Scientifically speaking, how far are we from uploading a single neuron into a computer?

Scientifically speaking, how far are we from uploading a single neuron into a computer?

Other urls found in this thread:

google.com/patents/US4896053
twitter.com/NSFWRedditGif

black dick

a really long way away when all we do with the internet is watch porn?

I can already put a neuron in my USB port

It's been done before but every time they jews manage to replace the neuron with a jewish one so the project has to be abandoned.

>Scientifically speaking, how far are we from uploading a single neuron into a computer?

infinity amount of time.

your question is literally meaningless, thus the response you receive is equally meaningless.

A tripfag deserves an infinite amount of humiliation.

>Scientifically speaking, how far are we from uploading a single neuron into a computer?

A neuron is basically a transistor in computer terms, except it has far more links (dendrites) and has other abilities controled by DNA.

However a brain has approx 100 billion of these. Even if a computer was created with a billion transistors, there is likely orders of magnitude of complexity it would lack compared to the organic.

Not really

From what we understand I think the neuron is a pretty simple machine. We can already make one on the computer that approximates basic functioning.

Infact neural networks are small clusters of these doing specific tasks.

If you get enough of these are put them together do they make a consciousness?
Probably not.

I mean, if the brain is just made up of enough 'modular' parts.. then why do some parts of the brain have specific functions, the emotional center etc... Should it all just be neurons capable of doing each other's jobs?

Shouldn't you be able to format your post at least passingly interestingly.

>A tripfag deserves an infinite amount of humiliation.

that's not what your mom said last night.

Go away. You are not funny and you are not smart. Go read your inane drivel in the archives and understand why people respond to your posts the way they do. You offer nothing because you are nothing. Shut the fuck up until you actually have something worthwhile to contribute.

neurons aren't the most complicated of the cells, brainlet

>Shut the fuck up until you actually have something worthwhile to contribute.

You posting that in a thread like this, huh?

Are you actually defending this thread?

Yes. And you are defending your shitposts. At least sage you dumb nigger.

>Yes.

wow, that's sad.

you sir, have my pity.

We're 5 away

yes really. maybe not in the future, but as of right now; Yes.

what other cells give rise to the conscious perception of realty ya fuckin brainlet

And you now have my filter. Enjoy talking to yourself.

>And you now have my filter. Enjoy talking to yourself.

wow, they replace you Cucks quickly, don't they?

Don't bother answering me, I know you didn't filter me because you have to reply.

it's your job.

Nigger you're retarded

>Nigger you're retarded

So, you are seriously debating "Uploading a neuron" huh?

You just did

>Infact neural networks are small clusters of these doing specific tasks.

not to be pedantic, but computational neural networks bear little resemblance to their organic cousins nowadays. The original networks were indeed designed with a 1970's understanding of neuroscience, but progress in the machine learning field and greater understanding about the functioning of synapses in biological systems mean that the two fields have diverged quite a bit.

as says, the question is a bit meaningless. a more apt comparison is ":how close are we to replicating a neuron in a circuit" and the answer is that we already can, but there's not much advantage to doing so when our own methods of building learning systems seem to have clear-cut advantages in solving the problems we apply them to.

stem cells, BITCH

So could we possibly replicate a human mind using these man made learning systems?

Is uploading a consciousness feasible?

Would you be yourself still if you "uploaded"?

If not, how do you reconcile the fact that every cell in your body is replaced by a different one on a steady interval?

Am I just asking stupid questions that can be reduced to what is consciousness?

google.com/patents/US4896053


Fun fact, the man who authored this patent claims there is literally millions of dollars in gold bars buried in the desert in Arizona, and imaged it using 10,000 dollar military grade ground penetrating radar to image it.

Heres a pic

>So could we possibly replicate a human mind using these man made learning systems?

No. there isn't any consensus on this in any respectable community outside of trans-humanists holding their breath for their version of an afterlife, but it's pretty certain that if AI does arise from the application of these techniques, it will be it's own beast.

>Is uploading a consciousness feasible?

Reproducing your consciousness? see above. uploading your consciousness?that gets philosophical as shit. not the forum for this is the short answer.

>Would you be yourself still if you "uploaded"?

that's kind of the issue.

>If not, how do you reconcile the fact that every cell in your body is replaced by a different one on a steady interval?

