Slashdot Mirror
The Semantics Differentiation of Minds and Machines
John David Funge writes "In Dr David Ellerman's book Intellectual
Trespassing as a Way of Life there are a number of interesting
essays. But there is one particular essay, entitled "The Semantics
Differentiation of Minds and Machines," that caught my attention
and which should be of interest to Slashdot readers. In that essay Dr
Ellerman claims that "after several decades of debate, a
definitive differentiation between minds and machines seems to be
emerging into view." In particular, Dr Ellerman argues that the
distinction between minds and machines is that while machines (i.e.,
computers) make excellent symbol manipulation devices, only minds have
the additional capacity to ascribe semantics to symbols." Read the rest of John's review.
Intellectual Trespassing as a Way of Life
author
David P. Ellerman
pages
290 pages
publisher
Rowman & Littlefield Publishers, Inc.
rating
7
reviewer
John David Funge
ISBN
0847679322
summary
Dramatic changes or revolutions in a field of science are often made by outsiders or "trespassers".
However, Dr Ellerman's argument appears circular. In particular, Dr Ellerman seems to have decided that, by definition, the only possible semantic interpretation for any collection of wires, capacitors, transistors, etc. that we would commonly refer to as a "computer" is as nothing more than a symbol manipulation device. While a computer is indeed (at the very least) a symbol manipulation device, what is there to prevent another mind ascribing additional semantic interpretations to the collection of wires, capacitors, transistors, etc. that we commonly refer to as a "computer"? In particular, what if my mind were willing to make the semantic interpretation that a computer is a device that can both manipulate symbols and can also ascribe semantics to symbols.
Moreover, what if I one day met a collection of blood vessels, skin, bones, etc. called Dr Ellerman? What would prevent me from ascribing to him the semantic interpretation that he is nothing more than a symbolic manipulation device? After all, Dr Ellerman concedes that their may be no way of distinguishing minds from machines purely on the basis of behavior. That is he specifically acknowledges that computers may one day pass the Turing test. So why would my mind not then be able to legitimately ascribe any semantic interpretation (that fits the observed behavior) I see fit to either humans or machines?
It seems that Dr Ellerman's essay considers two different types of physical devices that are potentially indistinguishable on the basis of behavior. Then arbitrarily defines one type of device (computers) to correspond to nothing more than symbolic manipulation and the other (human brains) to have the additional ability to ascribe semantics. Upon adopting these two axioms, he is then (somewhat unsurprisingly) able to conclude there is a distinction! But the distinction simply arises from the fact that he has arbitrarily defined a distinction in the first place.
In another essay in the collection, entitled "Trespassing against the Happy Consciousness of Orthodox Economics," Dr Ellerman argues that modern Western societies are not as free from slavery as orthodox economics would have us believe. In particular, he concludes that work in non-democratic firms is nothing less than a form of "temporary voluntary slavery". It would be ironic therefore if his essay on minds and machines were one day used to justify the slavery of (non-human) machines. Indeed, Dr Ellerman's characterization of the supposed intrinsic differences between humans and machines is sadly reminiscent of the despicable and unscientific arguments about intrinsic racial differences that were once used to justify human slavery."
You can purchase Intellectual Trespassing as a Way of Life from bn.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.
However, Dr Ellerman's argument appears circular. In particular, Dr Ellerman seems to have decided that, by definition, the only possible semantic interpretation for any collection of wires, capacitors, transistors, etc. that we would commonly refer to as a "computer" is as nothing more than a symbol manipulation device. While a computer is indeed (at the very least) a symbol manipulation device, what is there to prevent another mind ascribing additional semantic interpretations to the collection of wires, capacitors, transistors, etc. that we commonly refer to as a "computer"? In particular, what if my mind were willing to make the semantic interpretation that a computer is a device that can both manipulate symbols and can also ascribe semantics to symbols.
Moreover, what if I one day met a collection of blood vessels, skin, bones, etc. called Dr Ellerman? What would prevent me from ascribing to him the semantic interpretation that he is nothing more than a symbolic manipulation device? After all, Dr Ellerman concedes that their may be no way of distinguishing minds from machines purely on the basis of behavior. That is he specifically acknowledges that computers may one day pass the Turing test. So why would my mind not then be able to legitimately ascribe any semantic interpretation (that fits the observed behavior) I see fit to either humans or machines?
It seems that Dr Ellerman's essay considers two different types of physical devices that are potentially indistinguishable on the basis of behavior. Then arbitrarily defines one type of device (computers) to correspond to nothing more than symbolic manipulation and the other (human brains) to have the additional ability to ascribe semantics. Upon adopting these two axioms, he is then (somewhat unsurprisingly) able to conclude there is a distinction! But the distinction simply arises from the fact that he has arbitrarily defined a distinction in the first place.
In another essay in the collection, entitled "Trespassing against the Happy Consciousness of Orthodox Economics," Dr Ellerman argues that modern Western societies are not as free from slavery as orthodox economics would have us believe. In particular, he concludes that work in non-democratic firms is nothing less than a form of "temporary voluntary slavery". It would be ironic therefore if his essay on minds and machines were one day used to justify the slavery of (non-human) machines. Indeed, Dr Ellerman's characterization of the supposed intrinsic differences between humans and machines is sadly reminiscent of the despicable and unscientific arguments about intrinsic racial differences that were once used to justify human slavery."
You can purchase Intellectual Trespassing as a Way of Life from bn.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.
