[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math


View post   

File: 27 KB, 355x432, 686W.jpg [View same] [iqdb] [saucenao] [google]
3988433 No.3988433 [Reply] [Original]

When will they be able to capture the sentience of a person?

Would a different ethical code and priorities such as the Nazis had be necessary for transhumanism to be achieved?

>> No.3988447

Mind-uploading will only be possible when we've made enough advances in neurophysics

>> No.3988448

Descartes found it centuries ago, in the 17th century. Sentience is in the pineal gland.

>> No.3988459

According to Penrose, the human mind works in ways that cannot be imitated by a computer.

>> No.3988469

>>3988459
Sounds to me like somebody doesn't understand computers very well.

>> No.3988494
File: 265 KB, 938x1200, penrose.jpg [View same] [iqdb] [saucenao] [google]
3988494

>>3988459

>"Guys, the brain uses quantum effects for consciousness!"
>"But the brain is at room temperature and a solution, how can quantum effects be present long enough to impart any change to otherwise largely deterministic molecular dynamics?"
>"Err, there's this little section of the microtubules where quantum effects MAY prevail for longer than a trillionth of a second!"

All of these questions are more are answered with SCIENCE in the Whole Brain Emulation Roadmap, written by a PhD in computational neuroscience.

http://www.fhi.ox.ac.uk/__data/assets/pdf_file/0019/3853/2008-3.pdf

Also, ATLUM, the only microtome+microscope with enough resolution to extract a connectome out of brain slices was invented by a transhumanist. Yay!

>> No.3988500

>>3988469
Yeah he only worked on QFT what does he know of complicated object like computers.

>> No.3988510

>>3988459
This statement is obviously incorrect, since a "computer" is just any input/output manipulator. Ergo, the brain is a computer. If he meant "any current computer" he might be right. Personally I think the right configuration of Turing machines would give rise to a mind, which in essence means an equation could grow sentient.

>> No.3988525

>>3988500
>being good at QFT means you understand computers
lol, physicists just can't help themselves

>> No.3988529

You people make me sick. Science cannot explain sentience or consciousness sufficient to fabricate reproductions thereof. If you think that the mind and its products are simply the results of physical processes--congratulations, you're a religious person!

>> No.3988531

>>3988529

Who said anything about explaining consciousness? Only the low level dynamics. That is enough.

>but but but I'm not a sum of my parts!!!dfdsjfdisfjdsifjidso

>> No.3988535

>>3988510
if that is true then every choice you made was predetermined since your thoughts are just the outputs of some computation.

>> No.3988542

>>3988529
do christfags come here often?

that's like niggers going to /new/

>> No.3988548

>>3988529
>If you think that the mind and its products are simply the results of physical processes--congratulations, you're a religious person!

But that is all any observed phenomenon has ever been due to. We have no reason to believe the human brain is any different. Making up a fantastic explanation like "a soul" for an unexplained observation is what religious people do.

>> No.3988550

>>3988535

Yeah, so? Don't tell me you still believe in free will.

>> No.3988565

>>3988535
For any given input yes. But the great variety in input makes for a practically unpredictable mind. We might say what will happen during a certain circumstance, but not what will happen in the nigh infinite complexity of the real world.

>> No.3988577

>>3988565
You can't predict it (yet) but it was always determined.

>> No.3988579

>>3988510
Except that the brain isn't just neurons, but also hormones. The exact same brain will respond differently, depending on the hormone level in the blood.
The effect of the hormones allows for chaos to enter the realm of the brain. It is impossible for any entity to perfectly emulate a chaotic system (such as the brain).
Which does not mean to say that a computer cannot perfectly act like a brain in general; it just cannot simulate a particular brain perfectly.

>> No.3988585

>>3988577
except quantum mechanics says "No."

>> No.3988596

>>3988579
The hormones are not magical. They're chemistry, which is applied physics. Thier effects could be perfectly emulated, given enough computer power.

>> No.3988610

>>3988548

Are thoughts phenomena that have been observed? Please, identify the qualitative or quantitaive units of measure for thoughts. I'd love to know.

