Like I said before an AI would not think conceptually in a way that is even comparable to how we do let alone understandable

No, nerves have a different communication system built on aggregate stimuli

Last I heard they were creating experimenting with reproductions of brains at a functional level

That was some years ago

And they also function on a much less efficient chemical circuit, as opposed to an electrical one

I'm guessing you don't have a source on that @Eccles

And also the input senses for the human brain are extremely subjective and flawed

Well, it's bioelectrical

It’s not hard to make inaccuracy or randomness in AI though

^

You can make random noise

https://www.humanbrainproject.eu/en/brain-simulation/

The brain is very efficient in the power it has for its size

Size yes

The human brain is quite decent, but the strength is in the focus on memory and expanse on processing that memory

@lunemarie yeah, but random noise isn't an accurate approximation of healthy human brain subjectivity. You'd just make your computer schizo with that

Mechanical computers also aren't built on nearly such a small scale as the brain

Lmfao true but you can have degrees of randomness

size is irrelevent

You can add risk taking etc

Size is never irrelevant <:Kappa:386676594120589312>

the important thing is if they can accurately model it

Is the brain really random or is it all deterministic based on measurable chemicals and some such

moore's law will do the rest

The first attempts at artificial brains will probably have such horrifying results that they'll be banned globally

Any reason to believe that?

I think if we were to try and make an AI based on what we know and have right now, scholarly and hardware-wise, we'd have things that can solve something we've given it information on how to solve in a very short timespan.

Moores law is in decline last time I checked

I don't think we have anything that can truly learn on its own. So far we've made programs that tell it _how_ to learn, not that it can just learn.

@Samaritan™ it's deterministic but the determinism is so minute that it's extremely difficult to emulate, so it appears nearly random
This is a big hurdle in emulating the brain, and also in understanding neurological processes

More specifically, programs that tell it _what_ to learn

Moore’s law isn’t that fucked right now

See AMD

It’s just harder to make die shrinks happen really which is what propelled it

@UnfilteredGarbage That is effectively my position on this

@Lios this. Deep learning is just efficiency-minded memorization and recall

And only on specific subjects

Well outside of what something one would call intelligence

Technological progression isn't at a consistent speed

Still kinda fast these past few decades