![latest[2].png](https://i.imgur.com/LmYllb1.jpg "latest[2].png")
Tell me something fascinating.
My fursona is a pubic lice
>>53730045
I'm eating cookies at 6 AM.
I'm about to go to bed.
My driver's license was just confiscated for >1 year because I was diagnosed with epilepsy.
chinese replacement screens don't work too well on slightly bent phones
the people i work with are flaming fucking retards
the secret big mac sauce is actually no secret
I'm eating cookies at 5 AM.
I just woke up.
The 1s go through ethernet and fiber cables sideways
>>53730083
I'm eating cheese ravioli at 6 AM.
I'm about to go to bed.
pic related.
>>53730169
Welp. Goodnight fellow night-owl.
>>53730104
Shit I always wondered if people prone to seizures could drive. So you got it back?
We are less free today in a literal sense than Roman citizens were.
>>53730210
Damn anon. Don't die in your sleep.
IBM chose the inferior X86 architecture because they didn't want to be viewed as a follower.
>>53730344
At the very least IBM hasn't lost all its roots just yet.
>>53730083
Literally same here, almost 7 now.
>Chocolate chips ftw.
>>53730179
and yet here we are on 4chan.org/g/
Recurrent neural networks are Turing complete.
>>53730045
>Scan a brain with an electron microscope, program algorithms that maps all neurons in the brain, and use the digital map as one single node in a gigantic simulated neural Network. Then you make copies of the brain/nettwork and run it in a evolution program, where small mutations is programmed to occur randomly for each generation. And for each generation they compete and are tested in intelligence. The system is optimized by natural selection. With enough computing power, artificial Intelligence would arise overnight. Why don't the scientific community come together to do this?
The A.I would be able to tell us data on optimization of resource use and estimates on absolutely everything. Nigger.
>>53731371
where?
>>53731381
>scan a brain with an electron microscope
can't be done in a useful way.
>gigantic simulated neural network
we don't have the necessary computing power in all likelihood, especially for the first several iterations where we have to simulate EVERYTHING because we don't know what's important and what's not
>run it in an evolution program...intelligence would arise overnight
as someone who actually has run some of these evolution programs... no. again, we don't have the computational power.
>why don't the scientific community come together to do this?
Because as has been shown, there are significant, non-trivial barriers to the experiment every step of the way.
This is what the real process would look like:
1. Scan a LIVING central nervous system simultaneously at all points in order to capture a state, somehow (we don't have a scanner that can do this right now and no technologies are on the horizon which would allow it)
2. Simulate that state in a system which can also simulate the action of further stimuli and that CNS's stateful response to the stimuli. (This is not only technically difficult because we may not have the processing power to do it unless we can somehow efficiently pare down the scanned data prior to attempting to simulate it, it also is conceptually difficult because we would have to figure out how to simulate appropriate stimuli if we wanted to get useful results out of the simulation.)
3. Evolve it using a GA. (This is conceptually difficult because we still don't actually have a good idea of how intelligence - especially general intelligence - works in humans, so we wouldn't have a decent way of imposing the right selective pressures).
4. Ask the resulting evolved intelligence the right questions. (Again, this is conceptually difficult because unless dumb vanilla humans can ask the AI the right questions, we are not likely to get useful responses from the system).
