/sci/ - Science & Math

>>8227054
Honestly, we only have indications. A is the closest we have.

But that's just what we think. We could be doing post hoc ergo propter hoc all the time with the brain, as it is the only thing we cannot really take apart to study the same way we can with a lot of other things. We need to see the reactions while they happen which we can't because they're in our heads.

B and C are nonsense until proven otherwise, I don't really believe in B but I believe consciousness and the soul are separate entities that don't interact at all - so I'm sort of a materialist.

>>8227091

>I believe consciousness and the soul are separate entities that don't interact at all
>soul

nigga, wat?

>>8227091

>as it is the only thing we cannot really take apart to study the same way

Lol.

>>8227125
Yeah?
>>8227125
Well, it's true. Almost all other things we can dissect and study and get all the info we need even though it's taken apart. But in the brain we need to see the specific interactions while it's still functional.

>>8227136

>What is FMRI?

>>8227137
You think FMRIs are enough to actually study the brain thoroughly?

you know that feeeling when they can resample voice... know splinter or ludwig?

>>8227108
yeah, I believe in a soul and in god

>>8227143

>confocal laser scanning microscopy
>beam scanning electron microscopy

>>8227160

Why?

Thought experiment for everyone :
If you could exactly replicate someones braincells to the atom along with their memory, would you also be replicating their consciousness ?

>>8227187

It would diverge from the original immediately.

However, it would appear to be identical to us.

>>8227180
No reasons that you would respect or consider.

>>8227054
Make D) Consciousness is natural occurrence in all systems (biological or artificial) once they pass a certain level of complexity.

>>8227243
That's already covered by A, you pleb.

>>8227243
It's definitely C. I know it's subjective but thats the only option I know for a fact that it's real.

>>8227386
C is literally solipsism, you fucktard.

>>8227247
A states in in terms of a 'feedback', whatever the fuck that actually means

There exists a D, that which is probably materialistic, but is not a feedback, and one we cannot speculate on further

>>8227390
how is that contradict to what I said fucktard

A and C, B will likely also be correct because the nukes are overdue

If you think consciousness has anything to do with magic spooky skeleton voodoo mysticism, you need to leave /sci/ right now and never come back.

t. AI researcher

>>8227898
I really dig the way you presented Data to back that up.

>>8227899

>>8227898
Watcha working on?

>>8227904
Vector spaces. Lots of vector spaces.

>>8227898
>t. AI researcher
you and every other cs undergrad looking for internet cred

>>8227924
It's okay anon, let the butthurt flow through you. It's okay to make things up and call names as long as it makes you feel better.

>>8227926
>It's okay to make things up [...] as long as it makes you feel better.
practicing what you preach i see

>>8227054
Your definition of materialism is too restrictive. For example, John Searle would fully disagree with A, yet he is a materialist.

I would instead define materialism (or physicalism) as "All mental states are contingent upon physical states. All causal mechanisms of mental states can be reduced to physics"

>>8227934
Any other baseless, angry accusations you would like to make kiddo?

Conciousness is created by the feedback loop. By our ability to think about our thoughts. Without this loop we would simply be zombies.

>>8227945
This precisely. Fucking finally somebody else get it.

This is why recurrent neural networks, with their internal state, give rise to consciousness. This is also why thought vectors represent internal state.

>>8227054
World is fundamentally ideas - consciousness although supported by neurons and their feedback seems to be a field. Or rather it seems to be suggested by nature itself that we are a simulation - possible computer one.

>>8227949
lol

>>8227961
>he doesn't read the papers coming out of Google Deepmind
get a load of this pleb

>>8227963
show me the deepmind paper that says RNNs give rise to consciousness

>>8227949
What we need is a pill that turns the person unconscious for one hour after each use - then take it when in severe pain or when close to death to not be aware of disappearing.

>>8227965
They're not the ones who discovered it. They're just using it to do stuff.

http://cercor.oxfordjournals.org/content/13/8/807.full

http://www.psych.utoronto.ca/users/reingold/courses/ai/cache/neur_net.htm

They allow networks to think about thinking, by responding to patterns in their own state.

>>8227980
these have nothing to do with what i asked for

and now i know why; you literally linked the first two results on google for "recurrent neural network consciousness"

arguing on /sci/ in a nutshell holy fuck

>>8228006
>>8228009
There were some specific papers, but I can't be assed to find them at the moment since I'm busy. :^)

>>8228012
yeah i'm sure that intro to java assignment is keeping you real occupied

god damn consciousness threads are pathetic

>>8228019
Even though Kurzweil is a crank, he gave some talks that offered some insight into what Deepmind was working on and how they were making use of recurrent neural networks to emulate conscious behavior. Maybe start by looking there.

>>8227054
Both A and C are correct, it's a property so it doesn't matter if you think it's there or not, it's defined based on it's effects. This is how properties of a system work.

>>8228024
you are one dense mother fucker

just do me a favor and stop calling yourself an AI researcher on the internet

>>8228024
I haven't seen one instance of NNs producing conscious behavior. Very good behavior in very limited situations that they are pre-configured for, yes, but not general conscious behavior. Best I have seen is by symbolic AI projects like Cog, Watson and Cyc. This statistics approach is good for problems where you know what the solution is and want to make a system that can reproduce it, but when you don't know what the solution is, or there is a possibility that there is no best solution or the solution is ambiguous, symbolic logic is what is used, which is why symbolic systems are what is actually getting somewhere in regards to knowledge bases, and statistical systems are stuck automating engineering work

>>8227909
That's pretty basic. Kalman filters?

Has the consciousness been measured as quantized bits ever? Has this been proven somehow?

>>8228088
not yet

We have no proof of 'spiritual' concepts and gods. We do have proof (as much as it can exist) of material phenomena. Therefore it is more likely to be the latter. That's all that can really be said. You can't currently prove that anything is conscious, except yourself, to yourself. So who fucking knows.

>>8227179
>studying dead, freeze-dried, metal-sprinkled brain tissue
not good enough

>>8227243
why do you think there's an arbitrary threshold?
that doesn't make any sense. According to your theory, almost anything has consciousness (a single neuron, anything that could act as a neuron etc)

>>8228088
are you people joking? HOW THE FUCK IS THAT EVEN POSSIBLE?

You need to define consciousness.

>>8227898
>implying my SQL database hasn't achieved hard consciousness
Do you even read popsci weekly bro

>172 total votes
so far so good ;)

>>8228088
how do you measure an immaterial concept ?

consciousness will never be able to get proven by science due to its nature

>>8229053
fake screenshot
>>>/pol/

If you believe in unique human perspectives you LITERALLY believe in a soul and shouldn't be on /sci/. It's amazing how many people delude themselves into thinking they follow logic, but then throw it all out when asked a question like
>>8227187

You aren't unique. You aren't special. Your life isn't the beginning of anything, and your death isn't the end of anything.

>>8229053
Surprisingly Omnipotence is in the first place with 55% on 308 votes