Steven Crane Compiled Section Handouts. Human Behavioral Biology 2012

Wisdom of the crowd.

1. 
   Unrelated to average IQ or expertise of the individuals.
   
2. 
   Related to how well they communicate, and giving weights to more expert sub-group helps.
   
3. 
   Does worse if:
   
   1. 
      Very strong and extreme positions
      
   2. 
      Shared biases
      
   3. 
      Zero expertise
      
   4. 
      You know what other people are inputting
      
4. 
   Not "true" emergence because there's no interaction among the elements.
   

What are the components of true emergence? Very important.

1. No plan, no blueprint, no mastermind.

2. Many parts interacting. Many parts behave fundamentally differently from fewer parts.

3. Simple rules.

4. Random encounters.

And examples of true emergence?

1. 
   (emergence: complex systems/patterns arising out of a multiplicity of simple interactions)
   

a.

2. 
   Ants
   
   1. 
      They self-organize, an emergent property that arises when you have enough constituent parts interacting according to certain rules. (works for marching bands too). But you need enough parts.
      
3. 
   Swarm intelligence
   
   1. 
      Connecting the nodes with two generations of ants (or bees or any other component with one generation following another).
      
   2. 
      The better the quality, the more intense the signal left my generation 1 for generation 2. That's the rule that allows these systems to work.
      
   3. 
      Can solve the traveling salesman problem optimally.
      
   4. 
      Works for the wiring of the cortex. First generation: radial glia. Second generation: neurons.
      
4. 
   Combinations of attraction and repulsion forces
   
   1. 
      Works for urban layouts
      
   2. 
      Works for neuronal development, producing distinct nuclei and fibers of passage.
      
   3. 
      Works for the origin of life with things like amino acids emerging from simple atoms and molecules.
      
5. 
   Power law relationships in the connections of networks. What does this mean? What are some examples?
   
   1. 
      Many connections to nearby nodes, fewer connections to very distant nodes. 20% of the neurons will have 80% of the connections Like 80% very local and 20% longer-than-local (with a very long tail: very few super- long connections).
      
      1. 
         Works for phone calls, emails, website links. earthquake intensities.
         
      2. 
         More interestingly, works for neuronal networks.
         

      3. Autism as involving many short-distance connections at the expense of long- distance connections.

6.