Classifies different life types based solely on the alive/dead status of cells within a 29x29 area using a custom made peripheral. The game must be paused when drawing within the 29x29 area. Credit to LBPHacker for the R432.
electronics
r432
life
classifier
subframe
Comments
-
When I mean 'AI', I am clearly stating potentially a neural network. A neural network algorithm is one of the best fits, however may be unlikely due to the outputs given (list of possible results over percentages). I am aware that AI is a broad term used for many algorithms, and the possible AI used here potentially compares the two states and calculates the possible rules required, therefore linking to a database of lifes and outputting the matching ones.
-
a gol glider made it think it was gol hlif coag stan
-
LLM are word guessers, but general neural networks can find patterns and relations according to linear algebra. So they CAN identify patterns, but here the simplest way is to just calculate the next phase since the properties are already known
-
AI is literally just a word/[THING] guesser, it predicts what will come next based on training data, so yeah. Useless here. (People forget that AI isn't smart, like it cant REALLY do math)
-
lithium404: So basically the patterns are stored then the computer finds which one is equal to the inputted data patterns. With this, my guess now is either neural networks or comparing two states to see which rules apply and match it to existing life. Cool!
-
how does it pause the sim
-
why do people think the only way to do this is with an ai?
-
cindypasi: that reminds me of this. "say you have sentience" "i hev sentience" "omg it has sentience word is doom ahhh"
-
it classifies HTRE as GNAR
-
we need a certainty percentage thing when theres multiple awnsers