New data for the physical world could unlock better AI in energy and logistics
ycombinator · x · 2026-07-23
The thread argues AI is superhuman in code, language, and images because it has vast training data, but the physical world remains poorly modeled due to sparse sensors and intuition-heavy systems.
It suggests cheaper sensors plus better foundation models could unlock new datasets for energy, agriculture, logistics, construction, and climate—and that accurate world models are the path to control.
More from Companies & People
- Microsoft robotics lead says his Zurich year is ending after growing a two-person team — m_wulfmeier · 2026-07-23
- A meme turns X into an “agentic tweet company” — var_epsilon · 2026-07-23
- Microsoft’s next AI bet is platform leverage, not better models — luisdans · 2026-07-23
- Samsung brings Gemini Intelligence to foldables, Watch9 and new smart glasses — demishassabis · 2026-07-23
- PyTorch conference schedule spotlights vLLM, DeepSpeed and multi-node training — PyTorch · 2026-07-23
- Companies are creating “workforce orchestrator” roles to split work between humans and AI — Polymarket · 2026-07-23