(untitled)

Is the human brain a computer? I have always taken for granted that it obviously is, but now I encountered people who seriously argued it's somehow different, but I wasn't convinced. My argument is that I can calculate in my head everything the computer can, therefore they are essentially the same (if it walks like a duck...). What do you think?

I guess it depends on what your definition of "computer" is. If it is just "any apparatus able to perform calculations", then sure, the brain is one. But is this any useful, does it provide any insight ? Beside some obvious though important technical points (brain is electro-chemical rather than purely electronic, brain is much more space and energy efficient, brain does not work in a binary mode (I think ?) ...), the one big difference in my mind is : Computers are manufactured by men, and we (at least some of us) understand how they work, and we can get them to do whatever we want, within their physical limits. Brains are not so. Our understanding of the brain is still very limited, and when it dysfunctions, we're left with very, very primitive ways of trying to fix it. Simply removing the non-functioning part, or trying to inject some chemical in the hope that it will fix the issue because it worked once in a somewhat similar case, with no idea as to why of how. Maybe one day our understanding of the brain will be such that it won't be meaningfully different from any old electro-mechanic computer ; but it is not the case today. t. sleep-deprived ESL, I hope my English is still readable

And now that I think about it, "performing calculations" isn't even what the brain is really made for. Where the computer handles numbers at its lowest level of operation, the human brain only does so at ridiculously high levels of abstractions, rendering the whole process of "performing calculations" so inefficient that we're much worse at it than computers which have millions time less working parts (transistors vs neurones) than us. So, just as the ability of a PC to emulate a NES does not make the PC a NES, the ability of the brain to emulate a computer does not make it a computer. Symmetrically, if some day we manage to simulate a brain on a computer, that will not in my mind make the computer a brain. (And then remains the question, is this distinction still useful at this point.) (Language only approximates reality, we have to chose which rounding errors are acceptable in this process, if this makes any sense to you.)

>>4270 it's a semantic argument. a brain is self evidently not a von neumann machine. for one human memory is not a continous addressable region, there aren't equivalents to processor registers or program counter. it would be perhaps easier to find similarities between the two, than list all the differences. so you could say maybe it's a different kind of computer, like a connection machine. but minsky who came up with the idea of perceptrons that lay at the foundation of the neural network approach to ai, and in turn influenced the connection machine design, soon after inventing the concept argued against its similarity to brain neurons. at this point for example we know that instead of there being a single signal path between neurons, there's actually a dozen of biomechanically distinct ways in which neurons communicate with each other: there's the 'known' electric signal, but then there is also various chemical feedback mechanisms, with dopamine and serotonin being the most well known neurotransmitters. the mutual crossover between these varied mechanisms of interaction is not understood, therefore it's not clear (and is unlikely) that the system is discrete. think of something like an audio effects pedal where the signal travels over the wire, but then also bleeds across the wires. now imagine that the crosswire bleed _is_ where the bulk of work happens. this is distinctly not how _any_ of the current computer designs work. the example that you bring up is notable, because the mechanism of action of a von neumann machine adding two numbers is entirely different from human brain doing the same. could we possibly build a computer that would be like human brain? that's for futurists to decide, but _none_ of the existing computers is anything like the brain purely from the mechanism of action perspective.

>>4272 PC and NES are both computers, at least. If brain can compute but isn't a computer, then what is it - a calculator? In my opinion calculators are computers too, just with a fixed program. Brains have a fixed program as well (other people cannot re-program you), but the person itself can run all algorithms in his head.

>>4273 Right, brain has obviously different architecture than von neumann, but there are other architectures too and they are also computers. Brain has this neural network architecture, unique to wetware, but it doesn't make it any less of a computer.

>>4275 i don't follow the whole argument, but brain is definitely 'reprogrammable'. we know because that's the foundation of nature vs nurture debate. intelligence is a social mechanism, and elsewhere on this board people are struggling with the kind of 'programming' that their parents and the society put in them. further examples are cults, children raised by wolves, meditation and psychoanalysis. an interesting one is examples of brain working around areas of brain damage, using other parts of the brain to cary the functions that used to be performed by damaged regions.

>>4275 >If brain can compute but isn't a computer, then what is it Brains are brains. We have never manufactured anything like that, not even approaching, so don't expect to find a good metaphor / model for the brain among our creations. >Brains have a fixed program as well I don't see what you mean by this. I think you're being figurative with the word "program", but I don't think it's a good image anyway, as far as I can see the concept does not map very well on the brain.

