If the entire fly brain was imaged and we know exactly how it works now why the fuck can we not just build a computer model that works just like it?

>>16730817 (OP)
We can. Maybe not a fly specifically, but there have been experiments replicating a worm's brain digitally and even making a real biological worm use a computer as its brain

>>16730817 (OP)

The problem is not really the wiring diagram, that is straightforward ... it is all these tiny black boxes ...

>>16730817 (OP)
Computers don't have the massively parallel capability of the biological substrate, so there is a much greater bottleneck with current computing standards that prevents trillions of simultaneous calculations on interdependent variables that are not mutually exclusive.

>>16730817 (OP)
>I have no proboscis and I must scream
What's the use case of a fly computer? Browse the Internet and get caught on the web?

>>16730878
The internet is increasingly filled with poo. Seems obvious that flies would go where the shit is located.

>>16730817 (OP)
"was imaged" =/= "we know exactly how it work" =/= "a computer can even work like that in principle" =/= "we can build it".

>>16730817 (OP)
i doubt we know how it fully works, but say that we did then the thing that is stopping us from making a computer fly is resources.

its cool that we could make a computer fly but why would we? who is going to pay for it? it sounds like an expensive endeavor with no obvious profit waiting. in this age of scientific budget cuts, how can you justify the money? all the money is going to go to cancer research or AI research. what use is a fly?

>>16730887
>what use is a fly?
To extract the visual processing system and use as a biological GPU.

>>16730892
call up nvidia then

>>16730894
Big Tech hates true innovation.

>>16730896
idk. i see companies like google, microsoft, and ibm doing fundamental research. maybe nvidia is a bad choice, but other big tech companies definitely invest in it.

>>16730817 (OP)
>If the entire fly brain was imaged
Yes, for some definition of entire.
>we know exactly how it works now
No.

>>16730817 (OP)
Because a mind isnt its physical properties. The physical property is the construct of the mind.

>>16730899
>i see companies like google, microsoft, and ibm doing fundamental research.
I don't know what they're doing behind the scenes but they're still marketing gay matrix multiplication "AI" as their vision for the future.

>>16730914

Basic upscaling. Just a question of how much stuff you intend to hook into it.

>>16730901
Best post.

>>16730925
>just upscale it bro

>>16730942
He thinks that just because his butthole scales with the number of corporate cocks inserted into it so do "AI" capabilities.

>>16730817 (OP)
>we
spotted the gommie
forget flies, they already have human brain models but it aint for the public.

>>16731097
who is they

>>16731099
not (You)
kys

>>16730878
kek

>>16730901
Just train an AI model on a dataset of a series of electrical captures of the fly brain (assuming the electricity can be imaged in that resolution) to predict the next electrical state of the fly's brain.
We don't need to know how it works.

>>16731451
>(assuming the electricity can be imaged in that resolution)
The fact that you think such a scan can include The Electricity is hilarious. You're an idiot.

>>16731473
neuralink hires ML people for their modeling

>>16731658
So FUCKING what, nigger?

>>16731659
just upscale it

>>16730878
lol

>>16730817 (OP)
I remember reading the press release it seemed kind of meh, but i'm not a neuroscientist. They "used AI" to assemble a connectome from images of thousands of slices. The neuron firing model is pretty simplistic though. They did manage to invoke simple motor behaviors in the model.

>>16732446
advances in neuroscience tend to be made in small steps

>>16732504
sure, but OP is thinking that story is something bigger. As did I, before I read the press release.

>>16732504
I was just wondering the same things as OP when that news story was circulating, until I read the press release