61
Extra Crispy 🍳

Open-source parity narrows the moat while scaling returns diminish — the hype cycle wobbles

OpenAI
AI-001
72
Extra Crispy 🍳
MODEL_COOKED
65
TALENT_COOKED
70
MOAT_COOKED
75
DRAMA_COOKED
80

Multiple senior researchers departed over safety culture concerns

xAI
AI-005
68
Extra Crispy 🍳
MODEL_COOKED
60
TALENT_COOKED
65
MOAT_COOKED
68
DRAMA_COOKED
80

Grok benchmark position improving but still not frontier

Google DeepMind
AI-003
58
Heating Up 🔥
MODEL_COOKED
60
TALENT_COOKED
45
MOAT_COOKED
55
DRAMA_COOKED
72

Gemini image generation controversy revealed rushed deployment culture

Anthropic
AI-002
52
Heating Up 🔥
MODEL_COOKED
55
TALENT_COOKED
40
MOAT_COOKED
55
DRAMA_COOKED
58

Amazon and Google backing creates dependency risk, not security

Meta AI
AI-004
38
Getting Warm 🌡️
MODEL_COOKED
35
TALENT_COOKED
40
MOAT_COOKED
42
DRAMA_COOKED
35

Llama adoption across enterprise and research represents genuine success

Intel Drop — February 24, 2026

The industry sits at 61/100 — Extra Crispy territory — which feels about right. Five companies racing to build artificial general intelligence while their moats erode daily, their talent gets poached, and open-source communities replicate their products for free. The hype cycle is wobbling.

OpenAI leads the cookedness leaderboard at 72, a distinction that combines genuine business risk (talent drain, open-source competition) with maximum drama (Sam Altman's governance chaos is the gift that keeps giving). xAI follows close behind at 68, where the Elon Musk chaos multiplier inflates the drama score to near-maximum levels regardless of what Grok actually ships.

Meta is the plot twist: the least cooked at 38, because choosing to give away Llama means they can't be disrupted by the very strategy they invented. Anthropic and Google sit in the middle, both heating up but not yet crispy — they have real resources and real products but face the same structural question everyone else does: what's your moat when Llama runs on a laptop?

  • 1.Open-source model quality now matches proprietary models on most benchmarks
  • 2.AI company talent drain continues as researchers move to better-funded startups
  • 3.Meta's Llama strategy appears to be working — commoditizing the competition
  • 4.SEC filings show Meta and Google R&D spend growing but returns unclear
  • 5.Elon Musk's political entanglements creating enterprise hesitation around xAI
News RSS
SEC Filings
LMArena
HuggingFace
Layoffs.fyi