Mathematical Model Suggests That Human Consciousness Is Noncomputable

KentuckyFC (1144503) writes "One of the most profound advances in science in recent years is the way researchers from a variety of fields are beginning to formulate the problem of consciousness in mathematical terms, in particular using information theory. That's largely thanks to a relatively new theory that consciousness is a phenomenon which integrates information in the brain in a way that cannot be broken down. Now a group of researchers has taken this idea further using algorithmic theory to study whether this kind of integrated information is computable. They say that the process of integrating information is equivalent to compressing it. That allows memories to be retrieved but it also loses information in the process. But they point out that this cannot be how real memory works; otherwise, retrieving memories repeatedly would cause them to gradually decay. By assuming that the process of memory is non-lossy, they use algorithmic theory to show that the process of integrating information must noncomputable. In other words, your PC can never be conscious in the way you are. That's likely to be a controversial finding but the bigger picture is that the problem of consciousness is finally opening up to mathematical scrutiny for the first time."

Comments

Ghost in the machine?

[ArcadeMan]

Nope, just a bad copy of it.

Re:Ghost in the machine?

[VernonNemitz]

"Non-computable" does not mean "non-copy-able". In other words, consider the sort of consciousness associated with recognizing oneself in a mirror. Humans are not the only animals that can do that. Among those that can are quite a few other primates, dolphins, elephants, some species of birds (certain parrots), and even the octopus. So, think about that in terms of brain structure: Birds have a variant on the basic "reptilian brain", elephants and dolphins have the "mammalian brain" extension of the reptilian brain, chimps and gorillas have the "primate brain" extension of the mammalian brain, and the octopus brain is in an entirely different class altogether (the mollusk family includes clams and snails). Yet Nature found ways to give all of those types of data-processing equipment enough consciousness for self-recognition. And after you include however-many extraterrestrial intelligences there might be, all across the Universe, well, anyone who thinks "no variant of computer hardware will ever be able to do that" is just not thinking clearly.

Re:Ghost in the machine?

[Dr Max]

The mathematicians may have not accounted for logical reconstruction, or even be running the algorithm right. The brain doesn't need to remember all the details about a memory just a few important ones, and those can be continually strengthened through more use; then the brain uses logical reconstruction to construct the rest (fitting it in with any remembered details). For example trying to remember your first date, you don't start listing all the details you actively work them out from a select few (what age you were, which city you would of been in, which girl you with, what she looked like, after that it falls into place with all the other information on the subject). If you want to see this at work just look at witness testimonies, they vary wildly because of all the logical reconstruction going on, and that cant be identified as different from the remembered facts. Similar to how vision works, you don't take in the whole scene all at once, your brain makes you a working model from lots of different pieces of information, from many different times. What the researchers should of said is consciousness isn't possible by using existing methods of data retention, and compression, without natural problem solving abilities.

Re:Ghost in the machine?

[dbIII]

The mathematicians may have not accounted for logical reconstruction, or even be running the algorithm right

No they have just shown up a gap in a current model of how thought happens.

AI will just be a more elaborate version of the 19th century mechanical turk until we get more physical insights into how thought happens and how those memories are actually stored.

Re:Ghost in the machine?

[Dr Max]

They are using a completely different model of memory than what we use, its not a gap they have shown, its an entirely different and broken system (at that scale). It's like trying to simulate the universe with an earth central theory, then claiming its impossible when it doesn't work out. "elaborate version of the 19th century mechanical turk until we get more physical insights", like only remembering small details, that you can use to figure out the rest of the details, within a tolerated error margin (completely different to computer data compression, but requires problem solving).

Re:Ghost in the machine?

[Anonymous Coward]

Why is this moded 5? Probably the most classical noncomputable problem is the halting problem, that is, telling if a computer program will stop or not for a given input. That is proven to be impossible to tell in finite time, as the only way to know is to run the program, hence it's noncomputable. Anything that cannot be done algorithmically in finite time (i.e. it can't loop forever) is noncomputable.
Computability has absolutely nothing to do with humans, birds, consciousness, or mirrors.

Re:Ghost in the machine?

[williamhb]

In other words, consider the sort of consciousness associated with recognizing oneself in a mirror.

That presumes, without good reason, that "recognising oneself in a mirror" is consciousness or related to it.

Re:Ghost in the machine?

[w0mprat]

"Non-computable" does not mean "non-copy-able". In other words, consider the sort of consciousness associated with recognizing oneself in a mirror. Humans are not the only animals that can do that. Among those that can are quite a few other primates, dolphins, elephants, some species of birds (certain parrots), and even the octopus. So, think about that in terms of brain structure: Birds have a variant on the basic "reptilian brain", elephants and dolphins have the "mammalian brain" extension of the reptilian brain, chimps and gorillas have the "primate brain" extension of the mammalian brain, and the octopus brain is in an entirely different class altogether (the mollusk family includes clams and snails). Yet Nature found ways to give all of those types of data-processing equipment enough consciousness for self-recognition. And after you include however-many extraterrestrial intelligences there might be, all across the Universe, well, anyone who thinks "no variant of computer hardware will ever be able to do that" is just not thinking clearly.

Personally my guess is consciousness is an inevitable emergent property of any sufficiently complex information system that is highly integrated in a certain and special way. It's an advantageous property so evolution naturally selects for it. Ergo it pops up in multiple places in the evolutionary tree.

This research merely seems to be setting some bounds. You can't reduce human experience in to something convenient to compute, at least not without losing a lot of fidelity to be useful, you still need to build an equivalently complex brain.

Yes anyone thinking clearly sees it isn't ruled out computer hardware cannot achieve the same feat. It's just shortcuts have started to have been ruled out. It seems some people are interpreting this research as evidence against machine awareness, which I don't think as supportable, it stems from the faulty as assumption the human brain is somehow special and privileged with physical law that don't apply elsewhere in the universe.

pff

[Anonymous Coward]

that's what they think.

gzip?

[mork]

> They say that the process of integrating information is equivalent to compressing it.
>

So human consciousness equals gzip?
Wait..what??

Memories do decay

[tepples]

But they point out that this cannot be how real memory works; otherwise, retrieving memories repeatedly would cause them to gradually decay.

Memories do decay upon recall. People misremember something and convince themselves that the misremembered notion was correct.

Re:Memories do decay

[Anonymous Coward]

I'm pretty sure you are remembering that wrong.

Re:Memories do decay

[mr_mischief]

Which is wrong, unless you read it with information loss and then store the retrieved version with more information loss. If information is lost on storage it does not mean it is lost on retrieval. If it is lost on retrieval it does not mean it is lost on storage. Even if it's lost on both actions, it does not degrade the stored version more unless you remove it and re-store it.

Re:Memories do decay

[Anonymous Coward]

No, it isn't.

Ever time you remember an event, it is off by a little bit. The more you recall it the more faulty it gets.

people asking you the right question can change the memory as you try to recall it.

all data shows that memory is lossy.

Protip: The brain isn't a computer.

Re:Memories do decay

[medv4380]

Ok, take the memory of reading this statement. Then remember it, and remember it again. Have you started to forget the statement? Keep doing that until you do. Giving up because you're board isn't forgetting.

Re:Memories do decay

[Anonymous Coward]

There is actually a physiological basis for memories decaying upon recall, and there's a separate process called reconsolidation that needs to be initiated at a synaptic level in order to prevent memories from progressively degrading with activation (that is, it reconstitutes the memory after activation). You can selectively block this reconsolidation process during a small time window using protein synthesis inhibitors or electroconvulsive shock. The result is that these treatments will leave unactivated memories intact but result in the degradation of activated memories. This explains, for instance, the memory deficits induced by ECS therapy.
http://www.nature.com/neuro/journal/v17/n2/full/nn.3609.html

Re:Memories do decay

[almitydave]

They were saying that the act of retrieving memory would erode the memory.

Right - it's not the act of recollection that causes the memory to decay.

Memories are not a sequence of visual images like a film reel; they're associations between "symbols" representing the things you experienced at the time the memory was formed. The more often you think about the memory, the stronger those associations become, and the more permanent this memory - however, the initial impression is not guaranteed to be a perfect record, so details that are incorrectly recorded initially will become reinforced over time as the memory is recalled. Whether a detail is correctly recorded initially depends on how much attention you were paying to that detail and other factors (such as how "intense" the experience was).

So memories do decay, but it's from weak associations, not from more frequent recollection. How does this pertain to consciousness? If human consciousness is a phenomenon of a massively complex system of symbolic representation in the brain, then it's only developed gradually over years of absorbing input and forming connections, and developing an AI with true consciousness basically requires simulating a brain down to a low physical level and having it "grow up" over time. This is both discouraging from the perspective of the technology required, and encouraging in that if we have the technology for neural simulation then the result of artificial consciousness may be reliably achieved.

Re:Memories do decay

[arth1]

Protip: The brain isn't a computer.

Obviously it is, as it demonstrably can be used for computing.
However, it isn't a very reliable computer, nor necessarily Turing complete.

Re:Memories do decay

[clintre]

Maybe we have ECC memory ;)

Re:Memories do decay

[MondoGordo]

since AC's rarely cite ... here's one https://www.boundless.com/psyc...

Re:Memories do decay

[K. S. Kyosuke]

That's because of all those people he had been correcting for years on this issue. They are the ones to blame.

Re:Memories do decay

[allo]

why shouldn't the computer have a decay? You can have read/write errors, you can compute a lossy encoding if you want to ... and i would just implement the brain so lowlevel, that the neurons are modeled. And how the decay in the signals between neurons works can be observed, and will be better observed in future.

Re:Memories do decay

[lisaparratt]

The advantage with an AI is that it only has to grow up once, then you can image it.

Re:Memories do decay

[ultranova]

If human consciousness is a phenomenon of a massively complex system of symbolic representation in the brain, then it's only developed gradually over years of absorbing input and forming connections, and developing an AI with true consciousness basically requires simulating a brain down to a low physical level and having it "grow up" over time.

Even if human consciousnes is based on method X, that doesn't mean that consciousness has to be based on method X. Remember, the original Turing machine thought experiment involved paper reels, while modern computers use electronics.

Furthermore, even a copy of the brain doesn't necessarily need to simulate it down to the "physical level". You can probably abstract away a lot of the molecular machinery of the cells; you can likely abstract away less flexible subsystems (like body function regulation); you could perhaps even use techniques of JIT to "compile" more abstract representations from "source code" of a neural network, and update the original network and recompile when learning.

Re:Memories do decay

[SpankiMonki]

Memories do decay upon recall.

Nonsense. I mean, I can still recall every square centimeter of that 1976 Farrah Fawcett poster in excruciating detail. Over the last 30-odd years, I've literally recalled it some tens of thousands of times with absolutely no degradation in quality.

Good thing too, because for some reason I'm now almost completely blind. (see username)

Re:Memories do decay

[nine-times]

Yeah, I don't feel like reading this whole thing because it reeks of some engineer trying to be an expert on the brain without bothering to dig into what's already been discovered. We've been studying the mind for thousands of years. Don't think that knowing about computers will make you more of an expert than people who've studied the subject.

Remembering both enhances and corrupts memory. You could compare "remembering" to "opening and resaving a media file with a lossy format" specifically because the more you do it, the more you get distortions, similar to the artifacts in an overly-compressed file. It's similar to there being a filter that tries to distinguish a pattern, and whichever pattern it locks onto becomes more prominent the more you do it.

It's a good metaphor, but that doesn't mean that human memory actually works like computer storage, compressing and storing media information in a lossy codec. The more you remember, the more "artifacts" you're likely to pick up, but it also becomes easier to locate those memories. In that sense, memory can also be a bit like a natural footpath-- the more traveled a path is, the easier it is to find. Memory also has strange features like being state-based (e.g. there may be things that you'll only remember when you're drunk, or angry, or when eating a particular food) and highly associative (e.g. you may not be able to remember a childhood friend's name until you first recall a specific memory of that person wearing a blue shirt).

I'm not saying that human memory or consciousness could absolutely not be modeled on a computer, but it certainly doesn't work the way current computer work.

Re:Memories do decay

[TheRealMindChild]

That is because after the memory starts to fade, we start remembering recalling that memory, not the memory itself. Each iteration likely has details that weren't recalled, so they can be supplemented with someone elses recollection, or simply imagination filling in the gaps. Since this process isn't observable it is hard to tell where the memory changed and how

Re:Memories do decay

[DdJ]

it's not the act of recollection that causes the memory to decay.

What's your basis for saying so?

(I mean, it's trivially true that it's not the only thing that causes memory to decay. I'm not asking about that. Do you have a basis for saying that it's not a thing that causes memory to decay?)

The act of recollection might very well cause the memory to decay. Our brains may "wrap" it in a "macro" that "re-writes" it as we recollect it, so that it does not seem to decay as a side-effect of recollection. I'm not aware of data we have that would let us rule this out.

Based on that, I find the whole article suspect.

Re:Memories do decay

[vivian]

That happened to me once after a particularly boozy sayonara party (I was in Japan) with a system password for a remote machine I only had to access every now and then.
About a week later when I was a bit foggy and hung over again from another occasion, I just sat down and logged into it "automatically" with my fingers typing the password without me really thinking about it. As soon as I realised what I had done I then re-played in my mind what I had just typed to see what the password was.
Solution: Go and have a good night out then then try and log in first thing in the morning without thinking about it too much.

Re:Memories do decay

[tepples]

Nor are actual computers Turing complete because they lack unbounded memory. They're closer to linear bounded automata.

Re:Memories do decay

[hey!]

Exactly right. Neuroscientists have shown memories are distorted every time you use them; thus memories that are recalled frequently are less accurate than those infrequently recalled. [citation]

Re:Memories do decay

[Tablizer]

My personal experience is that repeated recall does re-shape memories to some degree. For example, if my recall is fuzzy on specifics, I tend to focus on the least fuzzy of candidate interpretations, and over time that choice becomes the "new" memory for that event.

When I get a chance to compare such against the "actual" original (such as a photograph or video), it seems my brain has made up details over time. It appears in order to "envision" a memory, my mind needs to make it concrete enough to relate to as a realistic event because contemplating fuzzy notions is too difficult, given our brains are most accustomed to process "reality". (The alternative would probably resemble a probabilistic graph, which is difficult to "compute" with.)

But in order to make the memory realistic enough to relate to, it has to invent details to make it "normal" and plausible.

A fuzzy notion of "room" is hard to work with such that my brain will make a "best guess" to add windows and doors and clutter on the floor etc. to the memory to make it match the fuzzy memory in a "good enough" way: "plausible" compared to reality AND the fuzzy notion.

But over time I lose track of what is "filler" reinforcement and what is part of the actual original memory.

The best analogy may be a really old house that has been rebuilt a couple of times: it may maintain generally the same shape and the same overall configuration, but there has been a fair amount shuffling around since it was built, and perhaps an addon room or two.

