Twelve dimensions across seven providers. Re-graded as announcements land, with the market's reaction shown alongside.
Own the stack vs depend on partners · updated 2026-06-01
Owns its models and an open-weights strategy that shapes the whole field; building its own silicon.
Does the provider control its own model destiny, or rent intelligence from someone else? Owning frontier training, custom silicon, and a coherent small-to-large model ladder is durable leverage.
Meta's plan to ship four inference-optimized MTIA chips in two years: purpose-built silicon over general GPUs.
Profiles the ~50-person TBD Lab at the core of Meta's superintelligence reorganization.
Each giant's AI strategy flows from its computing philosophy; Meta's 'personal superintelligence' is augmentation.
How Apple, Google, Meta, Microsoft, and Amazon are differently positioned across infrastructure, models, and distribution.
The Llama API preview (OpenAI-SDK compatible) and Llama Stack partner integrations: Meta's developer-platform play.
Meta submitted an experimental chat variant to LMArena that differed from the public weights, a credibility hit.
Zuckerberg's four AI bets, and the argument that Meta spends enormously yet lacks a coherent direction.