1) this isn't true, especially in the brain. there's a reason people with brain damage go vegetable for the rest of their lives.

2) even if it is true, it's about the core "dogma" of DNA and how that results in the brain's structure to begin with.

>Am I just asking stupid questions that can be reduced to what is consciousness?

yeah more or less.

...

Does any of this make any sense to anyone?

yes. the circuits are based off the discovery that synapses function in a way nearly identical to adjustable integrator and differentiator circuits - essential filters - which are the bread and butter of analog design

it's essentially detailing the circuitry equivalent of a synaptic connection.

Fuck. My remaining optimism for life just got crushed hard by an oppressive existential curtain of lead.

Time to give up my goal of furthering humanity for my children and be a parasitic hedonistic narcissist.

welcome to Veeky Forums.

So life extension should be pursued from a biological and genetic perspective, not a technological angle?

you mean welcome to life as a human.

If only I was a lobster.

>From what we understand I think the neuron is a pretty simple machine
we still do not understand all the different signaling or organization processes, just some of them

Isn't it just chemical?

partially, yes. and there is plenty of active research down that road. learning to regenerate brain tissue with stem cells and deep mapping of human neural circuitry will go a long way to increasing our "useful" lifespan, and could actively augment our intelligence.

on the other hand. one of the many upsides of having remarkable results in machine learning is that machines could well augment our consciousness. in a sense, they already do, but I'm talking about a more direct connection to machine, a neural interface.

without getting to long-winded here on /scifi/ stuff I guess what I'm getting at is that the two are complementary - one isn't a substitute for the other.

technically all known biology is "just chemical"

but then you'd be delicious. being delicious is contrary to immortality.

So this regeneration of brain tissue to increase our useful lifespan. Is this field biochemistry? Genetics? What do I study to optimally impact this advancement? I'm currently a premed/chem e, do I have any applicable knowledge?

What is this consciousness augmentation you refer to? My only exposure is Elon's interviews of Neuralink and technology's impact on our biological maturation

:( humans are delicious :)

yeah, and look how long we live.

FUCK you're right

no wonder trees live so fucking long

god damn it this is Nobel prize worthy

>So this regeneration of brain tissue to increase our useful lifespan. Is this field biochemistry? Genetics? What do I study to optimally impact this advancement? I'm currently a premed/chem e, do I have any applicable knowledge?

oh, yeah, you definitely do. - biology, more precisely genomics and cellular biology, would probably be areas to focus on.

it's not even close to my area - I'm more on the "tech" side of things as an EE - but you're in the right alley. ask around your dept. if you're interested in stem cell research. I also know that brain mapping is active and ongoing at several institutions, probably plenty of internships available. .

>What is this consciousness augmentation you refer to? My only exposure is Elon's interviews of Neuralink and technology's impact on our biological maturation

it's more or less what Musk is talking about, though being musk he's memeing the shit out of it.

what we know is that very compact neural networks can learn very specific tasks with incredible speed, and be very good at them when handled properly. it's all very early to begin saying they outperform humans, which they generally don't, But the results are growing increasingly impressive.

outside of neural networks, Computers themselves have the advantage of being scalable , which brains currently aren't, are very good at high fidelity information storage and retrieval, and can run rote calculations, even incredibly high level ones, at speeds and precision that put us to shame.

so, if you could give the human consciousness the ability to tap directly into these advantages of computing - to use computational neural networks, as controlled by a human consciousness - ala a neural SCADA system, to directly impart the advantages of digital computation to humans - you could augment our abilities as a species by multiple orders of magnitude.

to dampen this a bit - next post.

Damn this is some incredibly interesting stuff.

What other mind blowing topics interest you or do you think will be revolutionary?

I swear, all the kids who go into EE are definitely a step above. They also have the passion I tend to see other engineers lack.

>the answer is that we already can
Wrong. We're not even close. If you are, feel free to collect your Nobel any time. I would, in fact, be very surprised if it happened in less than 10 years from now.

To dampen that all a bit

1) people were hot and horny over neural networks for a long time, and trans-humanists have been trumpeting the end of human supremacy for awhile, but recently, there's been some interesting emerging research coming out of computational research that suggests neural networks aren't even close. I'm not going to dig up exact information right now, but recent attempts to crowd-source high level abstract computation in the form of "games" have proven to be disgustingly effective, particularly at solving some problems in quantum mechanics. it also turns out that the human brain isn't nearly the slouch we thought it was, it's just hard to see that when people are busy being assholes and dealing with life as a whole.

computational neural networks also have some problems that are starting to emerge as a bit of a thorn in the side of substantial progress from here that isn't just scaling up of current techniques, training methods and over-training being among the most pertinent.

