AI summary by TwinMind
This New Idea Could Explain Complexity
Sabine Hossenfelder
6m 53s
Universe's Building Blocks and Emergence
- Universe made from 25 elementary particles held together by 4 forces
- Creates bewildering variety: galaxies, bacteria, planets, hurricanes, humans, societies, cat memes
- Complexity emerges from simplicity through emergence process
- Emergence: small objects collect into larger ones, larger ones follow new natural laws
- Problem: don't know what emergence actually means
Separation of Scales Concept
- Architects don't need quantum physics - laws of statics emerge from quantum physics
- Individual particle quantum behavior doesn't matter
- Can discuss planetary orbits without knowing exact composition
- Particle physics degree doesn't make good heart surgeon
- "Separation of scales" - different natural laws at different resolution levels
- Deepest level: standard model of particle physics
- Larger scales: atomic nuclei → atoms → molecules → materials → human beings
Complex Systems and Scale Mixing
- Most complex systems (humans, human societies) lack clean separation of scales
- Drug example: tiny molecular reactions affect entire body → affect other people
- Reverse effect: humans built particle colliders affecting shorter scales
- Earth's weather/climate depends on physics of many different scales, mixes them together
- Scale mixing seems essential to complexity
- Understanding emergence = key to understanding complexity
University of Sussex Research Paper
- New mathematical definition for emergence
- Quantifies independence of large-scale structures from underlying details
- Two approaches:
Causal Closure
- Can formulate natural laws at one resolution level without input from underlying detail
- Computer algorithm example: explain logical flow without knowing hardware composition
- Some think computer science more fundamental than particle physics (usually computer scientists)
Lumpability
- Measures redundancy of underlying structure details
- Traffic flow prediction example: don't need exact driver origins/destinations
- Need certain classes: commuters, occasional shoppers, long-distance traffic
- Can lump many drivers into classes - traffic flow is "highly lumpable"
- More lumpable = more causally closed
Research Applications and Examples
- Used two notions to quantify emergent properties in examples
- Cellular automata: simple algorithm updating pixel lines according to rules
- Some rules create very complex patterns
- Formalism correctly identifies emergent features in simulations
- Numbers built from pixels form new objects
- Can describe whether statements are false or true
Broader Applications
- Potential applications in artificial intelligence
- Other use cases: identifying weather systems, activist groups
- Emergence is everywhere
- Technical term "lumpability" instead of colloquial descriptions
