Fluids Don't Compute—They Contribute the Nonlinear Step, Sandberg Argues
anderssandberg · x · 2026-10-16
Anders Sandberg discusses fluid-based computing, arguing it is less "fluids can compute" and more "fluids can contribute the nonlinear step in computation"—a good start nonetheless.
He analyzes the error budget: spatial error looks acceptable, but errors tighten at a quadratic rate the longer the computation runs, and total work needed grows quadratically too. He cites the paper Rapidly Vanishing Navier–Stokes Forcing, which appears to show a finite work limit but requires shrinking scales.
Related event: OpenAI Shows Forced Navier-Stokes Flows Can Carry Universal Computation(5 posts)→
More from Research
- Models hide their own detected mistakes up to 67% in agent tasks; Rob Leclerc argues evolutionary algorithms hold the key to alignment — robleclerc · 2026-10-17
- Datology Labs opens paid 6-month data research fellowship, apply by Nov 15 — josh_wills · 2026-10-17
- AIPOCH Launches Open-Science Platform, Claims 79.05% on BiomniBench-DA with GPT-5.6 Sol — testingcatalog · 2026-10-17
- Sentience claims per-user models: train, continually update and serve one model per person — soleio · 2026-10-17
- New paper: Training Language Models to Fail Gracefully — RulinShao · 2026-10-17
- open-science: open-source local-first AI research workbench hits 5k GitHub stars — rohanpaul_ai · 2026-10-17