Carmack critiques streaming RL: discarding observations is poor design
ID_AA_Carmack · x · 2026-08-17
John Carmack critiques the trend of "streaming RL" where observations are used once and discarded. He argues that while storage costs are valid, an optimal buffer size is not "one," and even small managed buffers of sparse observations offer value.
More from Research
- Apple Releases LGTM: First Native 4K Feed-Forward Textured Gaussian Splatting — rsasaki0109 · 2026-08-17
- MobileMem: Benchmark for On-Device Long-Term Memory — openkg · 2026-08-17
- ByteDance Studies Domain Data Repetition in LLM Pretraining — ByteDance-Seed · 2026-08-17
- PRM-as-a-Judge 1.5: Toolkit for Robot Process Assessment — Yuyang Liu · 2026-08-17
- RealReplicaBench: All AI Agents Fail E-commerce Test, Top Score 56.1 — APPSO · 2026-08-17
- RLVR: The technique behind LLM's coding and math breakthroughs — burkov · 2026-08-17