Search For Powder Intelligence by Windspren

Windspren

30 / 4

2nd Jul 2016

29th Sep 2016

The reborn SFPI! Our goal is to develop and identify forms of pseudointelligence and life within TPT in the forms of PSTN bots, computing systems, genome emulators, brain architecture simulations, and anything else we (or TPT AIs we find :D) can think of

sfpi tpte3

Comments

• 

  QuanTech 3rd Jul 2016
  such as weights in perceptrons
• 

  QuanTech 3rd Jul 2016
  we don't just want complete randomness, we sorta need some control over the processing
• 

  QuanTech 3rd Jul 2016
  It won't be easy to formulate an algorithm for the behaviour of TRON!
• 

  wipyman 3rd Jul 2016
  @Kevino, you could have merc randomisers at either end of the neuron, as the neurotransmitter is detected it generates a charge that passes into the marc randomiser. You could even have the generation of neurotransmitters randomised. All of this would add to increase the random generation within the collective system.
• 

  Windspren 3rd Jul 2016
  @Quantech, right, but I need to decide on a medium for the genome first. @Wipyman that's a good idea, but it might take up too much space. MERC randomizers might work though.
• 

  QuanTech 3rd Jul 2016
  +Kevino36 That was exactly what i was saying! +wipyman That is a good idea!
• 

  Windspren 3rd Jul 2016
  It's probably because there's less room for the TRON to move in, therefore giving it a lesser opportunity to be random.
• 

  wipyman 3rd Jul 2016
  For the neurons, you could use gases between neurons to act as neurotransmitters.
• 

  QuanTech 3rd Jul 2016
  oh, and on the topic of natural selection, you could select a "fitness" for each individual genome by comparing it to a genome of your choice. (you've probably read on genetic algorithms, right?)
• 

  QuanTech 3rd Jul 2016
  i think that is because of the small space it is confined in. the algorithm tells the TRON to stay in the same path for a longer time.