- Published on
Frontier model training has quietly shifted from a chip problem to a power-and-capital problem. A grounded look at the four scarce inputs (compute, power, data, talent) and who should (and shouldn't) build from scratch.
Tagged
Economics
All Posts
- ai (3)
- llm (3)
- data-science (2)
- development-tools (2)
- gpu (2)
- aws (2)
- sql (1)
- parser (1)
- claude-code (1)
- rust (1)
- agent (1)
- tui (1)
- distributed-systems (1)
- redis (1)
- distributed-locking (1)
- fastapi (1)
- restful (1)
- h20 (1)
- auto-ml (1)
- r (1)
- ml (1)
- xcode (1)
- mac (1)
- ios (1)
- vllm (1)
- inference (1)
- systems-design (1)
- keras (1)
- machine-learning (1)
- deep-learning (1)
- time-series (1)
- lstm (1)
- python (1)
- alpine (1)
- docker (1)
- travel (1)
- photography (1)
- personal (1)
- agentcore (1)
- opencode (1)
- bedrock (1)
- terraform (1)
- data-lake (1)
- data-engineering (1)
- lakefs (1)
- data (1)
- infrastructure (1)
economics (1)
- strategy (1)