>>4276 >it doesn't make it any less of a computer. If you're willing to extend the definition like this, doesn't the word become so broad that it loses its meaning and usefulness ? At this point, an anthill is a computer too, no ? What are you going to do with this ? What was the deeper question in your mind when you posted the OP ?

>>4276 the minsky point that you missed and the whole paragraph that follows is about how brain is not a neural network. both the inventor of neural networks argued that and our current knowledge of brain mechanisms. as far as "other architecture", that's also the point that i'm making: show me an example of an architecture that's anything at all like human brain. it's not von neumann, it's not "neural networks". but again it's a semantic argument.

>>4278 I thought it in this way: (General purpose) computers are those which the user can program, namely input some algorithm and the computer will run it. Pocket calculators cannot be easily programmed although they can compute a set of pre-defined algorithms. One could in theory re-program them by wiring them differently or replacing logic gates. Brains are essentially calculators also, since there is no interface to reprogram them easily, yet they are running some HumanAlgorithm v. 2017 that makes them tick. I'd still call all these as computers.

>>4280 Alright, we don't know the exact architecture of brains, but that doesn't exclude them from being a computer. I mean, most people can instinctively know when something is a computer, even though it's all cyrillic and works on some mysterious russian ternary logic. According to "chinese room" argument, if we teach computer enough chinese, people start confusing computers into brains (because they don't see the physical manifestation of the intelligence). It's an interesting topic.

>>4281 And who is the user when it comes to brain ? Not us human since we understand jacksoykaf about it. As I said earlier, the day when we actually understand the brain and can control it in a deliberate manner, I'm fine with calling it a wetware computer. Until that day, it seems to me at best a thought experience, at worst a recipe to misguided actions, to treat the brain as a computer. Like advanced technology and magic you know ? Brains are still magic to us, not technology. Even if we know, 'cause we're meta and all, that the brain is in the end just advanced technology, we can't act on that knowledge, because we just don't master the technology yet. We can only treat the brain like magic, not like a computer. Following rules that we discern, but for reasons that are unknown. Magic. Basically I'm not really disagreeing with you, I just think you're too hasty, drawing conclusions before we get a full picture of the brain's functioning.

>>4283 I get your point, I just don't like treating anything as magic, I rather find some explanation, no matter how non-plausible. Not sure if this a fake, but they claim here's a rat's brains driving a small robot: https://www.youtube.com/watch?v=1QPiF4-iu6g

>>4284 Yeah, I saw this video as webm many times. From what I read (third hand account, so who knows), it's not really the brain of the rat, more like a solution of stuff and some 1000s of neurons on an electronic chip, somehow reacting to the stimulation of the camera and acting upon the wheels controls. I'm sure there are scientific publications to read somewhere if you're interested. (And ready to pay.) As to magic, by all mean, let's find explanations, advance the science and all, just be careful how you act on partial knowledge, especially when we're dealing with the core of our being. We've had enough issues with pesticides and cancer, plastic and infertility. Don't want to go crazy because I got a chip in my brain that was poorly designed.

The fundamental unit of the nervous system is the neuron, an extremely specialized cell that transmits an output signal along the path of it's axon if the input signals it receives are strong enough. Signals are encoded biochemically in the form of neurotransmitters such as acetylcholine or dopamine. They are released by neurons and will act on cells which have receptors for them, not necessarily other neurons. When a neurotransmitter chemically interacts with a receptor, it provokes a change in the cell's internal state. They have two general types: inhibitory or excitatory. Inhibitory neurotransmitters change the cell's state so that it requires more excitatory input to fire. Excitatory neurotransmitters change the cell's state so that it requires less excitatory input to fire. If the neuron is excited past some threshold, it will output a bioelectrical signal along it's axon, causing neurotransmitters to be sprayed near the dendrites of other cells, where receptors can be found. A complex graph of neurons receiving and emitting signals to each other forms a nervous system. Neurons are much more dynamic and complex when compared to semiconductors. The behavior of any single neuron may vary over time as it adapts to the surrounding environment. For example, overexcited neurons may reduce the number of receptors exposed on their membranes, thereby reducing the excitation provoked by a given concentration of neurotransmitters; this mechanism is implicated in the pathophysiology of drug withdrawal symptoms. The brain is a complex organ composed of several substructures each adapted for particular purposes. Like a computer, it does have inputs and outputs, referred to as afferent and efferent nerves respectively. Inputs include the senses but are not limited to them. We are not consciously aware of all the information our brain receives. For example, the carotid sinus contains baroreceptors which measure blood pressure and give the brain the information it needs to control it by modulating heart and blood vessel activity as part of homeostasis. This information does not come in the form of bits, it comes in the form of the excitatory level and firing pattern of the afferent nerve. Outputs include innervation of skeletal muscles, allowing us to move. The brain also controls functions like intestinal peristalsis, heart rate, glandular secretions. The architecture of the brain does not seem to match any man-made computer architecture I know. In particular, it exhibits plasticity; the hardware can change and rewire itself during it's lifetime. For example, visual information. Neurological processing begins in the retina, which contains neurons capable of performing functions like detecting edges. Once preprocessed by the retina, it is transmitted trough the optic nerve to the occipital lobe which integrates with the rest of the brain