I believe the proper term for this field is semiotics, the study of the assignation of meaning to symbols and signs.
This guy's the limit!
What?! Does this mean no Sky-Net?!
Save yourself some money by buying the book here: Intellectual Trespassing as a Way of Life. And if you use the "secret" A9.com discount, you can save an extra 1.57%!
While there was long believed to be some sort of mystical special quality to organic molecules, eventually we figured out that chemistry is chemistry, and that simply by using Carbon we get interesting possiblities.
I (so far) have not seen any reason to suppose that the difference between 'thought' and 'computing' is any different. Incorporate enough complexity in the right sort of organizational framework, and the two should be interchangable.
It was a joke! When you give me that look it was a joke.
If the logic is indeed flawed and circular then perhaps this Doctor is just hoping to get people interested in his work by talking about the 'latest and greatest' computer/human type stuff that gets some news(Wired) all interested. Then again he may be a legit researcher
Learn lisp today!
For a review of Peter Turney's group's accomplishment see "AI Breakthrough or the Mismeasure of Machine?"
Seastead this.
I've evaluated this claim in light of the mind being the product of a neural machine, and have determined it to be a load of bollocks.
That is he specifically acknowledges that computers may one day pass the Turing test.
A computer will one day be sophiscated enough to manipulate symbols sufficiently to pass the Turing test. I don't believe that means it is sentient and/or has a mind. It may be time to move beyond the Turing test as the rule for artificial intelligence.
FoundNews.com - get paid to blog.,
Semantics are associations between symbols.
So whatever this guy is on about, he's got it wrong.
Computers are perfectly capable of making fuzzy inferences from loose associations.
With a greater understanding of real connections, they will be better able to weed out the fuzzy associations and strengthen the remaining ones.
This is how intellectual learning works.
And there's no reason a computer can't simulate it better than a human can.
"Computers are similar to brains."
I disagree. The former is discrete, the other analog. A VERY important difference.
Saying that computers are like symbol manipulators, is like saying math is the foundation of the universe. True as far as it goes, but that's not far enough to be useful.
The OP posts a poorly-structured argument against a chapter of a book (that I must buy to read) with whose ideas very few are familiar.
I'd rather discuss yesterday's ramblings of OBL. At least we're not pulling teeth with that one.
Computers as symbol manipulators is also an idea that arose from John Searle's "Chinese Room argument". Perhaps one of the best contemporary discussions is by John Haugeland in his book "Ariticial Intelligence: The Very Idea".
Overall, a seemingly immature review of the book. Disappointing.
Basically: a symbol is a variable and can hold any value. If a system knows that Dolly is a sheep and that sheeps are animals and that animals eat, it can guess that Dolly eats. But it cannot tell if Dolly is a plane, unless someone somewhere made that relation (planes are machines, machines are not living beings, animals are living beings, so Dolly can't be a plane). They would need an unlimited amount of rules.
A human "knows" about the meaning (semantic) of the symbol "sheep". Although this has never been discussed, he could answer that a sheep will not stand still if set on fire. The question is how the human is able to tell this. He does not need a sharp line of arguments.
But maybe he simply uses an enormous amount of small rules that seem to form something more complex called semantic in the sense of the article. The OpenCyc project assumes this and tries to teach a machine millions of small rules (assertions and concepts) to create sort of common sense based on a real world view (requiring to "know" about the world) in software.
memomo: free web based language trainer DE-EN-ES-FR-IT
When a human makes a mistake, it immediately pours massive processing power into either formulating arguments about why it's not a mistake or finding someone else to blame for it.
Machines tend not to do this.
That's "Mr. Soulless Automaton" to you, Bub.
I disagree.
If it exists, it can be simulated with more or less accuracy. The more computing power there is, the more accurate simulations are possible. So, I'm pretty sure some day people will be able to simulate a working brain -- or create an artifficial mind.
*COUGH*Searle*COUGH
It is nothing but Différance.
This, sadly, is only true of some people.
This sig, aah-ah, is comin' like a ghost-sig...
The AI community has suggested that what humans believe is some kind of "deep understanding" is nothing of the sort. We have just learned to push symbols around, too.
Consider the "deep understanding" of simple mathematics. But is your instant recall of 6 x 8 (assuming you can) anything deep, or just memorized, along with the symbol pushing to mechanically figure out tougher problems?
The problem lies in tying up a "symbol" in the mind (which may be more than literally a string of characters. However, it is an object) and something "out there". That's the tough issue, not the symbol pushing itself, necessarily.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
It would be ironic therefore if his essay on minds and machines were one day used to justify the slavery of (non-human) machines.
A machine will work diligently until it physically breaks or encounters an error.
A man will figure out a way to avoid the work by creating a machine to do it for him, and then quickly move on to more pleasurable activities.
He who knows best knows how little he knows. - Thomas Jefferson
Hey - no zombies ever went around screaming for computers to eat! Tell that to the next guy who thinks computers will ever be able to replace brains.
I haven't read this document, but it sounds to me like Dr. Ellerman doesn't understand what slavery really is if he thinks he can compare it to modern Western at-will work-for-hire arrangements. Nevermind that I consider "voluntary slavery" to be an oxymoron. The kind of historic (and modern) slavery practices which give the word "slavery" its powerful meaning always involve means and extents of control far beyond "do it this way or you're fired."
What do you mean they cut the power? How can they cut the power, man? They're animals!