Also, I never mentioned the soul.

>> No.3988616

>>3988585
We have no reason to believe that quantum effects are significant in the machinery of the brain. It could however affect the input ("the real world").

>> No.3988620

>>3988585
Except that it does not.

>> No.3988625

>>3988610
Thoughts are bits of software, formed by the biochemical hardware that is the brain.

>> No.3988632

>>3988610
>units of measure for thoughts
bytes

>> No.3988642

>>3988596
>The hormones are not magical.
Strawman. I didn't say they are magical, I said they are chaotic.
>They're chemistry, which is applied physics.
Yup.
>Thier effects could be perfectly emulated, given enough computer power.
Nop. Chaos. Chaotic systems cannot be perfectly emulated.
Compare it to the weather. No amount of computing power can perfectly predict the weather. Even though the weather is basic physics.

>> No.3988641

>Nazis
>philosophy of subhumanism
>achieve transhumanism
what

>> No.3988649

>>3988642

Chaotic systems are deterministic, even if they vary greatly if you change the number of inputs.

Since you can't get a perfect measurement of brain chemistry, you'll have to narrow the concentrations down to real numbers. In any case, a simulation of brain chemistry would be a lot less demanding than one of the actual neurons, and sub-threshold neurotransmitters act very slowly, reducing the computational burden of trying to make a real-time model of them.

>> No.3988658

>>3988649

Also, this means that the upload is not identical to the original, in case you're going to argue that. Not that anybody cares.

>> No.3988664

>>3988642
Can one not find order in chaos?

>> No.3988665

What do you mean by "capture"? Surgically remove the neo-cortex? Conduct scientific research to try and identify a more precise cause of consciousness than "alive human brain = soul"? Sorry if I'm being dumb but I am missing something here.

>> No.3988668

>>3988642
But it isn't chaotic. The body produces hormones based on available nutrients, current hormone levels and external input.

>> No.3988671

>>3988649
Either, you simulate a brain by not being aware of every single hormone floating around, but using smart approximations; in that case, the actual brain will diverge from it's simulation. Or you have to admit that you cannot know the exact location of every hormone, but in that case, you have a decision procedure outside of the actual model, making the brain a non-deterministic turing machine.

Tl;dr a turing machine modelling a specific mind is either wrong or non-deterministic.

>> No.3988674

>>3988641
They were the at the cutting edge of science and would have no qualms about treating people worse than lab rats if it got results (as all ruthless unshackled study does)

>> No.3988683

>>3988642
What the?
are you high or do you know shit all about computation?

>> No.3988686

>>3988671

>the actual brain will diverge from it's simulation

This will happen to every and any simulation on a level of abstraction any greater than reality.

>> No.3988697

>>3988668
>>3988664
The word chaos doesn't mean what you think it means:
http://en.wikipedia.org/wiki/Chaos_theory

>> No.3988700

>>3988642
The weather models aren't perfect, because we haven't got all the data. And since it's practically impossible to know the positions and velocities of every involved particle, the models have to be based on statistical physics. Truly unpredictable systems exist only at the quantum level (or macroscopic systems tied to them).

>> No.3988717

>>3988686
Yes, so?
I was just stating a fact that countered the notion that 'a sufficiently strong deterministic turing machine can emulate a particular brain'.

>> No.3988728

>>3988700
Seriously?
Look, if you won't even read the link about chaos theory, what's the point.
No amount of precision for the input will yield a perfect model of the weather. It's not just too imprecise, it will always be too imprecise.

>> No.3988741

>>3988728
But that isn't to say that it cannot be determined. Nobody can objectively no anything, however that doesn't mean the objective truth is not out there.

>> No.3988745

>>3988642

>Chaotic systems cannot be perfectly emulated.
>Compare it to the weather. No amount of computing power can perfectly predict the weather.

Both these statements are incorrect!

>> No.3988748

>>3988741
know*

And by objective truth I mean objective perspective.

