Slashdot Mirror
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."
Nope, just a bad copy of it.
Get free satoshi (Bitcoin) and Dogecoins
that's what they think.
> They say that the process of integrating information is equivalent to compressing it.
>
So human consciousness equals gzip?
Wait..what??
How do we compute it then?
Is hard. Lets go shopping!
Any turing machine operation by definition must be completable by any turing machine, the only difference would be the time it takes. Sooooo, I don't know what these guys are smoking but I guess I'll take a look at the actual article.
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.
but then again, odds are anyone making such a claim will be long dead before it could be proven wrong.
"Have you ever thought about just turning off the TV, sitting down with your kids, and hitting them?"
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 lossy, ask any psychologist/psychiatrist.
Not retrieving memories is what causes them to decay. Ever hear of refresh?
I think it is plausible that memory can be lossy. It is a well accepted concept in psychology that memories are subject to modification when they are recalled. This implies that memories must be resaved after being recalled and that errors can occur during this process.
... it's backed up in the cloud.
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/...
"You might as well get your son a ticket to hell as give him a five string banjo." -unknown minister
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.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
Finally, concrete proof that there is a magical ghost inside each of us.
A. Memory *is* lossy. The more you remember something, the fuzzier the details get. You may *think* you're remembering it exactly as it was, but you just don't remember the previous versions to compare the current version to.
B. Everything is computable given the right models and starting conditions. Even uncertainty can be built into the model. That doesn't mean your brain sim and your real brain will be in sync, but the process could be identical.
C. We know hardly anything about how the brain works. To say consciousness is non-computable is like saying there's a Chuck E Cheese in the center of every black hole..
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.
http://www.geoffreylandis.com
...can someone please explain why this is not just a bunch of new age crap? Thanks!
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).
That was the turning point of my life--I went from negative zero to positive zero.
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.
> 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.
http://rareformnewmedia.com/
"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...
Bitcoin pyramid: Join here: http://www.bitcoinpyramid.com/r/1427 it's FREE!
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.
Tom Swiss | the infamous tms | my blog
You cannot wash away blood with blood
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.
How is the brain not a computer? Pfft...ridiculous conclusions.
Cognitive dissonance ALONE would defy mathematical understanding.
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.
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.
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.
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.
Tell me something new.
Hmmm. Do you find yourself occasionally having to re-learn your address or phone number ?
> otherwise, retrieving memories repeatedly would cause them to gradually decay
So, i guess this was never observed on real humans?
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.
excitingthingstodo.blogspot.com
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.
Belief is the currency of delusion.
Memories decay upon recall. Your brain basically alters the memory slightly each time. This can be used to erase or alter memories.
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.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
If this is true, what does that mean for wankers like Kurzweil and the fantasy of the 'Singularity'?
Yes. Although your definition of re-learn might not be broad enough.
Cannot be broken down?
Everything can be broken down, it's just a matter of learning how it works. The conclusions are half baked.
In practical experience we know that the more we continually recall a memory, the better we remember it. Meaning recalling a memory actually preserves it. This has nothing to do with our tendency to combine memories with other memories and fantasies (corrupting the data).
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.)
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.
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.
I'm a good cook. I'm a fantastic eater. - Steven Brust
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.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
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...
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
Thank you Dave Raggett
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"...
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/
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
You start off-kilter and just careen into a ditch of dumbness...
The researchers did *not* start with that at all...
Here's where they started, from TFA:
"cannot be broken down" but it can be modeled in a way that proves the theory
Here's how, note the distinction, from TFA:
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:
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.
Thank you Dave Raggett
"But they point out that this cannot be how real memory works; otherwise, retrieving memories repeatedly would cause them to gradually decay."
This is a non sequitur.
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...
Forget magic. Any technology distinguishable from divine power is insufficiently advanced.
you guys keep linking here lately
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.
If a theory leads us to an insane conclusion, I think we should believe the theory.
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)?
SURELY NOT!!!!!
> retrieving memories repeatedly would cause
> them to gradually decay.
Wait -- retrieving memories does cause them to decay!
Yes, I suppose at some point you must have had to re-learn it until it "took". Or do you think that somehow we're hard-wired to remember addresses and phone numbers instantly after hearing them once, but everything else we learn is a long and painful process?
"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".
Table-ized A.I.
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....
http://www.geoffreylandis.com
Then it's magic. If the brain is formed by neurons that work within a certain logic and mathematical model, then it's computable.
