Conversation
Edited 13 days ago
the fact the entire brain of a fruit fly can be scanned and simulated in a computer and even taught to do things it would have never done in nature is extremely impressive, but I honestly hope they never manage to do that to a rat or anything that can be reasonably said to be sentient
5
0
2

@wire_witch I think the claims that the scan -> computer program is any accurate are greatly exagerated
fruit flies are not as simple as they seem

1
0
1
@efi even if it's not entirely accurate, it still clearly works and does things
1
0
0

@wire_witch yeah, but I can make a fly simulation in python too

1
0
1
@wire_witch
https://www.eweek.com/news/rat-neurons-real-time-computing/
It's no simulating a rat inside a computer but some scientists DID grow rat neurons and force them to do computing tasks. so in a way humans are already making dubiously alive electronic rodent homunculi, just with hardware instead of software.
1
0
1

@wire_witch fortunately the simulations are all static connectomes which means it's basically just input → output with no memory or learning

0
0
1
@AppleAmps oh yeah they've been doing this for quite a while now
1
0
1
@wire_witch the connectome is not a functional brain scan. it's more like if you opened up an electronic device drew a dot for each component and a lines between them for each connection without noting what the components and their values are. it's just a graph. vibe coders can then use that graph for art projects, but theyre not simulating a fly's brain in any meaningfull sense of these words.
1
0
1
@wire_witch (it's ofc still super cool from a neuroscience standpoint, and gives a lot of structure to wonder abt, but the simulation gimmick is meaningless)
1
0
1