The Powder Toy

Search For Powder Intelligence by Windspren

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
    QuanTech     11th Sep 2016
    i'm on the neural network side...
  • Windspren
    Windspren     11th Sep 2016
    Guys, we should really get the evolutionary cell done. I'm working on a bisexual reproduction system that will require 2 bots to share genomic coding and then make a duplicate cell with the shared data.
  • QuanTech
    QuanTech     11th Sep 2016
    I think the neuron function should be like a NAND gate!
  • QuanTech
    QuanTech     11th Sep 2016
    @herp holy crap! that is some really damn strong alloy! useful!
  • herp
    herp     11th Sep 2016
    TNTPIGS bots +tmp 100.100=This. ID:2038770
  • QuanTech
    QuanTech     10th Sep 2016
    also, remember that "New Weight = Weight + Error*Input", where Error is the difference between the output and the desired output
  • QuanTech
    QuanTech     10th Sep 2016
    or reinforcement learning
  • QuanTech
    QuanTech     10th Sep 2016
    Our ultimate goal will be to get to Unsupervised Learning!
  • QuanTech
    QuanTech     10th Sep 2016
    for that, we would need efficient and small neurons
  • QuanTech
    QuanTech     10th Sep 2016
    btw, instead of programming AI on R16K1S60, we should maybe try something that better fits AI development (and we've tried it before too): Neural Networks!!!