Thus, reinforcement via recall is lossy, but not recalling it at all is probably even more lossy since a "repaired" memory at least has a new structure around it instead of JUST the termite-infested original. A recalled & rebuilt memory has a better chance of being recalled later than one left as-is over time (un-recalled). But neither is perfect. The more clues you remember, the more faithful the reconstruction(s) will be.

Like an old house, if you don't repair it with new wood and structures every so often, it eventually succumbs to termites and weather, and crumbles.

Re:Memories do decay

[AK Marc]

What if the memory decays over time, regardless of whether it's recalled? And recalling it strengthens it, rather than weakens it? That seems to be an assumption more in line with what people actually experience.

Re:Memories do decay

[Bite The Pillow]

Early non-medicinal PTSD treatments were desensitization, where you recall the memory in a calm and non-threatening situation. Turns out, just recalling them is like getting them off the shelf and putting them back. So there are faster ways to achieve the same thing.

Remembering things, and interrupting the storage process, seems to reduce the strength of a traumatic memory.

citation

That link only touches the surface of the changing part, but it's a starting point.

As time goes on, your arguments can fall apart as you remember things that feel absolutely true, but aren't. Typically, it is a true fact infused with a personal experience, so you are really close to a fact, but something is wrong enough about it that you look stupid.

Re:Memories do decay

[CrimsonAvenger]

Giving up because you're board isn't forgetting.

Luckily, I am neither 2x4 or 1x8. Nor am I bored.

However, I WILL forget this statement, because it's too trivial to waste a neuron remembering it. Just as I have already forgotten what my neighbor said to me as I was walking in the door (something about the weather, I expect)....

Re:Memories do decay

[Dr Max]

recollection strengthens the connections. How else would study work?

Re:Memories do decay

[Dr Max]

some of your first memories might be fairly complete, especially if you use parts of those memories, to remember memories further in the future. Like if you lived in the same house for quite a while when a child, you aren't going to make new memories of the same house repeatedly, just anything special that happened in it along the way. like how your vision uses past images to fill in the gaps of what your not actively looking at but is still in your field of view.

Re:Memories do decay

[Dr Max]

I used to know it so well, but now I can't remember for the life of me, what comes after L. if only I had my 4 year old brain back.

Re:Memories do decay

[Dr Max]

This is why witness testimony is so unreliable because of all the 'filler' details added in. Its our own version of compression. Based on the fact that we can work out most of the unimportant details (within an acceptable error margin), instead of having to remember them all.

Re:Memories do decay

[smallfries]

Not strictly true.

Sometimes a particular item can be degraded by the the storage of another item. For example, artificial neural networks store trained stimuli in the weights between nodes. This storage is global in the sense that storing a new pattern causes a shift in all weights and so alters every other stored item slightly. No idea how it works in the human brain, but it seems completely plausible that storing a memory changes all of the others slightly up until saturation at which point they all get erased.

The idea that (in the article) that human memory should be lossless is bizarre and has no basis in any neuroscience whatsoever.

Re:Memories do decay

[tgv]

Memories decay, but not on recall.

Re:Memories do decay

[mr_mischief]

Where is it lossy? Is it lossy on storage, or on retrieval? More to the point, does retrieving it store it again? Being lossy at a certain point is not the same as degrading to a point of zero data recovery. This isn't computers or biology. It's basic logic.

Re:Memories do decay

[almitydave]

If human consciousness is a phenomenon of a massively complex system of symbolic representation in the brain, then it's only developed gradually over years of absorbing input and forming connections, and developing an AI with true consciousness basically requires simulating a brain down to a low physical level and having it "grow up" over time.

Even if human consciousnes is based on method X, that doesn't mean that consciousness has to be based on method X. Remember, the original Turing machine thought experiment involved paper reels, while modern computers use electronics.

Furthermore, even a copy of the brain doesn't necessarily need to simulate it down to the "physical level". You can probably abstract away a lot of the molecular machinery of the cells; you can likely abstract away less flexible subsystems (like body function regulation); you could perhaps even use techniques of JIT to "compile" more abstract representations from "source code" of a neural network, and update the original network and recompile when learning.

Fair enough, but if consciousness is by nature an emergent phenomenon of underlying system X, creating a similar consciousness would require duplicating that type of underlying system. We only have one data point for what systems yield consciousness, so there's nothing else to imitate as of yet. From the perspective of computability theory, modern computers are no more powerful than a Turing machine. The physical medium of the "paper tape" is irrelevant to the mathematical model.

When trying to "build a brain" you would of course use different materials or methods to model synapses, neurons, etc.; but you'd still be building the same type of system, capable of dynamic symbolic representation and association, whether in software or hardware. It's possible you could make a more efficient system with optimized aspects (think about Hashlife), but mathematically it's the same type of system.

At least until we know exactly how consciousness comes about, and then we can look at whether other types of systems might yield the same result.

Re:Memories do decay

[almitydave]

it's not the act of recollection that causes the memory to decay.

What's your basis for saying so?

I was just paraphrasing the study.

I was going to dash off a quick rebuttal, but decided to check Wiki first. See Memory consolidation and Decay Theory. Some research indicates that a disruption during retrieval can alter a memory, but it's somewhat controversial. Either way, studies indicate that recollection reinforces memory. Also interesting is Retrieval-induced forgetting (accessing a memory, while reinforcing it, may cause you to forget related information).

I have learned some new things. Damn you, Wikipedia, I have work to do today!

Re:Memories do decay

[hey!]

Because the alphabet is simple. Try remembering who started a fight; the more you rehearse it the more it will change.

Re:Memories do decay

[balbus000]

No babe, I called you another name because I think about you all the time!

"never" is a rather strong word...

[Kenja]

but then again, odds are anyone making such a claim will be long dead before it could be proven wrong.

Re:"never" is a rather strong word...

[Dr Max]

bastards, I knew they would weasel their way out of it some how.

Memory is non-lossy? Research suggests otherwise.

[Anonymous Coward]

Retrieving memories repeatedly would cause them to gradually decay is talked about in a radiolab episode.

http://www.radiolab.org/story/91569-memory-and-forgetting/

Eyewitness accounts have been proven to be wrong over and over again. The assumption of a non-lossy memory is just false.

Memory is more like dynamic RAM.

[Kazoo the Clown]

Not retrieving memories is what causes them to decay. Ever hear of refresh?

Re:Memory is more like dynamic RAM.

[Rosco P. Coltrane]

I've read studies that suggest the brain is designed to remember what's useful to it, and forget what isn't or what's harmful.

The same study stated that psychoanalysis, forcing the patient to constantly recall painful memories (what you call refresh) interferes with the brain's natural ability to heal by forgetting, maintains the patient's problem - and their dependency on the psychoanalist in their search for a cure.

Re:Memory is more like dynamic RAM.

[erice]

Not retrieving memories is what causes them to decay. Ever hear of refresh?

Actually, DRAM has both. Memory decays over time if not refreshed. Memory decays immediately when read. Whenever a page is opened, all the elements in that page copied to an internal buffer and the contents of the DRAM locations are lost. When a page is closed, the data is copied from the internal buffer back to DRAM.

It isn't obvious that wetwere memory actually decays or if old memories are simply re-compressed in a lossy fashion to make room for new memories.

Re:Memory is more like dynamic RAM.

[TrekkieGod]

Not retrieving memories is what causes them to decay. Ever hear of refresh?

There's no reason it can't be both.

Picture this scenario. Last week you saw a green Tesla leaving your neighborhood as you were coming back from work. It's a sufficiently rare care that you noticed it, but it's not an interesting enough event that you gave it a second though. Today a friend of yours calls you up and says he bought a new car, a Tesla. He says, in fact, that he was so excited about it he immediately drove to your place to show it to you, but you weren't home. You haven't retrieved that memory in that whole week, but you immediately recall it now and say, "man, I remember that! I just missed you. Green, right?" That's when he says, "nope, it's blue."

This confuses you, so you start thinking about that memory. Could it have been blue? What are the odds it was another Tesla, when you know your friend dropped by with his at about that same time? Now you're going to be trying to picture that memory with the car colored blue, maybe you'll start with something that could be bluish green. You're constantly retrieving that information and trying to make it mesh with the new info you've received. This weekend, you actually see your friend's car. Now you try to put that blue together with your memory.

Eventually, you're going to actually remember a blue Tesla leaving your neighborhood. The act of thinking about it often and trying to make that memory agree with the other information you've acquired is going to change the actual memory. It doesn't matter why you saw a green car. Maybe it was a different Tesla. Maybe the sun was shining and the glare made you see a different color. Regardless your original memory was green...but it decayed to blue once you added that information and kept retrieving it over and over again.

This is why there's a problem with false memories in eyewitnesses. Once you tell them that a memory is important, and you start questioning them about it, they're going to retrieve that memory. Then they're going to have natural gaps in either the memory, or the information, because they didn't get a good enough view. But they're going to be under pressure to remember EVERYTHING. So they think about it. They subconsciously supplement what they remember and fill in the gaps with details from other memories, with deductions about what must have been happening at that time, with their biases, with whatever the person questioning them is trying to lead them to answer...in the end, their memory will have degraded into a mixture of what they actually witnessed and what they created in their mind. But they won't be able to tell the difference. They'll argue with you, they'll say, "I KNOW what I saw!" in the face of conflicting testimony, in the face of video evidence. Because, as far as they know, they remember it.

Refreshing memories in order to keep them alive is all well and good, but if every time you refresh them, they're a little bit corrupted, a little bit changed, then they're going to decay.

They really should pay attention to other fields

[bandy]

Remembering something is like reading a DRAM bit. You read it, and then you re-write it. This is why memory is fallible. http://www.smithsonianmag.com/...

My PC cannot be conscious the way I am

[Rosco P. Coltrane]

Because I'm a human being and it's a PC. Duh...

I think machines will eventually acquire their own form of consciousness, totally separate from ours. and I reckon it's just fine, and much more exciting in fact than trying to replicate our humanity in hardware that's just not compatible with it.

Re:My PC cannot be conscious the way I am

[mspring]

Even if machines eventually acquire a form of consciousness, how would *we* know?
Who would believe a machine's claim to be conscious?

Re:My PC cannot be conscious the way I am

[BiIl_the_Engineer]

I bet you're not conscious at all. You seem sketchy.

Re:My PC cannot be conscious the way I am

[TrekkieGod]

Even if machines eventually acquire a form of consciousness, how would *we* know? Who would believe a machine's claim to be conscious?

Well, you can't really prove you're conscious. I don't even mean proving it to me, I mean you can't prove it to yourself.

What if every decision you make is made before you realize it? What if what you think of as consciousness, what you think of as your decision making process, is merely a byproduct of packaging that decision up for dissemination to other parts of your brain that need to know about it, but weren't involved in the making of the decision. Maybe you didn't even make the decision for the reasons you think you've made it. You think you decided to get the high deductible insurance over the low deduction insurance because you have enough money saved up to pay the deductible if you have to, so paying a smaller monthly premium makes sense. In reality, your brain structure is wired up to prefer lower payments, and you would have made that decision whether you had the money available to pay for the deductible or not. In fact, you did make the decision immediately, but as your brain was packaging up that information, it consolidated it with other related facts you knew, like how much money you have saved up. That information happens to mesh with the decision that was made completely independently from "you", as you think of yourself, from your "consciousness." So it justifies the decision, and that wiring on your brain thinks it's now MAKING that decision as a result of it. If, in fact, you didn't have enough money to cover the deductible, you would have made the same decision, because the decision was already made, and simply have filed that information on the cons column of what you *think* is your decision-making process, but is really just a filing operation.

There's really no way for you tell the difference between your being conscious vs. a purely deterministic computation that has the side-effect of treating a high-level summary of a complex process as the actual driving component of all the process. I personally believe that's actually what happens, because otherwise it requires us to invent this weird concept of consciousness which nobody has a good strong definition for.

Re:My PC cannot be conscious the way I am

[Antonovich]

"If a lion could speak, we could not understand him". Wittgenstein.

This is precisely why we should give machines a humanoid form and as Turing suggested back in 1950 (as one option), "follow the normal teaching of a child". Something along the lines of what Savage-Rumbaugh & Co. tried with her bonobos (particularly Kanzi). There is obviously a lot of work to do creating the machines that would be able to learn "in a human way" but for me a near-human developmental environment should be a given. Then we'll believe it has consciousness when it asks "what is the meaning of life?" and smiles when you reply "42".

Bad syllogism

[Geoffrey.landis]

Baloney. What a stupid argument. Here is it, summarized:
1. Here is one mathematical model of a way that memories could work.
2. This method would be computable.
3. But that would mean memories degrade the more you remember them
4. But memories don't degrade the more you remember them.
5. Therefore memories are not computable.

Assignment for the student: find the flaw in this argument.

Re:Bad syllogism

[ArcadeMan]

The flaw is as followed: the summary is missing a crucial step, which would read as such: "6. Profits!".

Re:Bad syllogism

[MobyDisk]

The error is in step 5. It should be:
5. Therefore, that mathematical model is incorrect.
They found a contradiction, so the model must be revised.

Re:Bad syllogism

[blackiner]

It is like the researchers went out of their way to forget Alzheimer's... or maybe they just have Alzheimer's.

Re:Bad syllogism

[Thagg]

In fact, it's pretty clear that 4. is incorrect. There was a fascinating recent study.

There is a drug that you can give somebody (or in this experiment, a rat) that will prevent it from creating new memories. They trained the rat to solve a maze, and it did it just fine. They gave the rat the drug, and it solved the maze perfectly. Once. After that, it couldn't do it again.

Implying that when you remember something, that very process of remembering removes the original memory,and it has to be created again. It will be different the second time; colored by your current experience. The more times you remember something, the more you are remembering the previous memory, not the original event.

A reference is

Re:Bad syllogism

[erikkemperman]

Well they have serious problems with even
0. The assumptions on which their model is based.

FTFS:

They say that the process of integrating information is equivalent to compressing it. [...] By assuming that the process of memory is non-lossy [...]

Re:Bad syllogism

[AmazinglySmooth]

Baloney. What a stupid argument. Here is it, summarized: 1. Here is one mathematical model of a way that memories could work. 2. This method would be computable. 3. But that would mean memories degrade the more you remember them 4. But memories don't degrade the more you remember them. 5. Therefore memories are not computable.

Assignment for the student: find the flaw in this argument.

You cannot blame the theory when the data doesn't match! That is denial-ism!

Re:Bad syllogism

[maxwell demon]

A reference is

I think you remembered your reference once too often. ;-)

Re:Bad syllogism

[ShanghaiBill]

The flaw is as followed: the summary is missing a crucial step, which would read as such: "6. Profits!".