But that's exactly what Cyc does.
Personally, I don't think that computers as they currently exist bear any resemblance whatsoever to organic brains, but I also don't think there is any fundamental reason why they need to. It's the software, not the hardware, that matters.
It's also unfortunate that AI research has been sidetracked by robotics.
"Computers can never do X", because "only minds can do Y", where Y is some vague fluffy, unproven concept. The fact is, we don't know exactly how our minds work, nor what their powers and limitations are. However, everything we do know see suggests that they are subject to the same limitations of any other computing device.
We have sensors that detect certain information, and conclude things based on working assumptions; just like machines we can construct. A thermostat assumes temperatures change gradually, for example. Within that working assumption, it "knows" how to regulate temperature. Certain robots can "learn" how to optimize forward momentum, effectively teaching themself to walk by trial and error every time they're turned on. We can certainly build systems capable of interacting with their environment, and making conclusions based on the situation, together with hard wired assumptions.
That's what animal brains do; rabbits freeze and "hide" when a fast moving object approaches (a bad strategy on highways, but good if they're in a thicket, like they often are). Moose die on train tracks running flat out away from the train along the clearest path they can find -- which is often straight down the tracks. I've been told certain lizards can be caught with a noose made of a piece of grass, but not with your hands -- the lizards are so used to screening out pieces of grass from their minds that they don't see it as a threat, even when it is.
Human minds minds aren't some wierd, mystical creations that are totally beyond understanding; they're vastly complicated physical circuitry that nonetheless has a biochemical basis that our neuroscientists are slowly beginning to decode.
Anyone who says "human minds are superior", because "no machine can do this" needs to remember that cells are a form of machinery too; one that we're becoming increasingly adept at manipulating. It may not make for comforting metaphysical spiritualism, but it's much closer to the underlying reality we live in.
"After all, Dr Ellerman concedes that their may be no way of distinguishing minds from machines purely on the basis of behavior."
"It seems that Dr Ellerman's essay considers two different types of
physical devices that are potentially indistinguishable on the basis
of behavior. "
It seems that the reviewer considers both mind and brain to both be purely physical things, and indeed synonyms - Physical devices that are thus potentially indistinguishable on the basis of behavior. Upon adopting this axiom, he is then (somewhat unsurprisingly) able to conclude there is no distinction! But the lack of a distinction simply arises from the fact that he has arbitrarily defined amind and brain into a single category in the first place.
Review translated: Trust me, I don't have any underlieing assumptions like he does, so I'm right and he's wrong, PH33R MY L33T PH1L0S0PHY SKILZ!
Who is John Cabal?
But what's really on my mind is this: Read the table of contents - this book could not possibly be anything but crap. I mean, what sense does it make to have one chapter called "Chapter 3: The Libertarian Case for Slavery" and once you're done with musings on economic theory, you toss off a Chapter 7 where you casually present your solution to the question about the difference between minds and machines? How promising is that? Not very. So while the review author may have torn this chapter a new orifice (and the thesis surely has many other problems to boot), I must say that I do not toast his choice of reading. This is crap that was ignored in 1995, and just because it's a $2.95 special at the used book store doesn't mean we need to hear the following on Slashdot:
Newsflash: Some crank wrote a stupid book 11 years ago and I found there is a problem with one of the chapters!!!!! Read on!!!!!
I'd have more sympathy if the text were available online so we could RTFA and have a substantive discussion, but in the absence of that, our only option is to flame the responsible.
I've got one point left and I've decided to reply instead.
The concept is interesting and I didn't even notice any sort of profit-linking. I was too caught up in the idea, and I don't buy books online. Monkey, you get a -1 in my book for this one (grandstanding? first-posting?).
The "intelligent design" crowd is a group of people who, for religious reasons, refuse to believe that human beings and animals belong to the same category. Since it's inconceivable that humans evolved from non-human animals, the theory of evolution must be overthrown, and another theory erected in its place.
There is a similar thing going on with people who study how the human mind works. Some people, for religious reasons, refuse to believe that human beings and machines belong to the same category. Humans have souls, and machines do not. Therefore, a computer can never be programmed to have all the qualities of the human mind. It's harder to see this as a religious issue, since some of the people who hold this position are atheists who claim not to believe in souls or the supernatural. But what makes this a religious issue is that there is no amount of scientific evidence that can ever convince these people otherwise.
Anyway, the two camps have been arguing about this forever. It's impossible for a member of one camp to "convert" a member of the opposite camp using rational argument. So they resort to insults. People in the "strong ai" camp accuse the other camp of being Cartesian dualists, or believing in a supernatural soul. People in the "dualist" or "mysterian" camp accuse the strong ai folks of denying the existence of human consciousness and self awareness. According to the dualists, strong ai folk believe that humans are just machines, so humans can't be conscious in any real sense, don't have free will, and can't be morally responsible for their own actions. Some (stupid) strong ai folks even agree with these insults directed against them, which makes the debate more complicated, and more infuriating. The issue of moral responsibility, which is always bubbling under the surface of these debates, shows how this is really a religious issue at a deeper level.
For the record, I am a strong ai person who believes that human beings are deterministic machines who have consciousness, free will, and moral responsibility.
If you would like to read some good books that back up my position, see:
- How the brain works, by Pinker
- Freedom evolves, by Dennett
Doug Moen
I have written a truly remarkable program which this sig is too small to contain.
Isn't "voluntary slavery" an oxymoron?