>>4286 Everyone can see the differences between hard- and wetware, but we are classifying objects according to their behavior (here: their ability to compute), not how they are actually implemented in physics. At least to my knowledge, something called a computer doesn't need to be necessarily human-made of specific materials.

So OP, what I'm still wondering is, once you've established or managed to convince us that the human brain is indeed a computer, what's the next step ? Is this just a theoretical exercise on the definition of a computer and the nature of the brain, or are you trying to make a further point ?

>>4287 >their ability to compute I don't think brains compute anything directly. At least not in the traditional, exact sense. For example, brains don't seem to have a structure that functions as an arithmetic logic unit. Given enough biotechnological sophistication, I suppose you could build one. My understanding is that parts of the brain, like the prefrontal cortex, enable cognitive functions which support conscious computation, planning and decision making.

>>4272 >"performing calculations" isn't even what the brain is really made for. True. Nervous systems are for enhancing the fitness of biological organisms. Living beings inhabit an environment and attempt to survive by maintaining nominal biological function. The hostility of the environment pressures organisms to adapt. Early species adapted by developing contractility, irritability and conductivity. By contracting, the organism reduces it's own size and becomes less susceptible to damage. However, contracting also impairs other biological functions which are hindered by the smaller area occupied, like absorption of nutrients. So knowing when to contract is useful to the organism. Irritability is provided by cells which activate the contractile cells when disturbed; if something touches the organism, it will be detected by this cell which will cause it to contract. This allows the organism to defend itself when disturbed. Conductivity is provided by links between irritable cells, allowing them to irritate each other and produce a coordinated response. When one cell is irritated, it will react and cause all the other cells to react as well. So instead of just the location that was touched contracting, the whole organism will contract. It's easy to understand this small nervous system and the simple "computation" it performs. It maps a direct stimulus (irritation of the cells) to a defense mechanism (whole organism contraction). This system might vary over time. It may change it's sensitiveness to input, requiring weaker or stronger disturbances to trigger. It may change the strength of it's output, causing the organism to clam up shut or just curl up a little. However, the nature of the computation is fixed; unless new elements arrive and the structure of the cells change, they can't compute new things. They are stuck providing the same function. Over time, more complex life began to show up. The number of these cells began to increase. More complex networks began to appear. New senses, new inputs, new structural elements, new functions. Encephalization is the process through which the brain was formed: by transfer of these local nervous functions to a central location in the organism. If man-made computers operate on bits of information: given a signal representing bits as input, produces a signal representing bits as output... Then nervous systems operate on cells, changing their biochemical state: other neurons fire, muscles contract, glands secrete, etc.

