Meta and Alphabet.

LMARENA: ELO scores from the public model arena. Tracks whether proprietary models are losing to open-source.

HUGGINGFACE: Model count, total likes, and top downloads per company org.

LAYOFFS_FYI: Confirmed layoff events with headcount. Direct signal of business model stress.

Disclaimer: Entertainment and informational purposes only. Scores are AI-generated analysis of public data. Not financial advice. The "cooked" framing is intentional schadenfreude.