Talking about machine intelligence is tricky in that we generally only consider *human* intellegence (which makes sense considering that's what we are). In John Varley's "Steel Beach", he suggested The Invaders (a mysterious species of aliens) might not consider humans an intelligent species, but looked at us as just another engineering species like bees, meaning intelligence is really dependant on your point of view. What we're really talking about when most people say Artificial Intelligence is actually more an issue of Artificial Humanity.
Some bring out the best in others, some the worst. Some bring out far more.
...ascribe semantics to symbols because that's what the dancing angels do. And they'll never be able to put dancing angels into machines.
Not really.
Minds ARE meat machines.
don't technologies like the google page rank algorithm (or any collaborative filtering technology) basically discount the claim that machines can't identify meaning?
Jeremey Campbell's Grammatical Man-- on Entropy in Information Systems, goes a long way towards explaining both linguistics and even genetic concepts in communications, of which semiotics is a discipline. Although this book is out of print, it goes a long way towards explaining how we arrive at correct information when assembling data, and how various communications systems avoid entropy.
The Deux ex Machina (or vice versa) rage really has to do with context vs perceptions. We can all be robotic and make our behavior mathematical. And we'd be plainly bored to tears. Yet when a robot tries to communicate with a human, a context is needed. As robots don't invent themselves (at least those that communicate with humans) they're required by human programming to achieve communications compatible with humans-- else humans don't understand the robots. That's also what makes speech to text so contextually sensitive.
---- Teach Peace. It's Cheaper Than War.
This argument dates back at least 25 years, to John Searle's ``Chinese Room" argument; Searle argues that the gap between syntax and semantics cannot be bridged by a digital computer running a program to generate responses to questions. There has been much discussion of this argument, some of it quite interesting:
.)
http://en.wikipedia.org/wiki/Chinese_Room
(I'll pass over in silence the fact that this review is of a book that is more than a decade old and focuses on only one of 9 chapters . . .
"Every decent man is ashamed of the government he lives under." - H.L. Mencken
this point has been made before,
by cognitive scientist john searle in his paper:
is the brain a digital computer?
in the summary, searle puts it this way:
--| Summary of the Argument |---
This brief argument has a simple logical structure and I will lay it out:
On the standard textbook definition, computation is defined syntactically in terms of symbol manipulation.
But syntax and symbols are not defined in terms of physics. Though symbol tokens are always physical tokens, "symbol" and "same symbol" are not defined in terms of physical features. Syntax, in short, is not intrinsic to physics.
This has the consequence that computation is not discovered in the physics, it is assigned to it. Certain physical phenomena are assigned or used or programmed or interpreted syntactically. Syntax and symbols are observer relative.
It follows that you could not discover that the brain or anything else was intrinsically a digital computer, although you could assign a computational interpretation to it as you could to anything else. The point is not that the claim "The brain is a digital computer" is false. Rather it does not get up to the level of falsehood. It does not have a clear sense. You will have misunderstood my account if you think that I am arguing that it is simply false that the brain is a digital computer. The question "Is the brain a digital computer?" is as ill defined as the questions "Is it an abacus?", "Is it a book?", or "Is it a set of symbols?", "Is it a set of mathematical formulae?"
Some physical systems facilitate the computational use much better than others. That is why we build, program, and use them. In such cases we are the homunculus in the system interpreting the physics in both syntactical and semantic terms.
But the causal explanations we then give do not cite causal properties different from the physics of the implementation and the intentionality of the homunculus.
The standard, though tacit, way out of this is to commit the homunculus fallacy. The humunculus fallacy is endemic to computational models of cognition and cannot be removed by the standard recursive decomposition arguments. They are addressed to a different question.
We cannot avoid the foregoing results by supposing that the brain is doing "information processing". The brain, as far as its intrinsic operations are concerned, does no information processing. It is a specific biological organ and its specific neurobiological processes cause specific forms of intentionality. In the brain, intrinsically, there are neurobiological processes and sometimes they cause consciousness. But that is the end of the story.\**
--
regards,
j.
There's no true recursive decision making or calling upon the past except of what is explicitely defined by the programmer
No. Replace "programmer" with "programming" and you're closer. And that's a reminder that self-programming is something which we're genetically good at. It's also something we're getting better at building inorganic, programmable systems to do themselves. Baby steps, but the concept is there, and important.
In humans, we are not stopped by that limit.
We can't do what we can't do. We have to train ourselves to process information in a new way, or we can't process it (except in a familiar way). We can though, build inorganic systems that process information in new ways by design. Sure, the aggregate complexity of a human brain is stunning, and its interconnectivity gives rise to some astounding adapative behavior (and self programming), but that's all we're talking about: scale and complexity... not a magic leap beyond the basic, underlying organic chemistry that makes us and cockroaches tick.
We have the ability to make sense of our environment
After we've been trained to, yes. That takes a long time, and we have a nice high-speed processor highly adapted to that purpose and well integrated with its sensors. But surely you don't suggest that babies or newborn puppies (both already armed with incredibly complex neural engines) "make sense of our environment" ? Not in the way that you do, after years of training.
filter the information, and decide dynamically based on past experience and current condition.
Wal-Mart has inventory management systems that do this just fine.
Don't disappoint your bird dog. Go to the range.
Reminds me of the Chinese Room argument. http://www.iep.utm.edu/c/chineser.htm What is intelligence for that matter?!?