The Turing machine was created to specify work that could be carried out by man without any need for intelligence. While modern machines don't fit the strict definition in many ways, they're still close enough for it to be useful. The human brain is explicitly not a Turing machine and neural networks have always been an abstraction over how the human brain was thought to perhaps behave. Alan Turing experimented with them and early neural networks in particular were typically quite simple, with each neuron representing a very abstract notion such as ``Don't attack the crab.'' or pain. A key aspect of the human brain is a lack of instantaneous global updates, use of seemingly continuous rather than discrete units, resilience to damage, and the ability to operate at varying levels of proficiency based on current feedback, along other qualities we're all familiar with. This last paragraph may seem like I'm arguing as if the brain is a computer, but I don't believe it is. If formal systems are insufficient for revealing all truth and so also inherently limited, that's enough reason to believe the brain isn't a computer nor a formal system until learned otherwise. I won't accept being restricted by some formalism, certainly before the jury's even out on it. Considering how, apparently, physical operations involving atoms at a high enough level of precision cease to be predictable and observable, I believe that could possibly be a key aspect of how the human mind operates. It would be nice and poetic if the functioning of brains were intrinsically tied to physics that would prevent a larger brain (say, building size) from being developed. Of course, this is all simply a passing thought I'd use to argue against the point, if needed; it's not precious to me at all and it's mostly something that would perhaps be difficult to argue against. It's interesting to consider that physically random events could be responsible for human ingenuity and whatnot, but that the brain has enough control over this to remain reasonably consistent. I've at times considered that thought may be similar to tree traversal after all, but this feels more as if merely one mode of thought rather than the whole picture. It seems less meaningful to stretch the idea of tree traversal to apply to anything and considering that these trees would effectively be infinite, that also muddies things. I'm less interested in understanding how my mind works and far more interested in how to improve it; I find IA to be much better than AI as something to pursue. Even if an AI that is indistinguishable from a human is created, it's very questionable if it would actually be conscious or merely an opaque imitation of consciousness. So, here's something to consider: an AI is typically thought of as being fed large swathes of data to learn from; is something that never asks a question of another or even considers it conscious? As a human, it's so easy to get in a rut or not realize thoughts until having a discussion with another human.

>>4291 >Even if an AI that is indistinguishable from a human is created, it's very questionable if it would actually be conscious or merely an opaque imitation of consciousness. If it walks and talks like a human, it's a human. The question about consciousness is silly, I certainly don't have one, nor free will for that matter, I am a mere passer-by observing my AI doing decisions. If you think otherwise, you are living in an illusion.

>>4290 >If man-made computers operate on bits of information: given a signal representing bits as input, produces a signal representing bits as output... Then nervous systems operate on cells, changing their biochemical state: other neurons fire, muscles contract, glands secrete, etc. You make a mistake of confusing the software outputs to hardware outputs. The regular computer uses mainly display for its output, the display is just a grid of pixels each encoded by 24-bits or so. But just as well it could use hands and feet and sound for the output, to perform human-like things. Actually my Windows tries to do something like that: it has software called 'Cortana' which triest to portray the computer as a speaking human. So, if we put these provides into mechanical humanoid bodies, your argument falls apart, as both hard and wet computers would operate similarly, by directly manipulating their environment. At that point the question about consciousness becomes obsolete, as there is no way to test whether the machine has it, or even if you or I have it. Consciousness is not a scientific concept.

>>4289 >brains don't seem to have a structure that functions as an arithmetic logic unit I can prove you otherwise: 73 + 22 = 95, a result directly from my wetware. So certainly there is something, if not unit or whatever you want to call it, but ability to make arithmetic and logical computations.

>>4270 Short answer. I don't think a brain is a computer. Now I can elaborate on that. You give this argument >My argument is that I can calculate in my head everything the computer can, therefore they are essentially the same (if it walks like a duck...). Well that only showing implication in one way. What about the other way? Can a computer do anything that a brain can? This is something similar that >>4272 stated. If a PC can emulate a NES it doesn't make a PC a NES. I will give two other examples. The first I remember seeing a youtube video where they made logical gates out of dominos (I think i could been posted on Lainchan in the past). Using those logic gates thay could make simple computations. Does this means that dominos and a computer are the same thing? Or even the brain is just a set of dominos? For the second example think about a wagon and a car. They can do the same thing, transport things. One needs lets say horses to do that but those are some minor differences (like in the way they work, like a computer will compute a equation in a different way then a human). Would you call a wagon and a car the same only because they do the same thing? >>4275 >Brains have a fixed program Well I think that >>4277 has described it pretty good. The brain is reprogrammable. >>4281 >Brains are essentially calculators also, since there is no interface to reprogram them easily, yet they are running some HumanAlgorithm v. 2017 that makes them tick. I'm not sure you can say that a brain is running some HumanAlgorithm. A brain is more of a self organizing structure. You can't separete the hardware and the software like in the computer (at least according to current state of our knowlage). >>4282 >According to "chinese room" argument, if we teach computer enough chinese, people start confusing computers into brains (because they don't see the physical manifestation of the intelligence). Again tricking someone in thinking that a computer is a sentient being and really baing sentient is two different things. Its like counterfeiting, if someone things coin is real it doesn't make it real money you can use as currency. >>4294 I can prove you otherwise: 73 + 22 = 95, a result directly from my wetware. Again if a set of dominos would also produce this result would you call it a brain? Also computers work in a different way then a brain (something others have pointed out). Even if sometimes a brain and a computer can give the same result it dosn't mean there the same. Of course you can call a brain a computer but keep in mind that "brain" and "computer" are only labels used to help us communicate. If you going to missuse them then that will only confuse people, but it will not make those to things the same.