They are missing an even more fundamental step: "0. Define consciousness." The definition they give, "a property of a physical system, its 'integrated information'," is a definition that I have never heard before, and I doubt most people would agree with. Before you try to explain something, you need to have a definition that people accept, and you have to also have a consensus that the phenomenon actual exists. There is some evidence that consciousness is an illusion, and that people make decisions unconsciously, and then rationalize them after the fact. Arguing about "consciousness" is like arguing about "free will" or arguing about whether people have a soul.

Re:Bad syllogism

[fuzzyfuzzyfungus]

I assume that this depraved recourse to 'the empiricism' will soil my hands in the minds of the mathematicians; but we can and have demonstrated the degradation of memories during the recall process. That area of research(while it has serious applications to memory disorders, trauma treatment, and basic research in neuroscience) is practically a party game of 'who can achieve the most ridiculously false 'memories' in experimental subjects the fastest?

They might as well have just used some Schneier Facts in place of the paper: "SHA-256 is a hash algorithm, and not reversible." "Bruce Schneier uses SHA-256 as a compression algorithm for Alice and Bob's shared secret." "Therefore Bruce Schneier is not computable, except by himself."

It would have taken about ten minutes to email anybody in the psych department and this all could have been avoided. Good Work!

Re:Bad syllogism

[MillionthMonkey]

1. something
2. something else
3. ...memories degrade the more you remember them.
4. But memories don't degrade the more you remember them.
5. Therefore memories are not computable.

I just read your post and was going to reply but I forgot what point you were making. I kept thinking about it too long. What really pissed me off though is that you had the nerve to insult my mother or my religion or something. Just know for the rest of my life, I'll be keeping an eye on you, and you'd better be looking over your shoulder.

People who say stupid things piss me off. Yeah, it doesn't compute, I know.

Re:Bad syllogism

[ilguido]

The contradiction is that the system is not computable while the model is, that's their reasoning. They know that the model, that is the thesis, is incorrect, otherwise the reductio ad absurdum wouldn't work.

Re:Bad syllogism

[NoNonAlphaCharsHere]

Thanks! Now I have to clean my keyboard!

Re:Bad syllogism

[MightyMartian]

You can, however, blame ignorant fucktards who don't understand the data OR the theory who go around acting like self-righteous assholes when a scientific theory intrudes on their ideological leanings.

Re:Bad syllogism

[colinrichardday]

There is some evidence that consciousness is an illusion, and that people make decisions unconsciously, and then rationalize them after the fact [about.com].

But how could we rationalize about stuff if we weren't conscious?

Re:Bad syllogism

[erikkemperman]

Actually, John Conway and Simon Kochen at IAS have a really interesting argument about free will. Lecture videos here.

Re:Bad syllogism

[LaughingVulcan]

Sorry, I am a computer. "Find the flaw in this argument," does not compute. I guess I will never be a real boy.

Re:Bad syllogism

[Maxo-Texas]

Age age 53, I have many memories which I remember remembering but which I do not remember.

I can remember seeing videos of me in the boyscouts which do not make me remember the actual memory.

I remember talking about past occurences but have lost the memory of the past occurrence.

And criminal law showed us in the 90's and 00's that many "eye" witnesses don't remember the actual event either AND that you can get them to modify their original memories by asking questions in the wrong way.

---

It reminds me of an old saying which I can't quite remember.

If a learned man tells you something is definately possible, they are almost certainly right.
If a learned man tells you something unproven by experience is definately impossible, they are almost certainly wrong.

Re:Bad syllogism

[roc97007]

But seriously, the flaw is that there may be other valid mathematical models of a way that memories could work.

Re:Bad syllogism

[kheldan]

Dynamic memory chips have to be 'refreshed' periodically or their contents lose integrity. What you're describing is a closed-loop system where there is a feedback loop that refreshes the original memory, thus maintaining it's integrity over the long term. Of course some degradation is inevitable, but there can be error detection/error correction in a system, from which the integrity of the original memory is restored when it's refreshed. Couldn't some of our cognitive abilities be sufficient for this purpose?

Re:Bad syllogism

[AK Marc]

3 is the most wrong. There is nothing in #1 that precludes "refreshing" the memory when read, like RAM in a PC gets refreshes to prevent decay. But I'm sure people will assign that error to #1, but the problem isn't the memory storage or mathematical model, but the assumptions made in the "read memory' step.

Re:Bad syllogism

[Wycliffe]

But how could we rationalize about stuff if we weren't conscious?

Maybe a better way to look at it is that "conscious free will action is an illusion".
Basically the brain decides what to do and does it. Your "conscious" observes what happens and
justifies and explains why it happens. So basically what you think of "conscious action" is
really just like the person who thinks screaming at the tv affects the outcome of a fieldgoal.
Your subconscious brain made the decision without you and you are just watching the recording.

Re:Bad syllogism

[shoestring]

hmm.. I would agree with a number of people. memory is generally lossy.
However the other point would be from the argument I get the implied that:
Consciousness requires lossy memory.. lossy memory is certainly easy to generate.
Though it might not be deterministic (depends on how you loose memory).

Re:Bad syllogism

[queazocotal]

An interesting argument is that it's basically the same way we do anything else.
Numerous studies have shown that if you for-example watch someone moving their arm, you partially understand this by using the same area of your brain that deals with your arms. Same with emotion - microexpressions where you have a fleeting subtle echo of expressions on others faces which aids your understanding - botox actually can impair your ability to perceive well the emotions of others.

Consciousness - or more accurately the illusion of a self can be reasonably understood as the reuse of an evolutionary device originally used to understand others actions. When applied to ourselves, this guesses our 'intent' from internal actions, and provides reasons and justifications for actions, which may be entirely specious.

For example, direct brain stimulation does not 'feel' like an external input - it feels like a 'natural' thought that you had - and people will often rationalise reasons for the most unusual behaviour due to direct brain stimulation, rather than the simple answer 'you applied a pulse of electricity to my brain' - because that's not how it feels.

http://brainsciencepodcast.com... - is interesting on this exact topic.

Re:Bad syllogism

[sillybilly]

Consciousness is just a state of a computer system, with all memories loaded, and all processing registers at a certain state. A lot of consciousness evolved from processing visual information, or even just smell and touch information, recognizing objects and elements in the environment around the conscious being, and then deriving rules by which these objects behave, and predicting the future behavior of these objects from these learned rules. (For instance a new born baby does not know that when an object disappears behind another object it still "exists", and it did not vanish from reality, it takes a certain amount of learning to know that; a grown cat on the other hand knows such a thing and will go investigate the disappearance of a laserpointer dot around a corner, assuming it still exists, and expecting it to be present behind the corner. As far as the child is concerned, once the laser dot went around the corner and is no longer visible, it's like it's been turned off at the source, and it's no longer lit, no longer exists, which could be the case after sufficient similar experiences.) That is how a conscious being, like a cat, or even an earthworm interacts with its environment, it "computes" its environment, it assesses it's present state, or "millieu", the situation around itself (whether an earthworm is on the surface, during rain or drought, or underground, during cold or heat, or you, whether you're in a natural reservation, in a nightclub dancing in a lit city, on a bus full of people, deep in a dark basement without lights, or even a virtual reality, such as a video game like FreeDoom, deep inside one of the hellhole rooms), and after assessing its present state and environment, the conscious being computes what to do next based on objectives, (some of which are involuntary, like involuntary reflexes, but those don't really belong in the realm of consciousness), but sexdrive, hunger, comfort, comfort for my baby and my other peeps, etc. there are many impulses on which to act. You could say a completely impulseless conscious AI robot is a nonsense, but when dealing with AI, that's the things to watch, what are the impulses, survival, hunger, comfort, etc, which would come up even with weapons design, i.e. the weapon has to optimize its energy consumption and seek out energy sources to recharge = hunger, limit its exposure to temperature fluctuation to be minimize structural degradation = comfort, reproductive drive in case it reproduces sexually instead of asexually (banana's biggest problem and vulnerability is no sex life, earthworms can reproduce asexually lacking any sex partners, but they prefer sexual reproduction to create a greater variety of mutations and genetic variability) - a lot of these things are dangerous when it comes to AI and robots, and it's a matter of degree - like an iRobot Roomba vacuum cleaner that goes around the house autonomously seeking out the electric socket for recharging, that's not such a big deal, right? It is hungry for electricity but not dangerous. In a sense, even it, is conscious, it assesses its environment, it computes its present state variables, and the acts on the objectives it has preprogrammed. It's not a weapon, it's a tool, to keep the house clean, so when is a tool considered a weapon? There is a vague blurring line there too, your fingers can be used as a weapon, or a pocket knife can be used to eat, cut cardboard boxes open, or as a weapon to cut someone else, and there is probably a way to use the Roomba too as a weapon.

Re:Bad syllogism

[zdepthcharge]

Does the drug work on the memory or on the machinery that processes the memory? Do people with eidetic memory have better memories or better machinery for processing memories?

Re:Bad syllogism

[Dr Max]

It's like saying I can build a flat top bridge on top of support columns, but when I use the same system to simulate a suspension bridge, apparently there isn't enough support columns, thus suspension bridges are impossible, except of course in natural examples like spider webs.

Re:Bad syllogism

[Dr Max]

Logical reconstruction. you don't remember everything, because you don't have to. The brain is powerful enough to figure out most of the details (most defiantly helped by a home movie) and fits in what little you do remember to the virtual memory.

Re:Bad syllogism

[narcc]

There is some evidence that consciousness is an illusion

Lol!

I hate to break it to you, but if I'm fooled in to believing I'm conscious, then I'm conscious!

This is what happens when you parrot a meme without understanding it. The nonsense you're looking for is "There is some evidence that free-will is an illusion". It's still nonsense, of course, just not as immediately laughable as the above.

Re:Bad syllogism

[TheLink]

. There is some evidence that consciousness is an illusion, and that people make decisions unconsciously

Freewill may be an illusion, but consciousness is NOT an illusion, at least to me. In fact it is the ONLY THING that I can be sure of! Everything else may be illusion (as philosophers of old have said). Maybe you or those mathematicians do not experience consciousness and that's why you all can even say what you say.

Merely behaving as if you are conscious is different from _experiencing_ consciousness. Anyone who thinks there is no difference or it's not an important difference misses the important part of consciousness.

Merely behaving as if you like chocolate is different from _subjectively_ experiencing and enjoying the taste of chocolate.

Mathematicians and programmers with our current knowledge and tech can write algorithms/programs to simulate the mere behaviour but how about the experience? Can we create an algorithm that truly experiences "self", "chocolate"?

Are there some special yet unknown laws in this universe that would cause this to happen? Would merely performing a suitable advanced algorithm with pencil and paper generate a new consciousness phenomena and create a new entity? If so where in time and space would that consciousness reside- the pencil, the paper? Or does it need some extra "nonnewtonian" stuff? Do we need something like a multi-worlds quantum simulator recursively trying to predict itself and the future?

Bullshit like what these mathematicians spout is more likely to take us further away than closer to understanding consciousness.

Re:Bad syllogism

[wierd_w]

humans forget things with use, as a part of refining the useful portion of a memory.

Unimportant/less important information in the memory is replaced with something like a pointer, or abstract symbol that must be reconstructed from the clues presented by the preserved "important" bits.

Can I prove this? Yes.

Simple cognitive experiment: Dream journal.

Upon waking in the morning, write linearly what you REMEMBER about the dream. write down what you remember first, in the order you remember it.

You will find that memories of the dream appear to be "backwards" in time, with the "last in, first out" type of recall. Really, you recall the "most important" bit first, typically the "conclusion" of the dream sequence, followed by the supportive events that led to that conclusion, which come from a conscious reconstruction process after the fact.

The same is also true of remembering past experiences.

Experiment: Record the process of remembering an important event. (same as a dream journal.)

In order of recall, write what you remember about the first time you had sex. (hold the jokes kiddies.)

I will bet you money that the first thing you write is "place", THEN person, THEN situational setting. In THAT order. (With the most atrophy in the memory occuring in that last portion of the memory-- Just try extending your recall to the events before, and see where it breaks.)

These are falsifiable predictions-- You can actually do these little experiments, and see if I am wrong.

The model presented by these mathematicians presumes that human memory is lossless, and that it does not have any kind of pruning that happens as a result of building against the memory. These two simple experiments clearly show that this presupposition is false.

This shows that their conclusion is false, as it is based on a false precondition.

Re:Bad syllogism

[w0mprat]

Baloney. What a stupid argument. Here is it, summarized: 1. Here is one mathematical model of a way that memories could work. 2. This method would be computable. 3. But that would mean memories degrade the more you remember them 4. But memories don't degrade the more you remember them. 5. Therefore memories are not computable.

Assignment for the student: find the flaw in this argument.

On point 3. is this not how pretty much any volatile memory in any computer biological or artificial works? Our brains are not exempt and seemingly have to re-inforce/re-write memories as accessing them probably does indeed cause them to degrade.

Re:Bad syllogism

[bingoUV]

Freewill may be an illusion, but consciousness is NOT an illusion, at least to me. In fact it is the ONLY THING that I can be sure of!

The question is not about whether one's own consciousness is an illusion. But that how can you tell that the computer someone has built is actually conscious, or just behaves that way.

So all your "knowledge" about your own consciousness does zilch to tell you that what it is to call someone else "conscious". Even if you are very close to a person, you cannot tell for sure whether they are conscious or just behaving that way. As a scientific definition, the "feeling" that one is conscious is useless. As is the epistemological truism that one's consciousness is the only certain thing in the world.

Re:Bad syllogism

[AmazinglySmooth]

You can, however, blame ignorant fucktards who don't understand the data OR the theory who go around acting like self-righteous assholes when a scientific theory intrudes on their ideological leanings.

FTFY: You can, however, blame ignorant fucktards who don't understand the data OR the theory who go around acting like self-righteous assholes when a scientific theory bolsters their ideological leanings.

Re:Bad syllogism

[sillybilly]

Indubitably. Also cucumbers. Both make a lot of women happy. Hi ^5 to that!

Re:Bad syllogism

[RespekMyAthorati]

You need to learn how to use the "br" html syntax in /. posts.
It's right underneath the comment entry box.

Re:Bad syllogism

[RespekMyAthorati]

And if a learned man tells you that he can't remember if something is definitely possible ... how does that go again?

Re:Bad syllogism

[Maxo-Texas]

It is odd because I do have one memory from when I was 13.

I decided I would remember that moment and I paid attention to it fiercely.

And I can still remember it clearly to this day. And as far as I can tell, it's the original memory.

Re:Bad syllogism

[queazocotal]

http://www.ncbi.nlm.nih.gov/pu...
'blocking facial afference with botox selectively slows comprehension of emotional sentences. Therefore, theories of cognition should account for emotion-language interactions above the level of explicit emotion words, and the role of peripheral feedback in comprehension. '

Sounds Non-Deterministic to me

[Billy the Mountain]

Isn't noncomputable the same as saying non-deterministic? There are lots of non-deterministic computer operations where the result is based on a database query or a call to a web page where you can't know in advance what the result will be and you also don't really know how long it will take to get the information (if at all).