So, technically, if you can simulate the behavior of a volume of 1.5 liters of matter, you can simulate a human brain.
Also ... "retrieving memory causes it to decay"? That's what feedback loops are for. Retrieving (or doing nothing) with computer DRAM causes the data to decay, that's why it is refreshed regularly.
For perfect examples, see anyone in Washington, D.C.
Circle the wagons and fire inward. Entropy increases without bounds.
Unless consciousness turns out to rely on quantum mechanics. You can't image quantum.
... Now there's something that's non-computable for you.
If they say that they can't mimic human thought using the current algorithm, then they are using the wrong algorithm. This whole "Flight is only possible for birds" stuff is bunk. Likewise "its not safe for humans to travel faster than 30 miles per hour". Get a better algorithm; try a little harder.
Memories get modified when we relive them. Memory is lossy, and consciousness is computable. These scientists could have known that, it's clearly evidenced by amnesia curves (I worked with a professor that did that research)..
One reason a phone number or address aren't lossy is because of "feedback" or "new memories". When you see your phone number again it is a new memory, when you give out your phone number and it is successful, that is another NEW memory. Same with your address. You probably have thousands of different individual memories of your phone number and address.
What you are talking about is being able to successfully recall a fact, not a memory. By accumulating multiple memories on a fact you make the recall more robust and even create a feedback loop where you are creating new memories of the fact, keeping it fresh.
Do you remember the FIRST time you saw your phone number or address... this is a "specific" memory.
Why is this crap even posted here?
this reminds me of dr steven wolpert's much earlier work on the photographic natire of human law and the human experience of time as non-reversible. i wonder if the theory might work a bit better if they adopted an entropic approach, and declared human consciousness "highly unlikely" to be simulated by a machine, instead of outright impossible : in the same way that it is extraordinarily difficult thoguh still theoretically possible to know the microstate of a physical system.
this reminds me of dr steven wolpert's much earlier work on the photographic natire of human law and the human experience of time as non-reversible. i wonder if the theory might work a bit better if they adopted an entropic approach, and declared human consciousness "highly unlikely" to be simulated by a machine, instead of outright impossible : in the same way that it is extraordinarily difficult though still theoretically possible to know the microstate of a physical system.
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.
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.
Learn to love Alaska
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.
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.
Finally! A year of moderation! Ready for 2019?
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.
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.
"Who are you?" "No one of consequence." "I must know." "Get used to disappointment."
My guess is things like that are reinforced via some kind of mechanism explicitly designed to reinforce certain types of memory.
Yes, it was a trick question. The answer is that there was not one flaw in their logic.
http://www.geoffreylandis.com
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.
OK, Show me one part of the brain that is not just (chemical) wires and electrical current.
Troll is not a replacement for I disagree.
No, but that's because it's constantly *written to* as well. I frequently make use of that information, and in doing so, reinforce the memory.
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.
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?
I told a lie years ago. I know it was a lie. I thought about it in great details. Now when I remember that event, it is true and it is also false. I can pass a lie detector tell it. The more I recalled it, the more fine-adjustments i made to it. I remember doing it - clearly. It is in fact, a fact. But there is another version - another memory of me knowing I made it up. I don't recall all the adjustments I made to it, but I believe I recall the near-first version and the last version.
It reminds me of a linked-list in a way...
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.
The language of the universe is math
We are part of the universe
The language of us is math
Therefore we are computable
So tired of defeatists getting press...
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.
We all die...entropy doesn't apply to memories?
The way I remember something being non computable is that it would take a model of the same complexity as the real thing to predict what was going to happen.
ie. most mathematical models are simplifications of their respective problems.
Some cannot be. Physics has the multi-body problem. Chemistry has thermodynamics and the gas laws.
Considering the only demonstration of consciousness that we have is biological and the hierarchy of science is math -> physics -> chemistry -> biology. With uncertainty being introduced in the physics then is it any wonder that biology is non computable in a time shorter than would be needed to observe the subject of the model?
Sez who?
Employee Of the Month - Cyberdyne Systems Corporation - September 1997
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.
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.
That's the nondeterministic Turing machine. A deterministic one can emulate a nondeterministic one, making the nondeterministic one faster but not strictly more powerful.
"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.
"Inveniemus Viam Aut Faciemus" 'We will find a way... Or we will make one!' --Hannibal of Carthage
lot of assumptions and unproved assertions here...