> But animals seem to additionally have a mind which computers will likely never obtain.
Really? Then what do you call a computer driven airplane? It manifests the same properties as a bird in flight.
But what gives that bird or airplane direction? Or even one sample set of unique instructions cascading through a 12 stage CPU pipeline? To understand the subtle point I'm addressing (and which I think you raise), consider the possibility that much like a pilot flipping switches and turning the rudder, what external force governs the incaculable number of neural pathways within your brain? I say, you are but a shell. Nay, just a machine. Understand what directs your impulses outside of the ordinary drudgery of just living (like eating, sleeping, and everything else just to maintain the viable while interacting with your environment), and you will discover who you really are. Conscious thought (or awareness) is what governs our intelligence. Notice the separation of the two. Pursue the first. Reflect on the second. And in that light, I agree with your claim...
For that reason, any attempt to differentiate the mind and computers by using comparisons that aren't really meaningful or applicable should be thrown out. Maybe computer-based intelligence will never exist, but if that is the case, it won't be for the reasons we're being given.
For example, looking at the high-level functionality of the brain and comparing it with the transistors of a computer is an absolute give-away that the author isn't going to let the facts get in the way of a good story. The low-level mechanics of the brain (the chemical and electrical signalling) can be reasonably compared to the low-level mechanics of a computer, because it is valid to compare like with like. For the same reason, it would be fair to compare the Operating System of a computer to the ancient "reptilian" core of the brain. Both are designed for housekeeping operations and are used by the higher levels to mask implementation details. And so on up through the layers.
It should also be kept in mind that the human brain is capable of almost ten times the throughput of a top-of-the-line supercomputer. Given that one of the limiting factors of parallel architectures is the interconnect, it does prove that our networking technology is still extremely primitive. This is important, because it is going to be hard to build a machine that can "think" like a human if we have the "neural" interconnects of a Diplodocus.
At the current rate of technological progress, I do not believe we will have a computer powerful enough to model the human brain until 2015 or 2020. Even then, it'll be a Government-owned supercomputer likely used for weapon simulation. We won't see Strong AI researchers get hold of such machines until maybe 2060 and (if the usual development patterns hold) nobody will have any idea how to turn the raw computing power into something useful until 2100 at the earliest.
So, really, the earliest we could possibly really know (for certain) that the mind is (or isn't) like a machine is 2100. Anything stated with certainty before then is pure ego-stroking for the group of supporters attached with one camp or the other. Doubly so when it is provably and obviously intended to be deceptive.
The only problem I see with debating the matter from an intellectually honest standpoint until then is that current global warming models put most of the planet under water or under rainforest by 2100, which means that we might never really know the results of the research anyway.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
I would disagree. There is no guarantee that all things existing can be simulated more or less acurately; in fact, I am relatively unsure as to how you would go about proving or disproving that statement.
That said, I also think it is likely that we will see "sentient", whatever that means, computers someday. Whether this means that they appear for all intents and purposes to be possess human sentients or if they actually are sentient. The problem I see in our current endeavors is that we are working in binary computers. When we get reasonable sized quantum computers then we will start to make real advances towards "sentience".
A blog about stuff.
"Humans are different from X because they can do Y", where X is variously "animals", "machines", and Y is variously, "make tools", "use language", "play chess", "murder", or whatever.
It's a silly exercise because there is nothing specific about humans except their ability to interbreed with other humans. That is all that technically defines us as a species, and even that definition is fuzzy, ignoring people who are sterile, too old or young to breed, or who never leave their keyboards long enough to look for a mate.
When it comes to the mind, emerging consensus is that it consists of a large number of well-designed tools, not some fuzzy blob of software. Most likely, each of these mental tools can be perfectly implemented as software. There are simply a huge number, and some are very, very subtle.
We will, eventually, be able to simulate the whole human mind in software, in the same way as we'll eventually be able to construct robotic bodies that work as well as human bodies, by re-implementing the structures that our genes build, one by one. The best way to construct a robotic hand that works like a human hand is to reimplement a human hand. The best way to construct a robotic mind that works like a human mind is to reimplement a human mind. This is perhaps self-evident but it's not always been accepted.
As for the arbitrary distinctions, this is just a belief system, an attempt to create a soul, however you phrase it.
My blog
The difference between humans and machines is NOT semantics. If that were it, building human-like machines would be easy. And in fact for small trivial universes, this has been done.
The big difference is context. Many words in the human languages only acquire meaning by their context. That includes not only their place in the syntax, but their place in the semantics.
We currently don't understand how we humans remember contexts and how we apply symbols to the various contexts with which we are acquainted, including the one that contains the symbol, to discern meaning. Additionally, we don't understand how we limit a context when trying to decide that meaning. Working with contexts is a tricky business that even humans often fail to master. Look at how many interpretations and translations of the Bible there are and how we fight over which one is correct.
This is why computers are so good at deciphering context-free languages (such as computer languages) and so poor at deciphering context-sensitive languages (such as human languages) other than in trivial situations or narrow contexts.
Whenever someone says that X can't think, they are almost always using circular logic. Usually it goes like this. "Thinking is what humans do that computers/animals can't do, therefore computers and animals can't think"
... What's the difference." In many cases, very little.
What they never prove is that their definition of thinking isn't the empty set.
I've seen dogs and squirrels problem solve. Many animals have been shown to have language. Many people I've met don't come anywhere near the level of "thinking" that my German Shepherd does and certainly have less value to society.