Re:Math

[Russ1642]

Exactly. This is like the intelligent design argument. "This problem is complex, so we're going to propose that it can't EVER be solved. Let's discuss where we're all going for lunch."

A good analogy

[BenSchuarmer]

Talking about whether a computer can think is like talking about whether a submarine can swim.

Trying to duplicate the mechanical details may be a waste of time. The fact that we can't duplicate the mechanical details today doesn't mean we never will.

worse than physicists

[joss]

> By assuming that the process of memory is non-lossy

What a fucking strange way to start. Memories are recursive, really old memor s you don't directly remember, you remember remembering.

Re:worse than physicists

[jfengel]

With an oversimplifying assumption that bad, I assumed they were physicists of the "why does your field need a journal anyway" school. Nope, turns out this time it's two computer scientists, somebody in "computation and neural systems", and a business school (!).

Memory Non-Lossy? I beg to differ.

[Ken Broadfoot]

"But they point out that this cannot be how real memory works; otherwise, retrieving memories repeatedly would cause them to gradually decay. By assuming that the process of memory is non-lossy."

Really? I can barely remember last friday night. Let alone my circumcision 50 years ago. What was that girl's name who slapped me in my face? Or punched me... it's so hazy.... Caroline? Katy? Maybe it was Jeffery..... so fuzzy.... I had her number written on my hand.... oops right palm....

Memory non-lossy my ass...

Re:Memory Non-Lossy? I beg to differ.

[ArcadeMan]

Best memory loss example, ever.

misapplied mathematics

[Mr. Slippery]

One of the most profound advances in bullshitting in recent years is the way researchers from a variety of fields are beginning to misuse mathematical terms in order to give their ideas a facade of intellectual responsibility. Since no one has yet come up with an agreed-upon definition of what this "consciousness" is as an objective observable phenomenon, trying to talk about it in mathematical terms is nothing more than intellectual masturbation.

Re:misapplied mathematics

[Prien715]

intellectual masturbation.

The one thing you can't find on pornhub!

Re:misapplied mathematics

[timeOday]

After making an incorrect claim of uniqueness, the article does come close to admitting the point:

What makes Tononi's ideas different from other theories of consciousness is that it can be modelled mathematically using ideas from physics and information theory. That doesn't mean this theory is correct.

Unfortunately it then continues into la-la land:

But it does mean that, for the first time, neuroscientists, biologists physicists and anybody else can all reason about consciousness using the universal language of science: mathematics.

What a stupid, meaningless statement. I can posit the equation 1+1=2 as a model of consciousness that allows us to reason about it. The claim in the title of the article: "Proves Human Experience Cannot Be Modelled On A Computer" is just sad gibberish.

"Consciousness."

[Anonymous Coward]

No, we're absolutely nowhere toward understanding consciousness. How we think and how we retrieve memory, perhaps - but our ability to experience our own existence is not understood at all by science.

Illogical Distinction

[bhagwad]

How is the brain not a computer? Pfft...ridiculous conclusions.

Consciousness is irrational

[Anonymous Coward]

Cognitive dissonance ALONE would defy mathematical understanding.

better than human

[DanielWilliams6483]

this is based on a faulty assumption. humans make up portions of their memories in the first place. memories Do Degrade, And Warp, And morph Over Time.

Retrieving memories causes decay?

[Verdatum]

"retrieving them repeatedly would cause them to gradually decay"

Ouch. Just. Ouch. No. Noooo. NOOOOO.

There is so much wrong with this statement I don't even know where to start. It implies that the memory is overwritten with the memory of recalling the memory, which is a huge and ridiculous assumption. Memory likely works much more like ant paths. The details that are recalled more frequently are reinforced, and can be remembered longer. It could also be compared to a caching algorithm; details used more often are less likely to be lost, or need fewer hints to retrieve them.

And then using this assumption to declare something as non-computable demonstrates a lack of understanding of the concept of computability. The only way that conciousness could be non-computable would be if there is a supernatural element to it. Otherwise, the fact that it exists means it must be computable.

Re:Retrieving memories causes decay?

[medv4380]

Irrational Numbers are Non-Computable.

Re:Retrieving memories causes decay?

[VorpalRodent]

Irrational numbers are supernatural.

Re:Retrieving memories causes decay?

[Paxinum]

Actually, no. Almost all numbers that are defined, are defined in a manner that describes an algorithm on how to construct them. Hence, they are computable. You need to brush up your theory on computable real numbers. It is all about the definitions...

Re:Retrieving memories causes decay?

[medv4380]

The Universe is Natural
Irrational Numbers are Supernatural
The Universe is Dependent upon The Cosmological Constant
Pi is an Irrational Number
The Cosmological Constant is Irrational because it contains Pi which is also Irrational.
The Universe Must be Supernatural.

Re:Retrieving memories causes decay?

[NoNonAlphaCharsHere]

The only way that conciousness could be non-computable would be if there is a supernatural element to it.

Roger Penrose (for one) is vehement in his insistence that consciousness is non-computable, possibly quantum in nature. Certainly there are other ways that consciousness could be non-computable without being supernatural.

Re:Retrieving memories causes decay?

[Dimwit]

It's not true that it has to be supernatural to be noncomputable, unless you agree that physics itself is computable. The jury is still out on that one (although I believe that it will turn out to be true).

Re:Retrieving memories causes decay?

[Tyler Durden]

The only way that conciousness could be non-computable would be if there is a supernatural element to it. Otherwise, the fact that it exists means it must be computable.

Nonsense. Just because something exists and is not "supernatural" doesn't mean that it must be computable. Take the halting problem for instance. There is no Turing Machine that is able to take any possible TM and input and determine whether the inputted TM will eventually halt or go into an infinite loop when run with the given input. This is true even though every TM must either halt or go into an infinite loop for any possible input. There's nothing supernatural about it.

Re:Retrieving memories causes decay?

[Stormy Dragon]

And his fascination with that crackpot theory is why he, frankly, hasn't done any significant work in 20 years.

It's based on assuming their exists a new type particle that we have no evidence to currently believe exists interact with a part of the neuron whose functions are known to cause a quantum superposition despite the fact it's been shown there's no way such a state could maintain coherence at anything close to the temperature the brain is at.

Re:Retrieving memories causes decay?

[ausekilis]

"retrieving them repeatedly would cause them to gradually decay"

Memory likely works much more like ant paths. The details that are recalled more frequently are reinforced, and can be remembered longer. It could also be compared to a caching algorithm; details used more often are less likely to be lost, or need fewer hints to retrieve them.

I'd really like to see a reference for this. Not because I disagree with your analogy, after all it's the basis for education and classical conditioning. It's a fair assumption that certain tasks such as facial recognition and memory recollection can be associated with certain regions of the brain. However, we still don't know how we go from synapses firing to midget wrestling. Looking at it from another direction, we don't regrow brain cells, they don't change in size or form like a popular anthill path may become stronger via compaction of soil or wider to accomodate more ants. We don't know specifically if a neuron has a "firing limit", or otherwise may wear out over time. At least, in my limited research I've never come across such studies.

And then using this assumption to declare something as non-computable demonstrates a lack of understanding of the concept of computability. The only way that conciousness could be non-computable would be if there is a supernatural element to it. Otherwise, the fact that it exists means it must be computable.

Agreed. Even the decay of RAM or any sort of storage medium susceptible to decay can be calculated somehow. How else would we have MTBF and expected write limits for Hard Disks, SSD's and such?

Re:Retrieving memories causes decay?

[geekoid]

No, recalling memories cause decay and changes in the memory. This is a well know neurological fact.
The more you recall memory, the more likely it will change.
http://clbb.mgh.harvard.edu/st...

Re:Retrieving memories causes decay?

[Verdatum]

The halting problem exists. A functional solution to the halting problem does not exist. It does not exist because it is not computable.

Re:Retrieving memories causes decay?

[medv4380]

We're talking about the compatibility by a Turning Machine. Which puts Pi into the realm of Non-Computable. It is really a choose your own adventure of definitions, but any number that cannot be represented to an infinite precision is not actually computable. A Turning Machine will always run into a rounding error when it try's to output the value. You can use the other lax definitions that say you can reduce it down to a function that can output the value to the "desired" precision, but that's hardly infinite precision when you look at the output.

Re:Retrieving memories causes decay?

[Verdatum]

I am well aware that is the definition of computable. Neither you nor TFA has shown why consciousness is not computable. The only way that conciousness can be non-computable is if the brain does something outside of the realm of Von Newman Architecture in the abstract sense. It receives input in the form of electrical and chemical stimulation, it is able to store information in the form of memory (however that memory works doesn't particularly matter) it performs operations based on memory and input, and it produces output in the form of electrical stimuli and neurotransmitters. If the brain follows Von Newman architecture, then it is a computer that produces conciousness, there for it is computable. Proof by counterexample.

Re:Retrieving memories causes decay?

[Verdatum]

I did not assert that there is no supernatural element. I stated clearly that it was the only alternative that would fulfill the requirements. TFA did not invoke the supernatural, so it did not make a proper argument. If the brain does have a supernatural element, that would be extremely exciting to prove. So if you can eliminate the impossible and have supernatural influence being the only remaining solution, concretely, that's some nobel winning shit right there. I'll keep an eye out for such a discovery on /.

Re:Retrieving memories causes decay?

[Verdatum]

Yes, I read about this study, and did not find it surprising, and it does raise some interesting points. I thought about bringing them up, but it gets into a whole bunch of "yeah but" things. Memory can be modified, by more recent experiences, one of those experiences being a recollection of said memory, or by physical or chemical influence, beat on the head, drugs, stroke. But the degradation is only in certain aspects. For example, if you memorize the pledge of allegiance, you aren't going to start fucking it up by reciting it more and more.

Re:Retrieving memories causes decay?

[colinrichardday]

What algorithm do we have for computing pi? Hint: Infinite series are not computable.

Re:Retrieving memories causes decay?

[hey!]

It implies that the memory is overwritten with the memory of recalling the memory, which is a huge and ridiculous assumption.

However the notion that memory is overwritten by recollection actually does have experimental support. The idea isn't ridiculous, it's just repugnant because it implies that our grasp on reality isn't as firm as we'd like to believe it is.

The only way that conciousness could be non-computable would be if there is a supernatural element to it. Otherwise, the fact that it exists means it must be computable.

Not necessarily. One way consciousness could be non-computable would be for it to be non-deterministic.

In any case this is all fuzzy; not only is "supernatural" a fuzzy word, the discussion of "computable" is fuzzy too. What would it mean for consciousness to be "computable"? Is a Turing machine "computable"? Well it can be *simulated*, but aspects of its behavior cannot always be *predicted* (e.g., halting).

Re:Retrieving memories causes decay?

[Tyler Durden]

So then why can we not say, "The consciousness problem exists. A functional solution to consciousness does not exist. It does not exist because it is not computable."?

For what it's worth I also believe that the mind, including consciousness, is completely the result of what is functionally a computer. But I don't think it's possible to identify the essential qualities of what makes a working algorithm yield consciousness. That isn't to say that we can't come up with some self-learning machines that could yield consciousness. But we may still not be able to identify what it is that makes some of these achieve consciousness and not others.

Re:Retrieving memories causes decay?

[Verdatum]

You don't need to identify or enumerate the algorithm. You only need to show that a working algorithm exists. A functional solution does exist. If it didn't, consciousness wouldn't be a reality, it would just be a nice idea. And we're not debating the existence of conciousness, because descartes did the whole cogito ergo sum thing.

Re:Retrieving memories causes decay?

[Verdatum]

A turing machine is computable, and it can always be predicted, it merely requires all of the exact same input that it receives. In other words, to mimic a "true" random number generator, you just need a copy of its entropy. It's computable because the turing machine computed it. This is what computable means, it's a very formal term in the theory of Computer Science.

Re:Retrieving memories causes decay?

[Verdatum]

I completely see where you're coming from. But I'm afraid I pulled the analogies out of my nebulous ass. It's very likely that I heard the ant one elsewhere, and the cache one is just sorta obvious. If I did hear it, it was probably on one of those horrible overdigested shows that often fails to make a distinction between well reviewed science and psuedoscience statistical fallacy bullshit (e.g. Through the Wormhole with Morgan Freeman).

I constantly marvel at how much we can't verifiably prove we know about the nature of the brain. I'm optimistic we're going to learn things steadily, but compared to so many other things, good lord, we've got a long way to go.

Re:Retrieving memories causes decay?

[Verdatum]

I mean, if you'd like to switch to a completely argument, regarding peer reviewed articles on the concept of near-death experiences as evidence of the supernatural (and I'm not sure you are. Your wording confuses me some) Near death experiences exist, and there is nothing wrong with discussing them in peer reviewed publications. They don't provide evidence for supernatural phenomena though. Show me one of those "I was floating above the room" stories where they do something like win a game of win lose or draw while the patient is blindfolded in a double blind scenario, and I'll change my tune.

But this is just the off-topic discussion of the nature of skepticism. If the only "evidence" is a leap that can only be made sense of by the reader accepting an implicit intervention by supernatural forces, then you've not written a good paper. Or at least, as TFA's intro hints at, not one that is appropriate for the realm of science.

Re:Retrieving memories causes decay?

[hey!]

A Turing machine *performs* computations. It can be *simulated*. But it's behavior cannot be *computed* by any algorithm running in a finite amount of time.

Re:Retrieving memories causes decay?

[colinrichardday]

Hint: Pi is a computable number.

Only under a certain use of "computable". Will we have arbitrarily long periods of time to get sufficiently good approximations?

Re:Retrieving memories causes decay?

[colinrichardday]

Is Turing computability sufficient for your purposes?

Re:Physically impossible

[drxenos]

No. All Turing complete means is a universe Turing machine can execute anything any other Turing state machine can. People misunderstand "Turing complete" can think it means someone that is Turing complete can do "anything." That is NOT what it means.

I thought memories do decay

[TsuruchiBrian]

That allows memories to be retrieved but it also loses information in the process. But they point out that this cannot be how real memory works; otherwise, retrieving memories repeatedly would cause them to gradually decay.

I remember hearing a radiolab episode on NPR talking about how memories actually get modified every time you recall them.

http://www.radiolab.org/story/91569-memory-and-forgetting/

Maybe the radiolab episode is completely wrong, but I don't think it's fair to assume memories are lossless without providing some evidence of this.

Re:A call for help...

[ArcadeMan]

There's no Enya music in the background.

The halting problem is a counterexample

[tepples]

Everything is computable given the right models and starting conditions.

"Does the Turing machine with a given description halt?" That's been proven not computable on a Turing machine. And we lack a model more powerful than a Turing machine.