We are only using barely 5% of the human brain's potential. Every sensory input we recieve is being stored in our brain, every human has 'photographic memory' but no one is able to conciously access it. Under hynosis it is proven that every moment in our lives can be accessed in perfect detail. So if you were to flash all the books in the library of congress in front of any human's eyes, that person would have all the information stored in every book in their brain.

>>4300 >citation needed The 5% is well known bullsoykaf. How would you know that the person under hypnosis has perfect recollection ? Ask them about increasingly small details ? How do you know that what they tell you are not made up details ? All this is bullsoykaf until you provide ample evidences.

>>4301 hypnosis is used by police to get witnesses to recall details they cant consciously remember. There is ample evidence the human brain has perfect photographic memory.

>>4302 >remembering some extra details from a few days ago >perfect photographic memory That's quite the leap there Boris. But I doubt not that you have several peer-reviewed articles backing you up that you will provide to convince me.

>>4303 I'll see what I can find

>>4300 I very much doubt this and also question how it's relevant to the topic.

>>4300 > We are only using barely 5% of the human brain's potential This is flat wrong. There was a misinterpretation of the science some decades ago which made it into the media and the new agers ran with it because it fit with their idea that mankind was about to transform into something greater and harness our "true potential". That, or some enlightened humans were already doing so. It's just not true.

>>4305 >and also question how it's relevant to the topic. the thread topic is: >Is the human brain a computer? the problem with AI threads and AI hype in general is people dont take a step back and define what it means to *think*. People are considered 'smart' just by retaining a lot of book memory, which can be compared to a computer simply storing data in its memory. Then there is the kind of *thinking* that involves observation and making decisions on these observations. This kind of thinking cant be replicated in a simple computer that does nothing but store data and run specific programs on that data. This requires a huge understanding of environment that can only be achieved by having a huge store of sensory input and the ability to isolate the important ideas out of that sensory data in order to create context on which to make decisions.

>>4294 >I can prove you otherwise: 73 + 22 = 95, a result directly from my wetware. People who didn't learn basic mathematics will not be able to perform that computation even though their brains are functioning perfectly. Therefore it is not the structures of the brain that directly compute the result. A part of processors, the ALU, will directly operate on binary representations of those numbers; there is no equivalent brain structure. A combination of brain structures must be supporting the cognitive process of computing numbers, a process which was taught in schools in basic mathematics classes. What is the neurosynaptic "program" that gets imprinted on the brain after it learns mathematics? I wish I knew. >>4293 I can agree with this. If you have a computer program that models a brain and it's structures, you can emulate human cognition. If you learn how, you can emulate a computer in your head. It's very a blurry distinction. I'm just trying to understand the differences between man-made electronic computers and biological nervous systems. I could think of some fundamental differences: >electronics >electric signals = electric pins >electric state = circuit state >nervous system >biochemical signals = neurotransmitters >biochemical state = concentrations of ions, gene expression, lots more

>>4270 The brain is most certainly a type of computer. Complex, difficult to study, and quite different to any silicon system, but a computer nonetheless. A computer by definition is "an electronic device for storing and processing data, typically in binary form, according to instructions given to it in a variable program." By this present definition, the brain and its distinct neuronal states and firing patterns, protein structures and plastic responses is a computer. An atypical computer, yes, but still a computer. It takes inputs, performs operations on them, and yields outputs, nothing more. While it is impressive that a few pounds of fatty tissue can do all that the brain can, it is becoming clear that the brain has substantial limitations due to its composition and substrate. Time will tell if the brain is smart enough to build itself anew; to reinvent itself.

It seems reasonable to think that a brain could be simulated wholly by a turing machine with sufficient resources and physical knowledge, assuming that any physical system can be similarly perfectly simulated. And a brain can certainly simulate a universal turing machine. So, formally, the brain could probably be considered a computer by turing-reduceability. This is corroborated by that fact that humans seem to be able to generate any computable number and unable to generate any uncomputable number. However I don't think something being a turing machine, under certain assumptions, is really a satisfying answer. Something being a turing machine or not is only really relevant when talking about computing power, and it's obvious that a computer and a brain are exactly equivalent on that front, but that doesn't mean that a brain *is* a computer. I think it would be better characterized and studied(formally) as a dynamical system, which can show computational characteristics under some circumstances.