>> No.3988762

>>3988686
No. Only to chaotic systems.
An oven, with a bimetallic switch, is an example of a predictable system. It doesn't even matter if the current temperature is 63 or 64 degrees. It will heat up to about 220 degrees, then cool down to about 218, and keep between 220 and 218, until someone switches the oven of.
A brain consisting of only neurons is another example of such a system. Like a computer, or a turing machine, a neuron is digital, it either fires or it doesn't. Now the problem is, that when hormones are involved, they can surpress such a firing event (or activate). Hormones in the blood, however, are chaotic. They do not converge to the model.

>> No.3988772

>>3988762

Sorry, I meant simulations of the brain, or, well, chaotic systems.

>Like a computer, or a turing machine, a neuron is digital, it either fires or it doesn't.

Um, there's a bit more detail than that, but okay.

>Hormones in the blood, however, are chaotic. They do not converge to the model.

Fluid dynamics is pretty accurate. You can model the distribution of hormones in the blood. It won't be an accurate simulation. How is this news?

>> No.3988782

>>3988741
Fucking pay attention here.
I can predict the outcome of a computer, if I 'know' the starting configuration. So what does it mean to 'know' the starting configuration. For a computer, this is simply the 1's and 0's. We don't have to care about the exact charge present in particular transistors.
Hence, the computer is deterministic. Given a sufficiently strong, but finite, amount of information, we can perfectly predict the outcome.
In a chaotic system, however, a slight displacement can lead to a completely different configuration. Say, there is a professional pool player. The outcome of a single poolgame isn't determined, however. A slight misconfiguration of the first shot, leads to a ball being in a slightly different position. If you hit the ball with the white ball, the displacement of the white ball will be more then the displacement of the ball you just hit. Every single shot, the difference from the previous scenario will increase. There's nothing you can do about it.

>> No.3988792

>>3988782

>In a chaotic system, however, a slight displacement can lead to a completely different configuration. Say, there is a professional pool player. The outcome of a single poolgame isn't determined, however. A slight misconfiguration of the first shot, leads to a ball being in a slightly different position. If you hit the ball with the white ball, the displacement of the white ball will be more then the displacement of the ball you just hit. Every single shot, the difference from the previous scenario will increase. There's nothing you can do about it.

If you know all of the input data, you can predict it with perfect accuracy because it's still deterministic, only the initial conditions change, but this is intractable and unnecessary and probably impossible to achieve. A simplified model -- What's so wrong with that?

>> No.3988802

>>3988772
Chaos, you dense mother fucker.
In a simulation of a computer, there is no approximate error, that propagates, and becomes larger. The whole point of the design of a computer, is to diminish errors. The charge that runs through the transistors isn't always exactly the same. If you pretend they are always the same, you introduce an error. The error, however, doesn't propagate itself throughout the whole system.
A small error in your fluid dynamics will (if you wait long enough) make sure a neuron fires in the model, but not in reality (or vice versa). This neuron firing will cause others to fire, it will influence the hormone level, etc. Every prediction will eventually be wrong.

>> No.3988820

>>3988792
>What's so wrong with that?
Nothing. Where did you get the idea that I don't believe that one can make a very sofisticated model of a generic brain?
My point was, and is, that the brain is fundamentally different from a (deterministic) turing machine.

>> No.3988841

>>3988802
>>3988820

The simulation will surely diverge from the original (IF the original is still alive after the scanning), but the neural structure will be preserved, and neural structure determines a lot of high-level things, it's not just the result of the interactions between neurons.

If a person drowns or freezes and is brought back, after neural activity is cut off, they'll in some cases return to normality as long as their neural structure was not damaged, for example.

Moreover, there's little hope for non-destructive scanning. So you can't really compare an upload with the original, because the original is going to be a descerebrated corpse on a hospital bed next to an electron microscope.

>> No.3988863

>>3988841
Well, I didn't have person uploading in mind per se. A good simulation of you would probably believe he's really you, and in all respects be you. But even if the input is identical, the simulation and the real you will become different persona, even though both are really you in all respects.

>> No.3988870

>>3988863

Pretty much.