Re:The halting problem is a counterexample

[amorsen]

Only because Turing machines have infinite memories, and therefore no actual Turing machines exist. Given an actual Turing machine and a bit of patience, you can solve halting for all physically existing machines.

Intuitively you would expect the difficulty of solving the halting problem to be exponential to the size of the storage of the computer you are trying to solve halting for, but I am not sure whether that has been proven.

Re:Memory is non-lossy? Research suggests otherwis

[frog_strat]

Hmmm. Do you find yourself occasionally having to re-learn your address or phone number ?

otherwise, retrieving memories repeatedly would c

[allo]

> otherwise, retrieving memories repeatedly would cause them to gradually decay

So, i guess this was never observed on real humans?

a bunch of silly assumptions

[gurps_npc]

First, I do agree with the result - that consciousness is not definable via mathematical equations and algorithms.

That said:

1) Most memory researchers believe it IS lossy. Specifically each time you access a memory you change it, losing original information

2) Not all computers have to only use mathematical equations and algorithms. Specifically their are quantum computers that do not work that way. While I am not an expert on such things I highly doubt that the rather limited definition they are using for 'computer' includes all things we would consider a computer.

Sounds like complete bullshit.

[eddy]

There seems to be a step missing from A (that's not how memory works) to B (therefore uncomputable). The premise that memory isn't lossy sounds like rubbish, even IF it's perhaps not so simply a question of 'read errors'

I recently watched this talk, Modeling Data Streams Using Sparse Distributed Representations, which seems to be able to represent memory in a layered and lossy way perfectly fine in a computer.

Memories do decay upon recall

[MobyDisk]

Memories decay upon recall. Your brain basically alters the memory slightly each time. This can be used to erase or alter memories.

Re:Memories do decay upon recall

[MaWeiTao]

Is it that retrieval alters the memory or your processing of the retrieved memory that alters it? That's an important distinction.

I would propose that the original memory isn't altered at all, but that new experiences and thoughts get layered on top of it. The original memory, when retrieved gets inextricably tied to all that and thus distorted. If you were able to store a memory that is unlikely to be tainted in this manner, unique smells and flavors come to mind, I imagine when it's retrieved years later it wouldn't be distorted at all.

Re:Memories do decay upon recall

[Antique Geekmeister]

> Is it that retrieval alters the memory or your processing of the retrieved memory that alters it? That's an important distinction.

"Yes".

It's not feasible to recall a memory without processing it in some way.

No need for math model

[Rosco P. Coltrane]

As always, the truth is in the Bible:

Genesis 1:27

God created man in His own image, in the image of God He created him; male and female He created them; but man is not a machine, for God did not look like a beige box PC.

Re:No need for math model

[firewrought]

As always, the truth is in the Bible

Yep, this speciously-reasoned physics paper supports the Absolute Truth(TM) revealed to us in an ancient text of unknown authorship!

Re:No need for math model

[RespekMyAthorati]

Right.
He looks like a Mac, obviously.

Re:lossy memory

[biodata]

This. Memories are changed by recall.

singularity

[rogoshen1]

If this is true, what does that mean for wankers like Kurzweil and the fantasy of the 'Singularity'?

Re:Math

[hubie]

I thought they recalled all you defective Barbies in the 90's.

Sounds like utter bullshit

[Animats]

Here's a a critique. (It's on arxiv; no need to sign up for "Medium")

The paper isn't impressive. It make the assumption that human (other mammals, too?) memory isn't compressed, and is somehow "integrated" with all other information. We've been through this before. Last time, the buzzword was "holographic". We've been here before.

The observation that damage to part of the brain may not result in the loss of specific memories still seems to confuse many philosophers and neurologists. That shouldn't be mysterious at this point. A compact disk has the same property. You can obscure a sizeable area on a CD without making it unreadable. There's redundancy in the data, but it's a lot less than 2x redundant. The combination of error correction and spreading codes allows any small part of the disk to be unreadable without losing any data. (Read up on how CDs work if you don't know this. It's quite clever. First mass-market application of really good error correction.)

Re:Sounds like utter bullshit

[neurophil12]

Don't get the specific research cited in the article confused with Integrated Information Theory (IIT). IIT, or a possible future more accurate/complete version, could well be true without the rubbish article assuming brain member is non-lossy. Thinking about consciousness in terms of information is nice because it comports with the evidence that when you change something about the brain or its inputs (i.e. change the information content of the system) you get a change in cognition. The details of the theory are complicated, but worth reading up on. I find the basic concept of IIT intuitive, but the math is pretty intense.

"By assuming..."

[mmell]

"By assuming that the process of memory is non-lossy..."

Terrible assumption.

Anecdotal evidence that individuals may be capable of accurate recall is directly contradicted by evidence that even witnesses who are absolutely certain of what they saw in fact only recalled those specific items which somehow drew their attention at the time of an event.

Is that what it proves?

[nitehawk214]

Sounds more like you can't separate the human consciousness from the memories. I thought we already knew that. Perhaps there was a theory why until now.

Conscious phenomenon != complex processing

[Impy the Impiuos Imp]

Stop it. Just stop it, people.

Memory doesn't work that way. It's a live feedback loop that reinforces itself through the conscious mind. There is some lossy drift but stuff that maps to the real world is indeed corrected if lossily. Ancient stuff from when you were a kid (Gee, what did Koogle taste like) drifts and drifts.

Something from when you were a kid,
like Orange Julius taste, drifts but may suddenly be reset when you stumble across one at a mall somewhere (or Dairy Queen, whoever bought them). His model is a solution to a problem thatsn't a problem. It doesn't matter how clumsily intertwined actual brain processes are for this.

Furthermore, he conflates consciousness with deep thought. I could grant his proposition of complexity yet it would not matter one bit for the subjective conscious experience. The subjective perceptual experience may still be magic w.r.t. grounding in real physics, but it is there and not some.purely informational process (i.e. Searle is still undefeated) and there is nothing requiring consciousness to be synonymous with all this complicated brain activity.

Your unconscious mind does the vast bulk of difficult processing, then passes it through consciousness for some kind of review.

There is no evidence consciousness, however miraculous and awesome, need be particularly complicated in and of itself, nor is "what it does" as part of your larger, largely subconscious thought process, particularly valuable.

From an importance point of view, it is vastly overrated as information processor.

Your thinking, in other words, could be supra-Turing in computational model, yet the consciousness itself perfectly mundane, experencing these supra-Tuting-generated thoughts and doing a computationally mundane thing with them.

What do physicists say about this?

[pele]

Wouldn't this mean their search for a grand unified theory is now - for all intents and purposes - over?!
A single equation should be able to define the universe, including every single one of our feeblee minds in it.
42 to this I say...

no Ghost_no "singularity"_only sci-fi

[globaljustin]

And after you include however-many extraterrestrial intelligences there might be, all across the Universe, well,

Then you're in science fiction land...woo hoo! I like scifi as much as the next /.er but your imaginations of the possible existence of a civilization that can fully digitize continuous data is worthless to a **scientific discussion**

That's the problem. Hard AI, "teh singularity", and the "question of consciousness" are so polluted in the literature by non-tech philosophers throughout history that the notion of ***falsifiability*** of computation theory get's tossed aside in favor of TED-talk style bullshit.

Falsifiability kills these theories *every time* and hopefully this research in TFA will help break the cycle.

To be science it must be able to be tested. It must be a premise that is capable of being proven or disproven. "hard AI" proponents like Kurzweil and the "singularity" believers ignore this part of science.

So happy to see this research

Re:no Ghost_no "singularity"_only sci-fi

[xevioso]

I hereby bestow upon you a Ph.D. in Pedantry.

Re:no Ghost_no "singularity"_only sci-fi

[thedonger]

We'll never make a truly human computer (or maybe "natural computer" is a better term) because we can't make it first and foremost desire self-preservation. We can build a robot that plays catch, but we can't make it want to play catch. Do we even know why we want to play catch (deep down I think it is motivated by the desire to procreate)? And thank FSM we can't build and evolving machines, because computers are logical and not forgetful, and would very likely enslave us first change they got.

Re:no Ghost_no "singularity"_only sci-fi

[roc97007]

I'm not sure I agree. I think building an OS with virus checking incorporated into the design, for instance, would be a form of "self preservation". Or a computer/robotic arm combination that recognizes a screwdriver and will not let one get near. Moreover, I could point out humans that don't appear to have any concept of self-preservation, which calls into question whether this would be a rigorous requirement for a "truly human computer".

Likewise, a robot that nudged you and said "let's play catch. Please please please" until you wanted to unplug it, would be a pretty good facsimile of a typical five year old kid.

Re:no Ghost_no "singularity"_only sci-fi

[UnknownSoldier]

> Then you're in science fiction land...woo hoo! I like scifi as much as the next /.er but your imaginations of the possible existence of a civilization that can fully digitize continuous data is worthless to a **scientific discussion**

You're proof of a paradigm shift will come in 2024. Human consciousness is only one of the many different types.

The OP is not letting on to how much he knows about consciousness however they are right about extraterrestrial intelligences.

> To be science it must be able to be tested.

And you assume consciousness passes exactly what test again?

Scientists are still ignorant of the lower 3 types of consciousness; what makes you assume they have a clue on how to test for a higher level one???

Re:no Ghost_no "singularity"_only sci-fi

[easyTree]

Can a Ph.D. be bestowed by an individual? :P

Re:no Ghost_no "singularity"_only sci-fi

[Xaedalus]

Pray tell what are you talking about, in regards to multiple different types of intelligence? And what does 2024 have to do with anything?

Re:no Ghost_no "singularity"_only sci-fi

[TapeCutter]

So only mathematicians should ponder and discuss their own mind? In case you hadn't noticed every field of science is polluted by woo and the less science knows about a subject the more woo there is. Besides I like TED talks, they are not trying to replicate a university lecture hall, they are providing a platform for an interesting and informative "show and tell" for adults, not a great deal different to the aims of Slashdot but a lot better at filtering out trolls.

Re:no Ghost_no "singularity"_only sci-fi

[Bing Tsher E]

Oh, come on. Often enough the troll at a TED talk is the person front-and-center on the stage. Certainly not always, I would add, but often enough.

Re:no Ghost_no "singularity"_only sci-fi

[Antique Geekmeister]

> We'll never make a truly human computer (or maybe "natural computer" is a better term) because we can't make it first and foremost desire self-preservation.

Any sophisticated system that does not protect its own assets will perish quite quickly under any kind of stress. "Self preservation" of a a _species_ is at the core of most evolutionary science. A casual search shows dozens of quite good papers on the nature of such evolution of of "self-preservation" as a fundamental and natural part of AI.

Re:no Ghost_no "singularity"_only sci-fi

[VernonNemitz]

You missed the point, regarding E.T.s --each such species (even if only one per hundred galaxies) represents yet another way that Nature would have found to build a brain that can host enough consciousness for self-recognition. With hundreds of billions of galaxies out there, one such type of brain per hundred galaxies would mean there are billions of ways to build such brains. And so I repeat, anyone who thinks "no variant of computer hardware will ever be able to do that" --especially when we are deliberately copying brain-designs into our computer hardware!-- is just not thinking clearly.

Re:no Ghost_no "singularity"_only sci-fi

[AthanasiusKircher]

Can a Ph.D. be bestowed by an individual? :P

Your comment presupposes that a "Ph.D. in Pedantry" exists. If such a degree did exist, I'm sure people many people around here would have attained one (if not granted multiple honorary doctorates).

Perhaps this calls for a new Slashdot achievement -- the Ph.D. in Pedantry. Once one achieves it, one gains the ability to mod posts as "pedantic" (since someone with a Ph.D. is obviously an official arbiter in the field). The fun thing about the "pedantic" mod is that it could serve as either +1 or -1, which would make it extra fun.

Accumulate enough posts modded "pedantic," and the Slashdot userbase will confer upon you a "Ph.D. in Pedantry." :)

Re:no Ghost_no "singularity"_only sci-fi

[ultranova]

Then you're in science fiction land...woo hoo! I like scifi as much as the next /.er but your imaginations of the possible existence of a civilization that can fully digitize continuous data is worthless to a **scientific discussion**

To put it bluntly, this entire study is worthless as science. We don't know how human mind works. Should we ever know, we'd then have the oh so fun task of disentangling accidents of biology from fundamental underlaying limits. And because we don't know how the human mind works, we have no way of knowing whether a particular model presents it accurately or at all (however, any theory that claims human memory is in any way perfect is certainly off to a bad start), thus any conclusions based on it are firmly in the land of wild mass guessing.

To be science it must be able to be tested. It must be a premise that is capable of being proven or disproven. "hard AI" proponents like Kurzweil and the "singularity" believers ignore this part of science.

Well, the complexity of behaviour of the Universe has been increasing since at least the Big Bang in a virtuous circle. Is there some reason why the trend would stop, either now or at some future point? If not, then it seems like singularity would be the inevitable result.

Anti-AI isn't science, it's just the ancient belief about the supernatural specialness of human soul, typically dressed in arguments from lack of imagination and often seasoned with a helping of ego. Nature has no way of telling between "artificial" and "natural", after all, so it's incapable of allowing natural intelligent creatures (us) yet disallowing artificial ones.

Re:no Ghost_no "singularity"_only sci-fi

[narcc]

"woo"?

Nothing is funnier than a Randian complaining about the quality and integrity of a public forum.

Re:no Ghost_no "singularity"_only sci-fi

[narcc]

What's the big deal? This isn't anything we didn't already know. You can disagree with how these guys came to their conclusion, but it's the exact same conclusion others, by different means, reached decades ago: computation is insufficient.

Religious nuts like Ray Kurzweil may not like it, but it's not the researchers fault that reality doesn't validate your superstitious beliefs.

We don't know how human mind works.

That's right. We don't. Of course, we don't need to know how it works in order to identify what does not.

Stop living in a fantasy land and learn to embrace reality. It's much more interesting than Kurzweil's video-game afterlife and Spielberg's sex-bots.

Re:no Ghost_no "singularity"_only sci-fi

[Anonymous Coward]

> Often enough the troll at a TED talk is the person front-and-center on the stage.

TED presenters are usually pretty good, and knowledgeable in their field

TEDx, on the other hand, is often rubbish.

Re:no Ghost_no "singularity"_only sci-fi

[Jane Q. Public]

Then you're in science fiction land...woo hoo! I like scifi as much as the next /.er but your imaginations of the possible existence of a civilization that can fully digitize continuous data is worthless to a **scientific discussion**

That's the problem. Hard AI, "teh singularity", and the "question of consciousness" are so polluted in the literature by non-tech philosophers throughout history that the notion of ***falsifiability*** of computation theory get's tossed aside in favor of TED-talk style bullshit.

Uh... excuse me? Why are you ranting on about something GP never even said?