>>4313 I agree with this analysis completely. A brain is a computer, but not merely a computer, more like a complex system of integrated information in constant flux. Dynamical would be an apt description for sure.

>>4313 >However I don't think something being a turing machine, under certain assumptions, is really a satisfying answer. The human mind can really only think of one thing at a time, at least consciously, humans dont really have control of recall, humans can try and remember things, but it is very difficult for humans to remember random information. There are lots of psychology games where an examiner say a random list of words and then ask the person being tested repeat what was said. The average person would be lucky to repeat 10 or 20 words. Also plenty of studies have been done testing humans to multitask, that is solve problems simultaneously, and it has been shown that humans can only keep about 7 or 8 different problem sets in their head at the same time. This is the reason humans use computer, because computers have infallible memory recall no matter how large the list of things to remember.

>>4313 So you would classify computers as hardware-only. What about quantum computers, they are different architecture from hard- and wetware, do we need a new name for them also? Some researches have presented an idea that biological computers could be grown, so when you need more ram, no need to buy it, just feed some nutrients and it will grow more storage. These also would need a new name.

>>4316 brains can be trained to do computational tasks, sometimes faster than a computer could. It just takes lots of practice. >>4323 >What about quantum computers, they are different architecture from hard- and wetware, do we need a new name for them also? but a quantum computer can only compute, and similarly with the biological computers. The brain does many things, in fact I would say most of the things it does, that are not computation.

>>4337 >I would say most of the things it does, that are not computation How about the humanoid-like robots they have built? It walks, talks and manipulates objects, it's all computation though since it runs on a human-made CPU. Are you saying that human brains do the same thing but do not compute anything in the process? I'd argue that neither the robot or you and I are aware of the actual numbers that are crunched, it happens on the background.

>>4351 >Are you saying that human brains do the same thing but do not compute anything in the process? Yep, among many other things.

>>4279 >>4288 These lainons are on point. "Is the human brain a computer" isn't an empirical question - whether or not you want to include human brain's in the same category as computers isn't a "fact", it's just a distinction you're choosing whether or not to make. Which is not necessarily a _pointless_ question, but I think it does want a little more context. I think most of the people on this board have a fairly solid grasp of the similarities and differences between the human brain and a computer. Besides the one or two posters that are clearly drinking some kind of soykaf, it seems like the argument is basically just "what is the salient feature of a 'computer'?" And that's a really contextual question. Whether or not you want to call a brain a computer or a computer a brain depends a lot what purpose it serves the discussion you're having. And I think most lainon's are smart enough to know when they are worth treating as part of the same category, and when they are worth distinguishing. Is there a specific context where you think people are making an inappropriate distinction, or failing to make a necessary one? Or is this just a "brain and computers: compare and contrast" thread?

>>4270 I'd say it's computer-like. It can be compared to how a computer works but it is NOT a computer itself. (I got this idea from Prometheus Rising)

As many other people stated, it depends on what you define as a computer. In my opinion we will never be able to recreate our brain, the thing so special about humans are that we know that we exist, we know about our existence. And no-one will ever recreate an AI that knows that itself exists.

The brain is fundamentally different from a computer because it is capable of abstraction, it forgets too easily, and it lets emotions influence the logical instructions given. And I'm not trying to imply the last one is bad either, that's how a brain should be. Let's say there's an "if" statement in Python with a ton of elifs. What if the brain decides to ignore some if and elif statements simply because it feels like it? For example, that one time when that Russian guy at the nuke controls refused to retaliate with nuclear weapons at a false alarm - if the brain was just like a computer it would have followed instructions exactly, and we would have all died.

it is capable of imitating (emphasis on this) logic, but theres the grand element of consciousness that screws with everything.

>>7169 This is a very good explanation, thank you lainon!

>>4270 It can do everything the computer can, but there are things a computer can't do yet: http://www.deanradin.com/evidence/evidence.htm

A "Computer," is anything that can compute (solve math problems), so yes. Hell, "Computer," was originally a job title.

>>7166 >And no-one will ever recreate an AI that knows that itself exists. Why not? All that's going on in the brain are electro-chemical reactions. Why couldn't those reactions ever be simulated? It's orders of magnitude too complicated to do NOW, but that's just because there's what amounts to way too many tiny moving parts; it's still just a function of the way certain atoms an electrons are interacting (admittedly, loads and loads of them), so with enough processing power and memory (and you would need a LOT), you should be able to simulate it like any other physical process.

