Discovering New Solutions to Century-Old Problems in Fluid Dynamics
Posted4 months agoActive4 months ago
deepmind.googleResearchstory
calmpositive
Debate
20/100
Fluid DynamicsAI ResearchMathematics
Key topics
Fluid Dynamics
AI Research
Mathematics
Google DeepMind's blog post shares new solutions to century-old problems in fluid dynamics using AI, sparking discussion on the precision and computability of unstable singularities.
Snapshot generated from the HN discussion
Discussion Activity
Light discussionFirst comment
3d
Peak period
2
78-84h
Avg / period
2
Key moments
- 01Story posted
Sep 18, 2025 at 11:09 AM EDT
4 months ago
Step 01 - 02First comment
Sep 21, 2025 at 9:26 PM EDT
3d after posting
Step 02 - 03Peak activity
2 comments in 78-84h
Hottest window of the conversation
Step 03 - 04Latest activity
Sep 21, 2025 at 9:30 PM EDT
4 months ago
Step 04
Generating AI Summary...
Analyzing up to 500 comments to identify key contributors and discussion patterns
ID: 45290636Type: storyLast synced: 11/20/2025, 5:51:32 PM
Want the full context?
Jump to the original sources
Read the primary article or dive into the live Hacker News thread when you're ready.
Can one "find" an unstable singularity if writing it down requires an endless computational process to produce the incremental digits as one marches towards infinite precision? It sounds like chasing a set of measure zero.
Anyone have an ELI5? I can't quite put my finger on what's happening here. It sounds like the PINNs are being used as fancy discrete basis functions to solve similarity transformed fluid equations so they can chase the unstable solutions comprising a set of measure zero. I don't know how one thinks about grid convergence in these settings. Does one need PINNs or could traditional discretization approaches accomplish the same?