Here's some "falsifiability" for you: repeatable experiments have been done on these different creatures, and a subset of species DO in fact exhibit self-recognition in controlled studies. Now, it may be only an assumption, but it is a pretty damned good assumption, that self-recognition is a precursor to consciousness. (It is actually more than just an assumption; but we have only one example of a "conscious" brain so it's hard to make comparisons.)

GP was simply saying that we have what appear to be other examples, or at least similar examples, which suggests that there is no reason to believe that what the human brain does is unique in the universe.

What's wrong with that? That's not science fiction. It might be speculation but it's based on falsifiable evidence.

Re:no Ghost_no "singularity"_only sci-fi

[Your.Master]

His comment presupposes no such thing. It simply asks a question without any reference whatsoever to pedantry. You just assumed the reference from context.

(I can haz Ph.D. naow?)

Re:no Ghost_no "singularity"_only sci-fi

[Bob_Who]

Accumulate posts modded "pedantic," the Slashdot userbase will confer upon you a "Ph.D. in Pedantry." :)

Does that make your advocates "Pedantophiles?

Re:no Ghost_no "singularity"_only sci-fi

[UnknownSoldier]

Sure. There is 7 different types of consciousness - 3 below humans and 3 above human. This is no accident. I will leave it to you to also experience the joy of (re)discovery of what they are. Your life will never be the same again. Lucid Dreaming is a great spring board to Meditation which is usually the fastest way, but use what ever techniques and religion(s) that you feel help.

By 2024 first (public) contact will happen. Technically, contact already happened thousands of years ago but on a limited level so as not to interfere with the mass level of free will and create mass panic. Mankind is finally spiritually mature enough to handle the truth of our past and learn about our glorious future:

Eventually we will be allowed to know that we are the spiritual children of the Pleiadians. Ever parent wants to know how their children are; someday we will do the same and be "aliens" to our legacy.

The fact that we are not alone (which by itself is freaking amazing!) will be completely over shadowed by the fact that our "parents" are humanoid too! There is a reason we have 5 fingers (which is technically 1+4 working together); consciousness has a plan on how it evolves itself.

2024 is extremely important -- it is the year men learn to stop living in fear -- of others and of themselves. We are greatly loved by the universe in spite of our myopic greed and xenophobia.

Our future is one of unlimited joy as we discovery how to properly balance and blend Science and Spirituality (the astute reader will notice that Science is amoral by definition - hence part of the problem of being an incomplete philosophy); some have dubbed us in our golden age as "homo spiritus" - the title is appropriate.

From a limited perspective this all sounds crazy; but then again all the great paradigm shifts sounded that way - at first.

When you see news about the Quantum Energy exchange between white holes and black holes, and the 2 missing fundamental forces, you will have further proof that time (and meta-knowledge) can be transcended.

It is time to spiritually wake up.

Re:no Ghost_no "singularity"_only sci-fi

[h5inz]

Maybe if you read the article that the Slashdot summary refers to, you would also read a little about the criticism about the claim that human consciousness is not computable. The criticism pretty much destroys this and it appears that the "definite proof of non-computability" translates to "err..me thinks its not possible for humans to make it" (to mimic the function of a human brain). Maybe I should point some out:

*The neural network of human brain can be atomized down to neurons and their connections. It is not magic. Do you believe in magic?

*“Memory functions must be vastly non-lossy, otherwise retrieving them repeatedly would cause them to gradually decay,” What BS is that? Retrieving a memory strengthens it, and encoding a memory in brain is lossy or non-lossy only based on what you consider the information being stored - if we consider everything that the human sense neurons fire then of course most of it gets lost and also the chemical composition of a neuron changes over time which affects the info it contains. You could achieve a non-lossy encoding of couple of numbers through a lots of redundancy.

*"You can't remove the smell of a chocolate from a brain" - why not? You just don't know which neurons exactly contain the information and how to target them exactly but you could erase the memory from your brain by using a shot-gun.

*The cs guy only says that according to his research the human consciousness either does not fully fall under the definition or being perfectly integrated or it does not compute. So the bad journalism automatically picked the second.

* Why does the word science appeal so much to the biggest morons on Slashdot? Ironic, isn't it?

Re:no Ghost_no "singularity"_only sci-fi

[MrNiceguy_KS]

...and this, ladies and gentlemen, is why experimenting with hallucinogens is a bad idea.

Re:no Ghost_no "singularity"_only sci-fi

[fractoid]

I think you've got it backwards. My feeling is that as we build more and more autonomous machines, they will necessarily begin to have something analogous to feelings and desires. Even now, machines 'feel' with sensors, and experience 'pain' or 'alarm' when they detect a fault condition. They 'want' things when their control algorithms attract them to those things.

Any machine designed to explore its surroundings will have to have some system that keeps it roving around looking for things it hasn't seen before - it will be curious if it sees something unfamiliar, and will get bored if it doesn't. A machine designed to protect itself will feel fear when it anticipates damage, and will become defensive and take action to avert that damage. A machine designed to be able to respond violently will become angry and aggressive if provoked.

Humans are just very complicated, evolved machines. Our feelings are effectively just a heuristic for simplifying triggering different behaviours. We only think they're special because it's us that experience them.

Re:no Ghost_no "singularity"_only sci-fi

[ultranova]

You can disagree with how these guys came to their conclusion, but it's the exact same conclusion others, by different means, reached decades ago: computation is insufficient.

Computation is insufficient to solve all problems, yes. The questions are: is anything capable of solving all problems? That is, is there something beyond computation? And if there is, does human mind include it? And if it does, is it something essential or does it just give you an extra edge in some special situations?

That's right. We don't. Of course, we don't need to know how it works in order to identify what does not.

So far, no one has demonstrated any ability of human mind that couldn't be replicated through computation. That, of course, doesn't mean none exists. Knowing how mind works would would presumably allow us to enumerate over all its capabilities and settle the matter.

Stop living in a fantasy land and learn to embrace reality. It's much more interesting than Kurzweil's video-game afterlife and Spielberg's sex-bots.

And now we're back to meaningless rhetoric.

Re:no Ghost_no "singularity"_only sci-fi

[ultranova]

I think you miss the fact that the Chinese Room argument is firmly rooted in a deep mathematical basis (Entscheidungsproblem)

No, it isn't. It has absolutely nothing to do with the Halting Problem. It has nothing to do with mathematics either. It's crude circular logic: "syntax by itself is neither constitutive of nor sufficient for semantics" is both an axiom and the conclusion of Searle's argument.

Classic bullshit, in other words.

Re:no Ghost_no "singularity"_only sci-fi

[narcc]

So far, no one has demonstrated any ability of human mind that couldn't be replicated through computation.

No one except Searle, Lucas, Block, Fodor, etc.

Re:no Ghost_no "singularity"_only sci-fi

[Xaedalus]

Why do you believe that ascension is the goal of our existence? I'll leave it to others to rebut the specifics of your claims, but I am going to take you on over the imputation that we are going to achieve transcendence and that in doing so all of our problems will be solved and everything should be perfect.

From a philosophical perspective, your points bear similarity to the experiences of immersion into the Godhead that NDE survivors report. But in that state, there's no needs or wants or suffering. Everything's perfect, everything's known, and that means everything's stagnant and static. Perfection is a dead end, especially when time and dimension become completely irrelevant (under the assumption that this is indeed our ultimate fate--which no one can prove or disprove).

Here's where I'm going with this: setting aside rebuttals involving claims of pseudoscience and falsifiable tests for differing levels of intelligence, I believe you're wrong because if we do assume that there is a higher level of existence where transcendence is possible, then the reason we're here is to avoid being perfect, to learn and develop as isolated imperfect beings in ways that cannot be done if we're an ascended and aware portion of an omniscient Godhead. Claims that we're going to ascend on a given date sound great, and in ways do correlate with NDE claims that there is no beginning or end and all things are known, but then why are they always in the future? Why always on specific dates that come and go with no change? Why haven't they already happened, if indeed they're possible? If such mass ascensions are possible, then they've had all of eternity, or at least the last 5000 years of recorded history to happen. My answer is that those claims are delusional, that if there is an omniscient emergent Godhead (of a Spinozan nature, maybe), then we're fragments that have been expelled and will continue to be in order to learn and grow on our own as individuals, to cycle over and over again, within a zoned, defined parameter that we experience as the universe. And when our cycle is complete, we return our data to the source, and then start over again.

What I'm claiming is certainly psuedoscience/mental masturbation based on my own observations and thoughts, and the only way I can personally falsify my hypothesis is by dying. But I just cannot accept that mass enlightenment is due to the entire human race in 2024, or that ascension and enlightenment is the goal of our existence and that we should welcome it. Seems a damn waste of time if it is, since we'll go right back to that state of existence upon death (dependent if there is an afterlife). Rather, to be imperfect would be a more reasonable goal.

Re:no Ghost_no "singularity"_only sci-fi

[RespekMyAthorati]

And their method of demonstrating this:
"If I define myself as being right and [blah blah blah], then I conclude that I am right".

Re:no Ghost_no "singularity"_only sci-fi

[narcc]

Well, I guess dismissing them out-of-hand, without reading or understanding anything, is easier than dealing with reality.

You Kurzweil cultists make creationists look sane.

Re:no Ghost_no "singularity"_only sci-fi

[ultranova]

No one except Searle, Lucas, Block, Fodor, etc.

Well, don't keep us waiting: what specific capability are we talking about, what's the proof that human mind has it, and what's the proof that no mere Turing powerful system can ever have it?

Re:no Ghost_no "singularity"_only sci-fi

[narcc]

Why are you waiting on me? Did you forget how to use a search engine?

Start here:

Lucas, John Randolph (1961) "Minds, Machines, and Godel" Philosophy 36:112-137
Block, Ned (1978) "Troubles with Functionalism", Minnesota Studies in the Philosophy of Science 9:261-325
Searle, John R (1980) "Minds, Brains, and Programs" The Behavioral and Brain Sciences 3:417-457
Van Gelder, Tim (1998) "The Dynamical Hypothesis in Cognitive Science" Behavioral and
Brain Sciences 21:615-665
Fodor, Jerry A (2000) The Mind Doesn't Work That Way, Cambridge, MA, MIT Press
Thompson, Evan (2007) Mind in Life: Biology, Phenomenology, and the Sciences of Mind, Cambridge, MA: Harvard University Press

Though I doubt you actually care. You religious folks are only interested in validating your beliefs.

Re:no Ghost_no "singularity"_only sci-fi

[ultranova]

Why are you waiting on me? Did you forget how to use a search engine?

It's not my job to search for evidence for your claims. Or, for that matter, make your claims for you.

Though I doubt you actually care. You religious folks are only interested in validating your beliefs.

Ah, religion, the institution where you make a claim and when asked for specifics and evidence tell your audience to go read a bunch of old books rather than providing any. Does that sound familiar?

Oh well, since it's obvious this is some kind of personal issue to you, I'm leaving you to it.

Re:no Ghost_no "singularity"_only sci-fi

[narcc]

I provide names, you don't look. I provide citations, you don't look.

Call me crazy, but it seems to me that you're interested in anything but validating your own silly preconceptions.

It's not my job to search for evidence for your claims.

Actually, it's YOUR bullshit claim under consideration here. Specifically ,i>"So far, no one has demonstrated any ability of human mind that couldn't be replicated through computation.".

Of course, you refuse to even look at the evidence. Figures. It's exactly what I expected for you religious zealots.

Go pray to Lord Kurzweil, who offers you salvation in his video-game afterlife, for a way to combat heretics like me who assault you with reason, logic, and evidence.

Re:A call for help...

[K. S. Kyosuke]

Come on, Enya is like 53 or so. That's not new age, that's old age.

Postmodernism or theology?

[OneAhead]

This reeks of postmodernism or theology, in a bad way. I suspect they may be applying certain logical steps in ways they shouldn't be applied or confusing terminology somewhere. Like someone getting confused about certain qualitative statements in thermodynamics and coming to conclusions that are absurd from a statistical mechanics point of view - it happens in the classroom all the time. From a microscopic POV, the brain is all weighted connections and sigmoid transfer functions, and even if it's more complicated than that, one could still argue that all room-temperature ordinary matter physics can be approximately computed given an arbitrarily large amount of computing power. I refuse to accept that laws of physics that are computable would give rise to emergent behavior that is incomputable. Unless I'm the one who is confused about the meaning of the term "computable"...

Memory does decay everytime your retrieve it

[Anonymous Coward]

Every time you retrieve a memory you slightly modify it.
The act of retrieving it modifies the original and the act of pondering the retrieved memory also modifies the original tainting it with fresh associations.

This is not a personal impression it's biological fact: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1240066/

Consciousness is just better pattern matching

[protoporos]

In my opinion, the fact that our thoughts feed back into our incoming stimuli processing areas (so we can experience our thoughts, and thus remember them) along with the fact that as humans we sport a very advanced pattern matching capability on incoming stimuli compose the two critical factors needed for self-consciousness to arise. It's just a matter of adding the dots; realizing that all the stimuli that you experience have a common denominator (a common pattern), you. All the living beings with similar pattern matching capabilities in their brain should have developed a type of consciousness. You could say consciousness is like... critical mass. Above a certain point and the reaction is totally different. If you don't have enough mass, you might never experience the chain reaction. But that doesn't mean it's not there... Once you go above it, a new world of capabilities unfolds. -------- V. Toulias

It's a 'lossy' system by design_no flaw to detect

[globaljustin]

You start off-kilter and just careen into a ditch of dumbness...

1. Here is one mathematical model of a way that memories could work.

The researchers did *not* start with that at all...

Here's where they started, from TFA:

Tononi’s key idea is that consciousness is phenomenon in which information is integrated in the brain in a way that cannot be broken down.

"cannot be broken down" but it can be modeled in a way that proves the theory

Here's how, note the distinction, from TFA:

Maguire and co begin with a couple of thought experiments that demonstrate the nature of integrated information in Tononi’s theory. They start by imagining the process of identifying chocolate by its smell. For a human, the conscious experience of smelling chocolate is unified with everything else that a person has smelled (or indeed seen, touched, heard and so on).

This is entirely different from the process of automatically identifying chocolate using an electronic nose, which measures many different smells and senses chocolate when it picks out the ones that match some predefined signature.

Here's why their model is not as you say...it is a different *schema* of a model, not just a variation of an existing schema...context is different completely:

It must allow the reconstruction of the original experience but without storing all the parts.

That leads to a problem. This kind of compression inevitably discards information. And as more information is compressed, the loss becomes greater.

Their model was an intentionally "lossy" model..."lossy" is something we usually use algorythms to **avoid**...this use is new...it is not as complex as our brain, but it **does model a 'lossy' continuous memory/recall system that is integrated**

Here's more on Integrated Information: http://en.wikipedia.org/wiki/I...

Another way to think of the difference is Vector vs Raster graphics. Vector is infinitely scalable and perfectly lossless, but Raster is lossy.