>>4271 >Maybe one day our understanding of the brain will be such that it won't be meaningfully different from any old electro-mechanic computer This is what I've been looking forward to my whole life, can't wait to code in Human

>>8084 Imagine working in architecture where each element works in negative/positive feedback loops or feeds back to something else and every invidual pathway is different based on what it connects to. Also don't forget that paths will change over time, new ones will be created and old ones will be weaker and weaker. I can't imagine level of abstraction needed to program in "Human" "programming proteins" in DNA code at the other hand...

>>4270 The human brain is probably *not* a computer. Why? By historical example. Would you say the human brain is like a clock? Because when clocks were the new thing, statements to the effect that humans were like clocks were common. Now computers are the new thing, and suddenly the brain is a computer. Probably the analogy will be updated again in the future.

I think this question is dumb, you are just playing with concepts and definitions, it'll change nothing in reality.

>Is the human brain a computer? can you tell me how much is 2+2?

>>7268 >All that's going on in the brain are electro-chemical reactions. We don't know that. >Why couldn't those reactions ever be simulated? Most reactions can't be. >enough processing power and memory Even simple systems we can't always simulate accurately. Maybe there just isn't enough energy in the whole universe for that. And you're talking about simulating the most complex ever observed. Complexity-wise our brain is like 99.99% of the know universe, but probably much more than that.

>>8440 well the human brain is like a clock in a lot of ways. And clocks are like computers. And computers are like brains. And things are like other things. It's a stupid question to ask but it comes from the human desire to categorize things, in reality everything is gray, nothing fits perfectly into a box. We end up arguing about what the boxes are instead of the things we're putting in them.

The way I see it is like this: A computer in and of itself is not like a humans brain. But, let's think about this for a second. Humans tend to first develop and build things at a large scale, then as they gain more understanding of how those things work, they make them smaller. So my theory is that a computer is more like a neuron. And a network is more like a human brain. So the internet is a brain spread over the entire earth and is not connected in a solid way. If you had control over every server in the world and used smaller networks of them for specific tasks it would probably start to look like a brain. Maybe like using each continent for different things, like Africa for data/memory, Asia for speech, etc. So eventually if we have more computers over smaller networks it would get smaller and smaller. Within 15 to 20 years each home could have at least 20 devices on a home network thanks to IoT. Take that network and times it by 10,000(a small portion of a large city) and you have 200,000 computers on 10,000 home networks that could all be attached to a larger network. This would only go up, and then you would see better swarm robotics and AI in general "Indeed, much recent research has focused on microbot communication, including a 1,024 robot swarm at Harvard University that assembles itself into various shapes" -2014 Ars Technica

You're really only hurting yourself when you oversimplify for the purpose of making meaningless analogies. Your brain isn't a computer, your eyes aren't LCD screens, and your hands aren't keyboards, but go ahead and believe they are.

I believe this thread was inspired by a discussion in the Lainchan IRC and so I believe I was the inspiration for this thread. To clarify my position, I argued that the human brain is not a formal system, because I don't believe there are problems a human being is inherently incapable of solving, which wouldn't be the case if it were a formal system. I don't believe there are programs that a human is inherently incapable of determining whether they halt or not, as an example.

This is an interesting listen and seems relevant to the discussion, so I'll share it. https://www.youtube.com/watch?v=cMxbSsRntv4

>>9904 >Your brain isn't a computer, your eyes aren't LCD screens, and your hands aren't keyboards, but go ahead and believe they are. Not yet at least

>>9998 What I meant was; it isn't for you primitive squares. My brain, eyes, and hands are all speak and spells and I eat salt and vinegar embedded linux chips every day for a snack.

No, and neither is the electrical impulse engine you probably posted that on. Both can be modeled as computers, however.

An argument ive seen used a lot is that the brain is a computer because it can compute. I ask this: Is the brain a synthethizer because it can imagine music? Is the brain a camera because it can capture images? Maybe the brain has a computing subcircuit, but calling the whole thing a computer is akin to calling the forest a tree. Humans are more than commodities.

>>4270 It's worth noting that we don't fully understand how the brain works, and don't understand at all how a physical object can result in consciousness. So this might be one question we never answer. As the saying goes, if the human brain were so simple we could understand it, we would be so simple we couldn't. Personally I lean towards the brain being more than a computer, mainly thanks to Diamond Age.