Everyone seems to jump on recognizing ones self when they think about consciousness, but you have to think about the two "branches" on consciousness: One of those branches is highly computational and correlates with intelligence... and yes, you need a certain amount of intelligence to be able to include "self" in your mental model in such a way that you can recognize self. But when people think of recognizing self with respect to consciousness, I think they're talking about the "experience" or "sensation" or "qualia" of that "oh my gosh moment" when you comprehend and "feel" your own existance, and like many other kinds of conscious experience, like color, it doesn't seem sufficient to explain that in terms of computation in the way a computer does computation.
Current wisdom is that your memories are kind of like diffs. They only store the important points, and the rest is (re)constructed in your mind from pattern matching routines. And I mean (re)constructed. You don't see what you think you see, too much info to handle and move around. Think of join-the-dots, that's a fairly good approximation of your memory. It's a little sad when you consider its implications. After about 6 years old, your brain is full. Then the tar.gz process begins. Then you lose memories. Depending on your "bioware", you may not have very good compression ratios - you are forgetful whilst others' seem limitless in their capacity to learn, but I would say (pure conjecture) that this is another reason for dying. You can't operate anymore. Everything gets so compressed and interlinked that it becomes a foggy hazy world (alzheimers?). I don't believe we'll get 200 year long lifespans till that's sorted. That'll take a few centuries more understanding of the brain first.
According to recent papers they also appear to be stored chronologically (not associatively, as you would think), meaning that there's possible loss just getting to the memory in the first place, if there are no synapses close, all the fresher memories need to be tunneled through to get it.So you've got entropic loss right there, just from the storage and access routines. Think of your memory as a pointer stack list or whatever you computer guys call them exactly, but the iteration through the values is analogue, not digital, so you don't get a clean iteration and so you can expect some things to go into the stack and then never be retrievable.
- prof
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.
-- Alastair
Let us explore the common concept of memory: Most people are probably talking about 1. all the things I can consciously recall. and 2. think things are forever stored in their mind like a filing cabinet. The above 2 points are very general, and I admit they may be wrong.
Lets look at memory moving from behaviorist to more cognitive understanding: Classical Conditioning:
Pavlov makes a dog salivate when ringing a bell. Memory or recall at an unconscious level is responsible for the reaction.
Operant Conditioning: Punishment for a particular behavior will reduce the likelihood of that behavior occurring again. Again memory is somewhere in the mix. I have a response I will do less often because I don't want to get punished.
From the above orientation, we see memory as being inherent in the behavior.
Moving more cognitively: People who have damage to their hippo campus can develop anterograde amnesia. These people can not form new memories. (think Memento). Studies have been done on them to determine if indeed they no longer have the ability to produce memories. One experiment, used an aversive event on subjects, then distracted them long enough that they forgot the event. It was found that even though they would forget the event, they would still physiologically react to it before it occured again. In this case the reaction would be increased heart rate and sweat.
More cognitively (and more salient to this topic): Elizabeth Loftus has found that memories can be changed after the fact. She has found that that certain ques in questions can change the way we remember things. For instance, in one experiment participants viewed a car accident on video and then answered a question about how much broken glass was on the street at the accident. Participants who were prompted with the question "How much glass was on the street" reported remembering glass significantly more often than those who were ask "Was there glass on the street?" This shows that memory retrevial is fallible.
I think you mean 1. here is one mathematical model that doesn't allow memories to work.
Rocket Surgeon.
Maybe you don't compress/uncompress your phone number to recall it...and this is different from other memories you do compress/uncompress...the authors claim that it is compression/uncompression that leads to lossiness...
...theory to show that the process of integrating information must noncomputable.
"The 9000 series use holographic memory, so chronological erasures would not work."
Same thing here
There are others as well, e.g., and may even be more convincing:
http://www.goertzel.org/books/complex/ch8.html
"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.
Rocket Surgeon.
He's speculating but he does actually have a point.
Although we through around the term consciousness as if its a defined idea it really isn't. You should throw your claim of falsifiability to anyone that makes an absolute claim that that computers will never be able to achieve consciousness. (given we first should have a firm understanding of conscious before claiming with certainty we can't digitize it).
Science fiction may be fiction but some of it ends up as science fact once we fill in the gaps in our knowledge. Falsifiable is useful in the sense one shouldn't assert something as an absolute fact without concrete evidence but that doesn't mean we can't use our imagination to speculate. Science would be nowhere without the first baby steps of speculation.