2) In relation to the human brain not being a slouch - there are individuals whose brains break the rules, and defy everything we think we knew. Savants who can carry through immense calculations with precision at speed, people with hyperthymesiic memories, suggest that the advantages of a digital computation may not be so clear cut in some cases.

That the above suggests is that much, much more understanding is required on all fronts in the field - biological and technological. We have barely any real understanding of the fine structure of biological neural circuitry, and new research with deeper analysis of individual nerve cells suggest we may not even fully understand those.

In a similar fashion, we have strikingly little understanding of why neural networks work so well, and though we can explain in broad terms why we see certain phenomenon, deeper progress requires deeper understanding of what computational neural networks are doing.

Do you have a journal or something where you write your epiphanies, thoughts, ideas? Would love to read one written by someone with the technical knowledge from your field. Even just a few random thoughts a day

>Wrong. We're not even close.

we certainly can reproduce a single neuron in functionality - to be precise, we can reproduce the functionality, at a single point in it's development. biological wetware comes with an advantage of reconfigurability that our hardware lacks, I'll give you that.


>I would, in fact, be very surprised if it happened in less than 10 years from now.

No-one is trying, certainly not in any physical way.

you, as many others i tell this to, seem to think I am knocking neurons, but I'm not.

the reason we don't try and do this is that the circuit for a single neuron would be rather complicated - each synapse connection, on it's own, constitutes a driven integrator and differentiatior, which a given connection on a terminal can switch between depending on history of input. in terms of a physical circuit, having both of these available, with some circuit to moderate which is active, would probably entail hundreds of components. on inter-neurons, which form the bulk of "processing" structure in the brain, there can be many thousands of such connections. Already we are reaching into the millions of components.

there are other functionalities of each neuron and synaptic connection, including inhibitory response, something that appears to be similar to local back propagation, and basic "power-management" functionalities, all of which would translate into complicated circuits of their own.

In theory, The above is fully realizable on an integrated circuit.

however, the IC required to accomplish this is much, much larger than it's organic cousin, and is much, much harder to network to other neurons. who cares if you build a "Neuron" IC? how do you propose to network it to a thousand other "Neuron" ICs and make something interesting happen? that is why there is not much interesting in pursuing that direction of research.

the above is also why the "same number of transistors as neurons!!!" argument is bullshit. fuck kurzweil.

I do, but I'm not of a mind to publish it, and it relates to some of my current research, so it will end up in public domain in some form at some point.

Veeky Forums posting will have to do for now. maybe I'll start a blog one day, but not soon, life is busy enough as is, sorry.

>we certainly can reproduce a single neuron in functionality
You can put clamps and after fucking around a bit create a functional model that statistically behaves similarly enough (with regard to spiking). Problems begin when you realise that you have pretty much no idea how to capture any long term dynamics in your model. Accurately mapping gene expression and how do neurons interact in ways other than sending signals down axons is something that will probably take decades, if you are very optimistic. OpenWorm can't do it in a fucking nematode.

>You can put clamps and after fucking around a bit create a functional model that statistically behaves similarly enough (with regard to spiking)

if we're speaking to a computational model. I'm not.

>Problems begin when you realise that you have pretty much no idea how to capture any long term dynamics in your model.

which I acknowledged above

>Accurately mapping gene expression and how do neurons interact in ways other than sending signals down axons

also acknowledged above.

you aren't arguing against me here - you're making a network vs. node point, which is fully correct. I haven't denied that at any point. doesn't change the fact that the neuron, as a singular structural unit at a single point in time, can be reproduced.

Technically all know biology is just "energy"

You're debating not being a retard and failing. Now it's your turn to keep replying to a thread that has been successful, regardless of your unfunny shitposting. Do it. ITS YOUR JOB. Remember to sage you dumb nigger.

depends

how precisely do you want to approximate the neuron? depending on that answer ranges from impossibly far to already done long ago

>upload single neuron

We done that with an entire fruit fly brain. I think we can emulate it's brain at 1/10th real time.

Forgot pic