We're getting much better at understanding how the brain works. See this explanation of the predictive processing model: http://slatestarcodex.com/2017/09/05/book-review-surfing-uncertainty/ To summarize, predictive processing explains the brain as a multilayer prediction machine with a bottom-up stream of sense data and a top-down stream of predicted sense data. Each layer parses the sense data into a higher level format, and cashes out the predictions as lower-level expected sense data. Then, depending on how closely the top-down and bottom-up streams match and how wide the confidence intervals on each are, may override the sense data with the predictions, or signal surprise, causing the predictors to change their models. This, then, explains many things: dreaming, how motor control works, the placebo effect, what the different neurotransmitters actually mean, schizophrenia, and much else.

>>4271 >when it dysfunctions, we're left with very, very primitive ways of trying to fix it. you haven't seen me trouble shoot my computer issues apparently

>>4270 Read some of the classics on the theory of computation--Turing et al.--and read up on some basic neuroscience (an actual textbook, if possible). You'll see a few commonalities

The brain is a computer insomuch it has modelled the computer after itself. What/how we percieve. Everything, from our vantage point, can be reduced to systematics, as we do with the computer. Where i would differentiate though, is in the emergent process of consciousness, and perception itself. We know what makes it, biologically, but we can't explain why it is the way that it is.

I think you should watch "All Watched Over by Machines of Loving Grace". It's a documentary that gives examples of human failures in modelling the organic in terms of the mechanic. Does this topic really belong on this board?

>>4270 >Is the brain a computer? >Computers compute numbers. >Humans can do mental calculations. stop

>>4270 No, it's not a computer. Back before computers, people made analogies to the human brain being like the steam engine. People reach for what they know. We don't understand the brain yet.

>>13516 This

Using a strict definition of a computer being a machine that can execute an arbitrary set of logic instructions, then you would be asking whether it is equivalent to a Turing machine. No real computer is equivalent since they have finite memory, but we still call them computers as an approximation since the number of states is so large. Using that definition, the brain is only a computer given pen and paper. Otherwise, except in cases of savants, it's a very poor approximation of Turing machines given the limited memory.

>>4270 My reasoning is that a computer is a subset of a brain (in other words, all the elements of a computer are in a brain but not all of the elements of a brain are in a computer)

>>4279 He's not "extending the definition". The only defensible definition of a computer in discussions pertaining to comparability is something that computes. The meat of the question OP posed is whether the human mind is materialist or not. If that which is you is a product of the wet material in your skull, then human consciousness is commutable. Unless you're religious, it's a pretty defensible position. Detractors would be people like Roger Penrose, but he has no evidence for his ideas. He forms conclusions then goes fishing for evidence, and fails to even any.

I believe the universe is cellular autamata and so is your brain and so are computers >>13538 You would also need infinite paper and ink

>>4270 >Is the human brain a computer? You mean is there biological determinism? I dunno, and don't really care, other than for purposes of preventing neurological damage. Every year (since what, 1960?) someone makes some new computer program and the media yells OH NO WE ARE NO DIFFERENT THAN COMPUTERS LOOK WHAT THAT COMPUTER JUST DID! >My argument is that I can calculate in my head everything the computer can, therefore they are essentially the same (if it walks like a duck...). Non-sequiter.

Yes, but your brain can physically reorganise its hardware to adapt for future inputs. Don't think we'll see something similar based on semiconductors.

>>15626 I think ML self-modifying FPGA chips exist now.

I think you could call a brain a computer in certain ways. It has branching logic circuits. Not exactly the same as our computers do, but it's at least something you could make a credible comparison to. Computers are also kind of brains. Different chemistry and architecture, but once again they are a network of logic elements that take an input and organize an output based on some internal pattern of structures. But then again, so is a Rube Goldberg machine, or an office building full of people, and a bunch of other things. Calling these brains is kinda interesting but ultimately useless. Of course there's a matter of degree here. A computer is a very very crappy brain, or rather a brain is a transcendent computer. I think there's some use in defining both as having a certain basic structure that is at least superficially similar, but taking the analogy much further is likely useless. A network of computers might be more brain-like in that it changes its layout to become more optimal and learn how to grow around problems. It's certainly more effective than a single machine. A network of human brains like the aforementioned office has terrible network throughput—way slower than a computer network—but it's still likely more capable than a single individual. In certain ways, at least.