That said, my own (speculative) opinion as a programmer is that a purely silicon based computer will never be conscious but we will eventually be able to create programs that can simulate human behaviour enough to beat a Turing test. My rational (perhaps mistaken) -- that there is an analogue chemical aspect to consciousness. The precise chemicals that make up a human are probably a crucial factor to achieving consciousness as we experience it. We should be able to eventually get a toaster to sound like it's our best friend but from the toasters perspective it will still be just a calculator.
That also said, we still tend to think of computers and biology as separate ( a mistake Kurswell makes by setting up computers as some sort of ideal vessel for consciousness). Human consciousness in its present form is unlikely to be some evolutionary end point. Things will get much more interesting if/when humans start mixing organic materials with silicon in the not too distant future.This in a teaser sense is what TV, the Internet and video games represents. We are ceding part of our identity and physicality to a virtualized world while the rest of us remains organic. This trend is likely to continue with advanced interfaces that obscure the distinction between physicality and virtualility -- to the point the two are practically indistinguishable from a conscious perspective.
Although it still in the realm of science fiction, and we currently lack the knowledge to achieve this, from a theoretical physics perspective there is no reason I know of why it won't eventually reach the point where we can spend much of our time in virtualized space with our organic conscious side safely tucked away somewhere (hopefully not being used as batteries).
Amazing. I read through all the comments and see not one reference to "The Emperor's New Mind". Roger Penrose reached the same conclusion - for different reasons - over 20 years ago. (Full disclosure - I am in a footnote to his second book on the topic, "Shadows of the Mind".)
"retrieving memories repeatedly would cause them to gradually decay." What is the science that shows they don't?
computate stupidity
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
Thank you Dave Raggett
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
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
Thank you Dave Raggett
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
Thank you Dave Raggett
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
Thank you Dave Raggett
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.
Die dulci fruere. Have a nice day.
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.
> By assuming that the process of memory is non-lossy, [...]
Aha? The assumption itself is doubtful...
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.
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; ... it just seems like a pointless example; BUT
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?
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.
"Consensus" in science is _always_ a political construct.
what is consciousness again?
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.
Please do not read this sig. Thank you.
Is it the people from reddit that are completely closed minded? Until we either have or don't SA AI it will seem impossible for most. Even after we get it - if we do - Most people will not believe that it is really SA AI, They dumb it down to parlor ticks and the like.
"Just because it sounds alive and thinks as such does not mean that it is..."
It's a good thing that every memory is a 10 digit number, otherwise there might be some types of complex, non-symbolic information that can't be consciously checked for accuracy each time it's retrieved.
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.
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.
C - the footgun of programming languages
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.
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...
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 ;)
Thank you Dave Raggett
I define consciousness as the evolving state of the internal articulatory loop, sketchpad, and episodic memory.
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!!!)
This makes sense in the way that once information is accessed (remembered) it's memorised / remembered separately again and kept fresh while older unsed memories which aren't accessed or actively remembered wither away and eventually disappear.
That's my experience of memory as far as I can remember.
Remove one of the two "not's". My mistake, not God's!
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.
Learn to love Alaska
But I think he's living at his mother Jan Kowalski's basement at:
At least, that's where he wants users of his hostfile manager to send him money.
Arxiv paper claims that if ones new memories are formed and compressed using all his previous experience, then retreiving those memories would also require all previous experience. Like for a series defined using recursive formula.
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.
Yeah, we can do this:
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...
Thank you Dave Raggett
No, I haven't RTFA, just the Slashdot summary. But, having researched a related topic last year, my understanding is that recalling memories _does_ cause the original memory to decay, and quite rapidly, at that. One might even characterize the process of successively recalling an event as a series of remembrances of remembrances. That is, the third time we remember an event, we don't actually remember the original event for a third time, thus possibly gaining new detail. What actually happens is closer to us remembering the second time we remembered the event. Eventually, you wind up with a game of telephone, in which your memory has been successively filtered by your intrinsic biases into something that may bear little resemblance to the original event. Rashomon & all that. This topic has become a controversial issue in the legal system, which relies heavily upon eyewitness accounts as a particularly probative form of evidence. But it's turning out that it's just as easy to remember something that didn't happen as it is to forget something that did.
Just sayin'. OK, back to the main conversation.
Mathematics deals exclusively with the relations of concepts to each other without consideration of their relation to experience. -- Albert Einstein
also relevant: http://en.wikipedia.org/wiki/List_of_fallacies
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.