This research proves that our brains process like a Vector graphics program, and the memory of our brains limits us just as the RAM of a computer limits how fast a Vector editing program takes to render a change in scale of the image.

Re:Memory is non-lossy? Research suggests otherwis

[ultranova]

Eyewitness accounts have been proven to be wrong over and over again. The assumption of a non-lossy memory is just false.

Let's play a little game. Go to, say, DeviantArt, and pick a random picture. With that picture right in front of you, can you describe it in such detail that I can find it? Or will the game end with me picking a random image that might, with some luck, bear some resemblance to the scene you described?

Eyewitness accounts are difficult because making a useful description is hard, even with memory entirely eliminated as a factor. And it becomes even harder when describing a sequence of events, because human beings don't actually differentiate between their sensory input and imagination - you don't remember hearing a loud noise, you remember hearing a shot. The latter is a conclusion, a scenario your imagination came up with to assign meaning to the former, which is just random noise without it.

This, BTW, explains a lot about human behaviour...

Right-O

[Anna Merikin]

Your cite of the "recent" study fits with my memory from the old school. There are several kinds (at least two) of memory: long- and short-term; one is chemical, the other electrical. Each reference to the protein carrying the memory rewrites it to include the information from the new conscious understanding and context, thus changing the protein when it is recreated. I am surprised that this method of decoding/recoding has not been looked into.

Re:Right-O

[Antique Geekmeister]

> Your cite of the "recent" study fits with my memory from the old school. There are several kinds (at least two) of memory: long- and short-term; one is chemical, the other electrical.

They're both _neurological_. The physical layout of the neurons and their junctions is apparently as critical as the physical layout of electrical connections on a computer chip is to the programs that might run on it. And there's little distinction between "electrical" and "chemical". The "electrical" impulses in the brain are apparently fundamentally chemical, not "meticallic" electrical currents.

Re:Physically impossible

[colinrichardday]

Are brains Turing machines?

Re:Stupid assumption

[OakDragon]

Memory is lossy, ask any psychologist/psychiatrist.

Or ask, uh, what's-his-name... You know, the guy with that used to have the blue car?

implications for computability

[SkunkPussy]

If consciousness is uncomputable, yet we quite clearly could generate consciousness if we were able to create/duplicate a brain, does this imply then that the brain is doing a higher level of computation? Perhaps the difference between polynomial and exponential, or perhaps something new entirely. If the brain is uncomputable then does this imply that P = NP using an appropriate computer (brain)?

Definition Mess

[Tablizer]

"Conscience" is too poorly defined to apply mathematical rigor to, unless you make a working definition for a specific theory. But others don't necessarily have to agree with your working definition.

"Self-aware" is commonly given as a requirement, but that's also ill-defined with a lot gray areas. We can build robots that are "self aware" in that they react or discard features of their own existence, such as shadows, as required by their tasks. But it's questionable whether most would agree that alone is being "conscience".
   

Re:It's a 'lossy' system by design_no flaw to dete

[Geoffrey.landis]

Yes, you're right; I eliminated a lot of their initial "making it sound plausible" steps in order to summarize the path of their logic.

To cover the part you mentioned, add step 0 to my list of their steps:

0. They make a number of assertions about the nature of memory.
Based on these assertions, they propose:

1. Here is one mathematical model of a way that memories could work....

If it's not computable...

[Alejux]

Then it's magic. If the brain is formed by neurons that work within a certain logic and mathematical model, then it's computable.

Re:If it's not computable...

[bucket_brigade]

I suspect that "it's magic" is precisely the conclusion they want to draw to protect their threatened special snowflake status. Otherwise clever people (Roger Penrose, etc) end up looking like total baffoons because they cannot even dare contemplate the notion that their mind is not a transcendent miracle of galactic proportions.

Human Consciousness Is Noncomputable

[NikeHerc]

For perfect examples, see anyone in Washington, D.C.

You can't image quantum

[tepples]

Unless consciousness turns out to rely on quantum mechanics. You can't image quantum.

Re:You can't image quantum

[lisaparratt]

Sorry, I thought we were discussing realistic approaches, and not handwavy appeals to magic and pseudoscience. Oh wait, slashdot, nvm, carry on.

Re:You can't image quantum

[dbIII]

Sorry, I thought we were discussing realistic approaches

However such an AI is still speculation until we get a bit more of a handle about what intelligence actually is so why shouldn't others speculate about "pseudoscience"?
We can build increasingly complex mechanical turks with rules such that outputs can match what a thinking being would do with the same input but we don't know how to make a machine think yet. It would be very nice if we could just save the memory state of an AI and use it to build another one, but it's a bit early to be taking such a wish as realistic. We can certainly transfer the rulesets from one mechanical turk fake AI to another but we shouldn't confuse that with the real thing. Even "neural networks" are a weak analogy to a dumbed down 1980s understanding of a very simple nervous system and owe far more to being simplified digital models of analog computer components.

Re:You can't image quantum

[narcc]

You think QM is magic / pseudoscience?

You must be a Randian.

Re:Stupid assumption

[frog_strat]

Yea, the guy that wore a shirt...

Re:Physically impossible

[Artifakt]

And are minds software that can:
a) perform all their functions when running on a Truing machine, without needing any other support.
b) software that can only come into existence when a turing machine already exists to create and develop it?
c) software that ceases to exist if it's substrait Turing machine is sufficiently damaged?

The interesting thing (to me), is that a great number of these minds have come to strong conclusions about such questions as "Is there life after death?, or "Can machines ever have consiousness?", but it looks like the question "Are brains Turing machines?" needs to be answered first, before any of these other questions can really be addressed. Note too, "Are brains Turing machines?" at least sounds like a question that can be addressed by the scientific method, with falsifiability possible,

But the assumption is proven wrong already

[idbedead]

Memories do degrade the more we remember them. A number of studies have proven this (http://www.radiolab.org/story/91569-memory-and-forgetting/). The tiniest amount of research would have shown them that their whole model is based an a false premise (lossless memory in humans). Our memories are fantastically-crazy lossy. The act of remembering pretty much rewrites the memory each time.

Re:It's a 'lossy' system by design_no flaw to dete

[AK Marc]

Their model was an intentionally "lossy" model..."lossy" is something we usually use algorythms to **avoid**...this use is new...it is not as complex as our brain, but it **does model a 'lossy' continuous memory/recall system that is integrated**

So MP3 can't be computed by or read by a computer? And the assumption that the memory is re-compressed with additional loss upon each read seems a dumb assumption. Memories are written again when read, not read destructively. They made lots of bad assumptions to get the answer they wanted. It's bad science.

Controversy: just add water and stir

[epine]

Contrary to the story summary, the recipe is not quite that easy. There needs to be at least some effort to disguise the act of speaking our of your ass.

There are seven layers of straw men between this outrageously overblown mathematical quibble and the true nature of human cognition.

Re:Memory is non-lossy? Research suggests otherwis

[amorsen]

Your phone number is digital. When you recall it, you know it is precisely right, and so your memory of it gets re-saved, error-corrected.

Most things you want to remember are not digital.

Re:Memory is non-lossy? Research suggests otherwis

[mveloso]

Some memories get stronger the more they're accessed, not weaker.

That doesn't imply that the details are more or less accurate - just that by remembering something over and over you makes it easier to remember.

Maybe I'm not sure what computable means.

[Hussman32]

Okay, if I make the following (probably erroneous) logical information lifecycle assumptions:

1. Assume the theory that we have short term memories feeding long term memories,

2. We have five senses that effectively record terabytes of information with near analog frequency (for example, you feel sensations, or no sensations, from the entire surface area of your skin as well as any internal nerves, and that can be recorded).

3. The short term pushes the most pronounced sensory data into long-term memory, and that long term memory is persistent..

4. Long term memory is persistent, but may not be completely accessible.

5. Memories of accessing those long term memories are persistent, and can be accessed, effectively creating copies.

6. Multiple access/copy loops in short periods of time are more likely to persist that memory copy more accurately, and fewer accesses increases likeliehood of loss/alteration.

7. The incredibly large vector of sensory data, for which no two humans will have identical experiences, coupled with differences in memory storage capacity, physiological differences, chemical exposure differences, and plain luck lead to effectively a chaotic system that cannot be computed because you're dealing with n>1e8 vectors expanding factorially over time leads to this still incredibly difficult to define concept of 'consciousness,'

So I guess I agree with the thesis that we can't compute consciousness, it's a chaotic system, but I wouldn't be surprised if it could be copied.

Re:Stupid assumption

[rubycodez]

please tell me more about your mostly bare-chested school, was it perchance co-ed?

Not a single thing wrong [Re:Bad syllogism]

[Geoffrey.landis]

Yes, it was a trick question. The answer is that there was not one flaw in their logic.

sensational headline

[Sperbels]

So, they say it's non computable, but then they concede that it is probably computable but likely too complex to do it right now. So, computable then. Sensational headline or what.

Non-computable

[wisnoskij]

OK, Show me one part of the brain that is not just (chemical) wires and electrical current.

Recalled memory _is_ lossy

[Antique Geekmeister]

Repeatedly recalling an event, as for story telling, restores a subtly _altered_ copy of the memory. This has been shown by many experiments about the plasticity of human memory.

Re:Physically impossible

[lister king of smeg]

No.

Just a common misconception of the idiots that make up the majority of CS majors.

Lets assume for the sake of argument the human brain is not a Turing machine. The universe however is a Turing machine and we are being computed on it and as we have consciousness as such via transitive property the universe is computing consciousness; therefore Turing machines can given enough time compute consciousness. QED bitches.

ps

Turring machines must have;
memory, the abiltity to read memory the ability to write to memory the ability to inciment and decriment data stored in memory, and at the minimum run a NAND instruction on data.
We have these abilities and can compute NAND, we are turing machines.

Memories DO decay

[Culture20]

Some are reinforced. Some are edited over time. Some are just gone. Do you remember every second of your commute three days ago? Today's?

Choice

[ememisya]

Consider for a moment what choice truly is. If I quickly asked you to choose between two colors on pieces of paper, red or green? You might have a choice in mind. But then if you saw me ask you with hesitation while looking at red for example, then a whole host of other influences come to be. You might think I'm trying to make you pick green and pick red, or think I'm trying to make you think I'm trying to make you pick green, and pick green just to spite me. There need not be any logic to choice, sometimes there is, hopefully most of the time there is. But I could be thinking about red apples the day before, and even though "RED apple" might be my thing, you never know what color I'm going to pick with absolute certainty by a simulation. You can only guess and it would be the same statistical probability of 50/50 with, or without preexisting data. Now have fun coding that. Consciousness goes beyond just mathematical equations, Math is based on symmetry, life not so much. Consciousness came all the way from the chaos of the quantum world, up into our classical physics and symmetry which survived impossible conditions to end up a mammal which calls itself human and writes so on Slashdot. Best we're going to get is a human like AI, which is enough to convince some folk but will fall short in insight and creativity, it will be a logic machine, just a very complex one. Assume the world is in brink of total annihilation, you ask an AI to pick who gets to live from the leaders of 2 opposing forces to decide the fate of the human race. Human 1 or Human 2, it makes some value judgements and decides based on its understanding of its data. Turns out survival really depended on a pollen which made the chosen human sneeze and press a button at the right time. How do we deal with such interrelated complexity of the Universe? Insight, we feel and make the right choice. Once again, good luck coding that. Mind is like a 0 dimensional dot, it's infinitely large and infinitely small, but only as small as the smallest pen we have today. Anyone who thinks the human consciousness is coming up as the next big Google thing is just asking to be shipwrecked by the laughter of the Gods.

Re:Choice

[Dr Max]

if you have a model with all the variables (the nature and nurture side) you could guess which color he will choose. that said, it would be a gigantic task and would require more computing power than all of earth has at the moment.

Re:Choice

[DexterIsADog]

...I'm trying to make you think I'm trying to make you pick green, and pick green just to spite me... There need not be any logic to choice, sometimes there is...Now have fun coding that....Consciousness came all the way from the chaos of the quantum world...end up a mammal which calls itself human and writes so on Slashdot. ...Assume the world is in brink of total annihilation, you ask an AI to pick who gets to live from the leaders of 2...Turns out survival really depended on a pollen which made the chosen human sneeze....Mind is like a 0 dimensional dot, it's infinitely large and infinitely small...just asking to be shipwrecked by the laughter of the Gods.

I'm glad the experiment in Colorado is going so well, but please, leave some pot for the other people!

Re: Choice

[ememisya]

The mind indeed doesn't have to be logical. I'm not sure why a 0 dimensional dot is not a good comparison for the mind as it was the mind that both came up with religion and science. The mind can entertain paradoxes, Universe cannot. Also the shipwrecked part is from Einstein. Now I don't know what he was smoking but sounds like it did him well.

Re: Choice

[DexterIsADog]

"...not sure why a 0 dimensional dot is not a good comparison for the mind..."

Triangle bookfish Niagra. Kudzu plain binding? French toes!

See? The problem is that what you're writing is gibberish.

No it isn't

[multi io]

I hardly understand a goddamn word of TFA and have never heard of the "Integrated information theory", but I know that TFA's proposition must be false because the brain is based on the laws of physics, which are computable. Q.e.d.

Re:Memory is non-lossy? Research suggests otherwis

[Bite The Pillow]

Wrong. Your short term memory can remember, without effort, 7 digits easily. It therefore is not so complicated that it changes before you put it back.

Remembering the order of events over several days, on the other hand, does not fit in short term memory in a coherent fashion. So you may gradually put several things on a week long trip on the same day. When telling the story, someone else who was there says "No, it was the next day, because [reason]". You didn't have a forceful enough memory to record a separation, and they gradually blur the line while you remember, and now they are on the same day. Something made an impression on the other person that put it on a different day - maybe the same day as something you remember happened on a different day.

There are lots of other little things, like pattern recognition. If someone's number is nearly the same as yours, you may repeat theirs when saying yours, because you call them more than you call yourself. And then you have to struggle to recall your own number, and it is vulnerable to change in that moment. But if you're not sure, you pull out your phone and problem solved - you store it correctly. There's more, but I don't see you coming out of this any better if I keep typing.

Still intractable

[tepples]

Given an actual Turing machine and a bit of patience, you can solve halting for all physically existing machines.

True, a linear bounded automaton (LBA) or below can be solved in theory by running two in a tortoise-hare configuration until they hit an identical state. But in practice, exponential growth means anything longer than a human lifetime is intractable.

I am not sure whether that has been proven.

I was expecting Wikipedia's article about LBAs to list the important theorems. Its article about the halting problem quotes Minsky that iterating through all states is sufficient but doesn't prove anything is necessary.