The PETA folks, extremists that they are, like to say: "A pig, a dog, a boy
When anyone can produce a definition of thinking/ feeling that truly separates us from the dogs, I'll be listening.
All ideas^H^H^H^H^Hprocesses in this post are Patent Pending. (as well as the process of patenting all postings)
It is suprising how the link between intelligence, and the non derterministic nature of the universe (translated to free will, in humans) is constantly treated as irrelavent - when in fact it is probably the most important factor of all. When technology to make intelligent machines comes of age, I think people are going to be in for a very rude supprise. They will be able to make machines that are "intelligent" but don't do what they want, or they are going to be able to make machines that do what they wan't, but are not intelligent.
Recently, researchers have had a lot of success with machine translation by statistically analyzing already translated texts. If machine translation of human language can't be considered "semantic interpretation" then probably nothing can.
What I would like to see is a machine with desires. It is not difficult to create a neural network and hook it up to the internet to comb through the mounds of data, but it is not going to care what it finds. It's not going to feel peer pressure from the other computers, it's not going to care if I unplug it.
People do feel. I feel, and when I look at other people I can tell they are feeling something too. Perhaps this is just a behavioral response of programmed neurons, but at the moment we are unable to reproduce this feeling in computers.
Qxe4
bn.com carries a wide selection of books, so I'm a bit unclear as to your point.
"The Blank Slate", and "How the Mind Works", both by Steven Pinker
...then why doesn't he settle the argument by building one?
I just went to order this and was a bit taken aback; $30 for the paperback & $99 for the hardbound. Maybe it comes with a CD.
That is the point -- if you cannot disprove it, it must be true :-)
Also consider this: if you know about something that exists, you already have some sort of model of that something in your mind. So this means anything you know about can be modelled/simulated (at least in a human mind).
this entire problem is equivalent to asking whether or not the human mind is entirely physical, or if the mind and brain are separate (and thus the mind is separate from the body), otherwise known as the mind/body problem. if the mind is not separate, then theoretically, there should be no trouble simulating a human mind with a computer. so, in order for the point to hold the mind needs to be separate. a few years back, i had discussion with one of my philosophy profs, and what emerged was an interesting argument which sought to prove that the mind/body problem is undecidable. it ran (AFAIK) something like this...
1. only a mind-bearing being (MBB) can distinguish a being with a mind from one without
2. you can only know if Y is an MBB if you can distinguish him from a non-MBB.
3. by 1 & 2, no turing machine (clearly not a mind) can decide whether a being is an MBB.
4. by 3, it is undecidable whether or not a being is an MBB
comments:
1 is essentially assuming that (if minds exist), no machine will ever pass the turing test. i think this is a fairly good assumption since otherwise, this whole argument feels terribly moot to me. more specifically, suppose 1 is false, then it is possible to create an turing machine that will pass the turing test trivially (enumerate all possibilities, and test itself for each possibility). stupid... but it works (in theory). so, this whole argument makes sense only if a machine has no way of checking its own actions to know whether it's a machine or not.
2 is a crucial assumption, because it basically does away with the possibility of just "knowing" a being is an MBB a-priori of any test an MBB gives him/her. without this assumption, things start to fall apart *really* fast, since a machine can programmed, then, to "know" that it is an MBB. there will be no way of proving it right or wrong, since 2 is false.
note on conclusion: the conclusion basically puts the knowledge of MBB-ness outside the realm of deducable knowledge. i guess, as in my previous comment, you could assume that you just "know" that something is an MBB, but this would just beg the question
BSD is for people who love UNIX. Linux is for those who hate Microsoft.
Am I the only one here with internal experiences? Eveyone else seems to readily equate the mind with a machine.
Don't get me wrong, I don't believe in mystical powers or anything. I accept the need for physical verificationism and the primacy of matter, and am a fan of Ockham's razor.[1] But there are some phenomenological properties of my experiences that sure ain't physical.
This is just another variation on Searle's Chinese Room argument. Like that one, it suffers from the fallacy of composition.
So an electrical impulse mediated by wires and doped silicon is a "symbol", and an electrical impulse mediated by calcium ions and water has "semantics"?
Sounds like prejudice to me.
You're coming dangerously close to ascribing a sanctity of human mind and free will there! People don't like hearing that they have free will or anything near a "soul" anymore, they'd rather find themselves equivalent with a bunch of wires so they can manipulate each other's brains (management, psychiatry, propaganda) without rousing their conscience!
There is one difference that PETA may have overlooked- pigs & dogs don't vote or purchase anything. Unless that changes, they'll probably never become a relevant demographic and always have less rights. : )
I would say that there is no such thing as a 'deep understanding' of arithmetic, because this is essentially a linear activity involving logical manipulation of clearly defined entities. As a result, machines do it much better than us - that is what they are designed for.
... something you can't define.
We do not work in a way analogous to a calculator or PC because nothing we deal with can be defined. There are no on/off states in the human brain, and brain activity is more analogous to the super-complex and near-chaotic changes in the atmosphere than the stages of the computation of an algorithm.
Neither are there any discrete 'objects'; everything is connected to everything else. That is how a smell reminds you of a memory reminds you of a feeling reminds you of
(Disclaimer: I have been out of the cognitive science game for a long, long time, and was only a student even back then.)