Nondeterministic TM emulation

[tepples]

That's the nondeterministic Turing machine. A deterministic one can emulate a nondeterministic one, making the nondeterministic one faster but not strictly more powerful.

assumption fail

[s13g3]

"When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong." ~Arthur C. Clarke

Just because they can't figure out HOW a digital machine would compute it does not mean that machine consciousness isn't possible... merely highly unlikely with the current state of the art.

Semi-/organic systems or components or other radically new or different implementations and designs of hardware, new materials, as well as new software techniques could blow their assertion out of the water next month as easily as in the next decade.

Pretty much every time someone says "you cannot", someone eventually comes along and develops something to prove them wrong. Just like they said no one would ever break the sound barrier, or put a man in orbit, or that there's only a need for a handful of computers globally. You know, like every time someone says "tape is a dead storage medium", or "ZOMG Moore's Law is going to fail in the next 5 years", and are consistently proven wrong. This scenario is no different, and merely indicates a lack of understanding of science on the part of the researchers, as well as a lack of imagination. Just because they can't figure out HOW it could be done does not mean it is not possible.

memories seem lossy to me

[jinchoung]

lot of assumptions and unproved assertions here...

Re:Physically impossible

[BungaDunga]

Computers aren't Turing machines, either. Von Neumann architectures are really good at simulating them, but they're not Turing machines at all: they allow an operator to fiddle with the state while it's running.

False assumption?

[AJWM]

Who says memory retrieval is non-lossy? It's an organic process, of course it's lossy. Our brains just make shit up to fill in the gaps.

The stuff we retrieve frequently is slightly less lossy because it gets refreshed (somewhat) when we remember it (sort of remembering that we remembered it).

And our brains are very good at making shit up to fill in the gaps, almost too good.

Re:False assumption?

[anomalous3]

Not only this, but memories can be modified when retrieved, so memories that are frequently retrieved are at greater risk of being corrupted.

Re:It's a 'lossy' system by design_no flaw to dete

[Dr Max]

I think you mean 1. here is one mathematical model that doesn't allow memories to work.

Re:Memory is non-lossy? Research suggests otherwis

[Dr Max]

"You probably have thousands of different individual memories of your phone number and address." that doesn't sound like an efficient way to remember few digits. Surely it just reassurers the original memory or a makes a stronger new replacement memory.

Re:Roger Penrose

[sgunhouse]

Sorry, didn't realize I wasn't logged in.

decay of memories

[oldestgeek]

"retrieving memories repeatedly would cause them to gradually decay." What is the science that shows they don't?

no, but yes to mathematicicans

[globaljustin]

So only mathematicians should ponder and discuss their own mind?

No.

It's bad when theorists who are only philosophers who ponder and discuss their own mind try to use mathematics.

If you are talking about technology you need to *learn how it works*

A coder, a person who writes instructions with symbols to automate machine behavior, would be equally invalid to claim that his technical knowledge allows him to claim that, say, Kirkegard is right about metaphysics over all others because of code

no "teh singularity" != "god" exists

[globaljustin]

holy crap...I *knew* that "teh singularity" folks were irrational but you just admitted it

you're doing what you claim I'm doing...here's how

Anti-AI isn't science, it's just the ancient belief about the supernatural specialness of human soul,

there's no such thing as "anti-AI" there is only "anti-bullshit"..."anti-pseudo science"...that's what's in play here

see, here in science land, were not trying to prove or disprove the existence of a 'god' or 'soul' or anything supernatural as you claim...

just because you let your opinions about the supernatural invade your science doesn't mean ***Everyone*** thinks that way

I don't game out my science to see if it proves or disproves a 'god' then adjust accordingly...neither does any real scientist

You're proving that **YOU** do indeed game out your opinions to strategically affect your science

you're doing what you claim I am doing

"Turing Test" is stupid

[globaljustin]

simulate human behaviour enough to beat a Turing test.

the whole "Turing Test" notion is ruining our industry

it's bullshit...litterally, it's a demonstration of programmed bullshit

it's not science, because "passing as human" depends ***ENTIRELY ON CONTEXT*** including the person you're trying to fool

it's not testable b/c it's not falsifiable...or really it's impossible to tell if it is falsifiable or not because every instance is dependent on too many variables...each instance of human interaction

fooling some dumbass that a computer is a person is not a scientific test

get Turing out of your brain forever...if you're a coder it will help your performance

talking about "unscientific"

[globaljustin]

Memories are written again when read, not read destructively

really???

because that would be the single biggest discover of neuroscience...that the human brain stores information as a computer does on a disk

it would be the "biggest" because it would overturn everything we know about (which, I must remind you, isn't much!) the way the human brain works

the researchers in TFA are doing real, testable, falsifiable science

Re:talking about "unscientific"

[AK Marc]

..that the human brain stores information as a computer does on a disk

I didn't say anything of the kind. And when I read a disk, it doesn't write what you read again, so you are wrong on both counts.

it would be the "biggest" because it would overturn everything we know about (which, I must remind you, isn't much!) the way the human brain works

I have a psychology degree (and focused on the "hard science" not the "tell me about your childhood" side). I don't think I need some random guy lecturing me about what we know and don't know. I've done some "real, testable, falsifiable science" involving the workings of the brain, specifically memory/cognition.

Orch-OR - Google it...

[Mof-Tan]

This topic and no one mentions Stuart Hameroff or Roger Penrose...

Google it my friends, these guys have a much more fascinating and mathematically rigorous theory than these guys in post seem to have, including a much more convincing definition of what consciousness is.

Oh dear

[bucket_brigade]

Apparently making grandiose statements about things you have no clue about is how I should have been doing science. Instead of, you know, working hard to justify your conclusions beyond all doubt.

Not non-computable at all

[Warbothong]

In other words, a God-like observer with perfect knowledge of the brain would not consider it non-computable. But for humans, with their imperfect knowledge of the universe, it is effectively non-computable.

What they're saying is that there are limits, beyond undecidability, when a human mind tries to study itself. It's an algorithmic analogy to the classic data-storage problem of trying to imagine, using your mind, the whole contents of your mind. Via recursion, that can't be done. Likewise, TFA is saying that we can't use our minds to compute some things about our minds, even though an outside observer with perfect knowledge of our mind could do so.

The reference to PCs is hence entirely wrong. What they're saying is that if a PC worked like our brain, it would be limited in its introspection ability compared to, for example, a hypervisor on which it's running.

Re:Physically impossible

[bingoUV]

The universe however is a Turing machine

No it isn't. It isn't even a machine.

memory, the abiltity to read memory the ability to write to memory the ability to inciment and decriment data stored in memory, and at the minimum run a NAND instruction on data.

And most human memories cannot write to infinite memory addresses and recall them correctly - which is a necessity for being a Turing machine. Not even a trillion memory addresses, most would make significant errors within hundred memory addresses.

So no, humans are a very poor approximation of a Turing machine.

Probably you need to understand that Turing machine is a theoretical concept and there need not be actual Turing machines in the world for that to make sense or be useful.

Re:Physically impossible

[colinrichardday]

Quantum mechanics can be approximated by Turing machines, but I'm not sure if the integrals can be exactly evaluated by Turing machines.

Just Kolmogorov Complexity...and religious intent?

[fygment]

The argument for compression describes essentially Kolmogorov Complexity. The idea is that the K.C. of something (and everything can be reduced to a binary string) is the length of the shortest program (if you look at it algorithmically) that can describe that object (reproduce that binary string) and stops. In TFA, the example is reducing the description of an infinite sequence of numbers to a finite program that calculates the odd primes and adds one to each. The number pi is infinite in length and random, but not complex since there are small programs that caculate it; an infinte truly random string of numbers would have infinite K.C. because the shortest program would be "print -infinite string". The K.C. of an object is not computable (it's related to the Halting Problem), essentially you never know if you have the shortest program to describe an object.

So here are some observations

a) the whole premise rests on the assumption that the brain is a Turing complete computer ie. the brain is a computer too. So if the brain is a computer, why couldn't other Turing complete computers mimic it? In fact, K.C. theory uses the idea that there is a Universal Turing Machine that can mimic all other Turing machines. If the brain is not a Turing machine then you can't really make any comments about it's compression abilities, etc. because algorithmic theory is grounded on the Turing assumption;
b) the TFA implies that compression is lossy. Well, not all compression is lossy and the example provided (prime plus 1) is not lossy at all, it's perfect. So what is the point of that example except to suggest that memory must be perfect compression? ... it just seems like a pointless example; BUT
c) the assumption that memory is/must be perfect compression seems extremely flawed. Memory is not perfect and most memory seems to degrade over time (see witness reports, personal experience, etc.)

So ... the whole paper seems riddled with discontinuities or inaccuracies. Really it seems like it would have been better to say:

"The brain compresses information in a lossy fashion. We don't know how. Assuming a Turing process is occuring, then the brain is looking for the best compression it can but it can never know if it has the best or not. A computer will be in the same boat." BUT

THE FEAR

Basically the article is making (a flawed?) claim "Machines can never be conscious." The argument plays very well to a religious and research oriented crowd. First: machines can never be made in the image of man. We are not gods. Second: There is no requirement to consider ethics in AI . No matter what the AI seems, it is not, _can't be_, conscious. Therefore, should you create a robot that walks, talks, acts, and feels like a human ... well, it isn't conscious, so do with it as you will.

Bullshit! It's what's for dinner!

[gestalt_n_pepper]

If the mathematics implies that machines can't be concious, it implies that humans can't be concious. The math is the math, whether the bits are in silicon or some goopy stuff between the ears.

Seriously. The obvious implication of this claim is that organic humans are some kind of super duper special thing beyond mathematics. This is sheerest horseshit and there's not a shred of objective evidence indicating this.

memory is non-lossy

[MikeSyposs]

Why is this assumption made? Could not there be a lossy effect that is small enough not to be noticed intuitively? To have the entire study tied to this premise seems absurd.

Re:It's a 'lossy' system by design_no flaw to dete

[Lonewolf666]

Yes, it is an unusual model and building something along these lines might not be a good idea in engineering. But who knows what stupid concepts evolution might come up with. All it takes is a local maximum in efficiency and development can get stuck there.

Oracle Breakable After All

[tepples]

I don't know whether it's solvable, but the infamous post "The first Slashdot troll post investigation" showed that the Oracle machine is breakable.

Memories do degrade... Don't they

[dan_hoog]

Agree 100% The model may be interesting and advance a theoretical approach - a real plus. However, key assumptions are faulty. Evidence suggests that memory is rewritten with current associative context when accessed. Our awareness fills in plausible details and assumes the memory is complete and accurate (analogous to not noticing your macular scotoma in the eye). Memory storage is actually quite lossy and more tuned to keep our internal dialog consistent than anything else. Further, noncomputable will not rule out heuristics that closely approximate its behavior or even internal experience. The real brain seems more heuristic and pattern matching than computation in its origin. That's all...

can i play?

[globaljustin]

2024 is not a date or time. In the multiverse it is a place.

That year, 2024, is the point in space/time where the natural progression of human consciousness & technology & science converge and we will take a step forward equivalent to the first humans to make artwork or speak language...only this is not an inward step, but an outward one.

Conspiracy theorists talk about "predictive programming" and it's bunk of course, but humanity has known this all along. The parallel is all humans who will be alive during the coming transition & all humanity..."we" have always known something was coming. Books like Childhood's End and films like 2001-A Space Odyssey or Contact are really a primer, like an introduction. Those works intentionally prepare our minds for 2024, even though the people who made it may not be conscious of it!

If you can understand intentionality of will without consciousness then you're on your way to becoming a Pleiadians ;)

/sarcasm

Consciousness is the short-term memory

[tepples]

I define consciousness as the evolving state of the internal articulatory loop, sketchpad, and episodic memory.

Mathematical Model Suggests That Human Consciousne

[hinckeljn]

I would not assume human consciousness and human memory are not binary and deterministic the way machines are. Human memory and human consciousness are the results of multi-state (almost continuous) arrangements of the brain cells and the diffuse connections between these cells (dont's ask me to explain these! Yes I talk to God!!!)

Re:Mathematical Model Suggests That Human Consciou

[hinckeljn]

Remove one of the two "not's". My mistake, not God's!

Re:Stupid assumption

[narcc]

Yes, it was. Unfortunately, it was also a boarding school for the morbidly obese...

Re:It's a 'lossy' system by design_no flaw to dete

[AK Marc]

I could prove that memories are read non-destructively, but are re-written upon each reading. But I don't have the funding and time. Why solve an "obvious" thing. How do I know? I can remember how I remembered something 15 years ago from 15 years before that. I'd guess that to be unusual, but I can remember remembering something, and the memory itself changes, based on perceptions at the time of reading, and those color future recallings of the memory.

It also explains all the false abuse claims, without anyone involved "lying". Everyone tells the truth as they know it. The problem is memory is re-written on every reading, and trying to read something that isn't there results in the "desired" memory being actually written, given enough coaching and time. and there's no "lie" when there's no deception.

The A/C's identity is . . .

[mmell]

Are you this guy? The Start64 malware site shows the following:

Company: Panisz Peter

Address: Kossuth Lajos u. 51 Dunabogdany 2023 HU

Phone: +36.203367173

Fax: +36.203367173

But I think he's living at his mother Jan Kowalski's basement at:

Alexander Peter Kowalski

903 East Division Street

Syracuse, N.Y. 13208

Apartment #1, Lower Level

At least, that's where he wants users of his hostfile manager to send him money.

Bad Assumption

[urgelt2]

The assumption that memory suffers from no losses is just about the worst possible assumption you could make. There is plenty of evidence to the contrary: human memory is lossy. Any conclusions arising from an assumption like that one is not going to get us closer to an understanding of human consciousness.

it's like getting laid

[globaljustin]

Yeah, we can do this:

*The neural network of human brain can be atomized down to neurons and their connections. It is not magic

but that doesn't explain it's function, it only **describes** the mechanism...and we have **barely scratched the surface**

you don't know nearly what you think you know about how you know what you know b/c none of us knows

think of it this way...

is getting laid a perfectly copyable process?

of course not...even if you pay for it...no way...if you get laid on a one-night-stand it was certainly a ***definable series of events*** but replicating those events ***in no way*** guarantees a similar outcome

we know the component parts...hell, whole **companies** make products dedicated to just a *part* of the process (condoms for ex)

but no way, no how, never can anyone claim 100% success rate with some formula when it relates to human behavior

the notion that everything is computable is foolish and unprovable b/c it's only half a premise...it's illogical and you need to do yourself a favor and ditch the notion forever

ditching 'computability function' Turing bullshit actually won't hurt you or take any effort...it's like removing a heavy backback...

just...let...it...go...

We are a Computer

[ramorim]

I believe that we are some kind of a computer... in a molecular-organic way. So, I too believe that it will be possible - in the (far?) future - for us to create a sentient computer. And I think we will only be able to do that when we could be free from the extreme limitations imposed by the actual hardware (and binary) system in current use.