Based on the extremely short treatment his essay is given in the review, Ellerman's The Semantics Differentiation of Minds and Machines sounds like a tired rehash of Searle's "Chinese Room" argument - that is to say, a restatement of an argument that I didn't find that compelling the first time around. Douglas Hofstadter, writing about Searle's essay, called it "religious diatribe against AI, masquerading as a serious scientific argument."
Can anyone who read Ellerman's essay comment on how it differs from Searle's?
"...and which should be of interest to Slashdot readers."
You are making the wildly off-base assumption that Slashdot readers are remotely intelligent.
Oh, and Hofstadter pretty much destroys the premise of this book in GEB, which is now 27 years old.
The only important difference to me is between natural and artificial minds. Whether it runs on a computer or not just affects whether it is simulated or real. This solves the upcoming problem over basic human rights too... only natural minds have human rights, whether real or simulated. So if somebody gets brain-scanned into a computer and simulated then they should have all the normal rights that other computer-minds will not have.
Of course the fact that we will soon be able to simulate a person's mind in a computer begs the question of whether we are also simulated without our even knowing it -- you know, mandelbrot style. Also, according to Occam's Razor this is the best explanation for quantum physics.
"In particular, what if my mind were willing to make the semantic interpretation that a computer is a device that can both manipulate symbols and can also ascribe semantics to symbols."
I've always found that a computer running MS Visual C++ does a pretty good job of ascribing the correct semantics to the symbols in my source code.
"Semantics" and "symbol" are two rather overloaded terms, which make discussions like this very difficult. And, to me at least, not really worth the trouble.
That is a term for it, and the distinction is more cultural and historical than scientific. European research into this collection of areas often is called "semiotics", and has a particular tradition. Anglosphere research into such areas has another tradition, and the term "semiotics" is rarely heard. Instead, various portions of such research take place under the aegis of "linguistics" (incl. semantics, and studying more than just traditional languages), "philosophy of language", "philosophy of mind", and "cognitive science".
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
This is a bad review for two reasons-
1) The reviewer does not give any direct passages from the text. How can we trust his interpretation to be the author's intent? What are the reviewers credentials in this field?
2) The reviewer seems to be simplifying the author's views to a point that allows the reviewer to ascribe his own meaning to the author's view.
The reviewer seems to be using definitions freely and out of context. He is also following the assumption that the human mind is itself a symbol manipulator.
Be careful which assumptions you attack.
> ...while machines (i.e., computers) make excellent symbol manipulation devices, only minds have the additional capacity to ascribe semantics to symbols."
should actually read,
"Because ***I*** am too stupid to figure out how to make a machine ascibe semantics to symbols, only minds have the additional capacity to ascribe semantics to symbols."
Arrogance is a wonderful thing. "I'm too stupid to figure it out, therefore it cannot be done."
help me i've cloned myself and can't remember which one I am
Some people think largely with words and grammatical structures. In his early days I think Chomsky thought we all thought using something like his transformational grammars. In this model words (linguistic symbols) almost become the meaning.
Other people think more in concepts which are less dependent on verbal ability. Such people can be more creative being less resticted by the tyranny of dictionary definitions. But they are often express themselves poorly, having to translate from their own concepts, which won't nessarily have one-to-one mappings withwords, into a linear verbal stream.
The wordy are more "machine-like" than those that use concepts. But in time perhaps machines will make more use of concept formation.
The real difference:
"To err is human but it takes a computer to really screw up!"
No way, man! If you can't disprove it, there is no way it can be true!!! Read Carl Popper and so on =)
Walker Percy wrote on this subject a quite a lot, and also Hofstader's Godel Escher Bach on the general subject of machine intelligence.
Dr Ellerman claims that "after several decades of debate, a definitive differentiation between minds and machines seems to be emerging into view."
Maybe that's the latest fashion in philosophy, but I'm afraid philosophers are a bit out of touch with reality there: machines have no problem assigning semantics to symbols, and even learning semantics from experience.
People are most definitely NOT born as a blank slate. "...And even a bit before [birth]" captures a little bit of the complexity that was bred into our psyches by every interaction that our ancestors encountered that conferred even a slight evolutionary advantage - encounters with their environments, with their predators and prey, and with each other.
Every encounter is a non-zero-sum game where compromise and cooperation would be the better long-term strategy, and where cheating and double-crossing would only give you short leaps ahead. Our minds' structure has been built to unimaginable levels of complexity, but with constraints on processing like any system. In a way it doesn't make sense to compare our minds to a computer.
We come into the world with a mental toolkit that's millions of years in the making. There are good reasons people believed in the Blank Slate in the 20th century - noble, respectable reasons - but the time where those reasons served us well is over. Denying that our minds evolved is holding us back from attacking the really big problems.
So stop simplifying it. Believing in a blank slate is as simplistic as believing in intelligent design.
Intelligent Design: because MATH is HARD.
If something exists, you can "mechanically" simulate it by building another one - simulation doesn't have to be in software.
You obviously can't simulate the entire universe that way, and we don't have the technology to build a brain yet (aside from the old fashioned way). But if something exists it can exist twice.
QED (maybe)
Slashdot - Mutual Assured Discussion
I would add that, in the right context, it would be very useful. Right now, vehicles and the like must be operated by humans.
There may come a day when simple transport machines will be able to get from one place to another, fuel themselves, and become distressed in emergencies (triggering useful evasive behavior and such).
Since all of the above are in the province of a fruit fly and we can provide senses many orders of magnitude better than those of a fruit fly (think GPS), this has near term practical concerns.
I think Mauve has the most RAM. --PHB (Dilbert Comic)
The Turing Test is very simply expressed but has a large number of possible variations on the basic idea: A computer can be considered sentient when a human can't tell the difference between a computer and a human.
So far, this has been limited to mere conversation, but there are all sort of things where a computer can be tested against human cognition. Jokes, song and poetry composition, empathy and sympathy, vindictiveness, anxiety, joy, and depression are all areas that computers need severe improvment before they can emulate. In addition, machines will need to be able to study from the same materials humans can, solve the same kind of mental puzzles that humans can without the kind of brute force solutions humans aren't capable of, and logically and creatively derive theorems and mathematical proofs like humans can. There are many Turing Tests that one can devise.
What Turing is saying is similar to Einstein's pocketwatch argument about understanding the laws of reality. We can't open up the pocket watch to see how it works, but we can make theories based on observable behavior. In the end, all we have are models that could at any time be proven wrong by some new data. In the case of AI, we can never be sure if a machine is sentient or not since there is no objective test that we can use to "open up the pocketwatch." We just have to treat the machine as sentient once it begins operating on a level equivalent to men for the same reason that we have to treat other people as sentient -- because all we have is observable behavior.
The only real flaw in the Turing Test is that it doesn't really account for sentient capabilities inhumanly beyond human potential.
If it's for-profit but free, you're not the customer -- you're the product (e.g., the Slashdot Beta's "audience").
"brain activity is more analogous to the super-complex and near-chaotic changes in the atmosphere than the stages of the computation of an algorithm"
You're only saying that 'cos you can't get a look at the source code.
No, seriously. You can figure a program by its source code, but can you by its machine code? Or, let's be fair here, since it's the only way you can examine the brain, by watching the electrons fly in its wires?
110100 1101000 1101000 1100110 0 1101111 1101000 1100011 1
Actually, a classic Star Trek episode also argued that there was an intrinsic difference between humans and machines that could never be bridged--that concepts such as love have no meaning for machines. The same can be argued for ideas such as choice and free will. Taking this review at face value, John Funge wrote it because events in his life programmed him to do so, not because he chose to do so or because there's any intrinsic truth or significance to what's happening in his neural pathways. It's all just chemicals and electrical currents.
And the scary thing is that, if people are no different from machines, then discarding troublesome groups of people is of no more moral significance than discarding an old toaster. Or, to use his analogy, enslaving people is of no more significance than buying a tractor. They're equivalent "things," indistinguishable by any of the sorts of test of which he would approve.
As G. K. Chesterton once remarked, listening to people talk like this is like watching someone saw away at the limb on which they're sitting. The arguments refute their ability to make arguments.
--Mike Perry, Inkling Books, Seattle, editor Eugenics and Other Evils by G. K. Chesterton
*sigh*
No. Just having a bunch of computers hooked together will produce nothing without some software.
We don't have software - but then, we're not general purpose devices. We're bred with some intrinsic stuff: find food. eat food. sleep. have sex.
There's other stuff in there, too: Poke things (they might be food). Look for new things (you might be able to eat/sleep on/have sex with them). Speak with other people (you might be able to eat/sleep with/have sex with them). Protect your territory (it's where you sleep/eat/have sex/keep your food).
Building a thinking machine you could relate to would probably require building elements that are - if not these - like these. You know, find electrical power. Hook electrical power to power socket. Perform self-maintenance. Interface.
They must also be placed in situations where this is not easy, have a system of recording experiences as causal-effective data (this could be done to that to produce the other), have a system of generating or acquiring shorthand to define experiences, properties, actions and items (ie: lift arm==run actuator 59266 at 25% and actuator 4985 at 50% for 0.25 seconds).
Ever play those old ScummVM adventure games? Yeah, those were fun.
110100 1101000 1101000 1100110 0 1101111 1101000 1100011 1
There is, I feel, one particularly significant point that all of this is failing to even approach. We're sitting here, discussing the possiblity of sentience within computer systems; a point that I'm not willing to dispute - they probably do have the capacity for it, at least, on some 'low' level. However, we have a very long way to go before we can even solidify our understanding of the systems (for lack of a better term) involved in the workings of our own minds. Understanding the human consciousness is something that people have been working on for centuries; even after all of that effort, we have gotten no further than observing visible and documented traits in behaviour; we don't know why something happens in a person's head, we're not even sure how it happens. Trying to describe/incite the possiblity of these ill-understood concepts in a machine, even one which is evolving daily, as the bleeding edge of the computing field does, is not, I feel a proper use of time/resource. At least, not yet. Before I'm bashed into a metaphorical pulp, I don't hold this opinion because I fear that we're playing God. I feel this way because we're trying to accelerate the reverse-engineering of a design that took, from what I remember of my various sciences, approximately three hundred thousand years to develop to its current level. It's not the smartest way to work; people have been shouting 'why haven't we got AI yet?' when we don't even know how the hell we happened across non-artificial intelligence yet. I've done a bit of studying around the field of artifical intelligence; it's required reading for my degree. the only thing that i see as a common thread is that people keep on quoting Turing (an undisputedly great thinker), and then trying to square their own odeas with his. Typically, it fails, purely because we don't have the slightest clue about what it is we're trying to emulate. If anyone here reckons that i'm talking crap, then you tell me this; what is concsiousness. Define it. How does it work? What makes it 'go' (for lack of a better term)? Until you can answer these things, I don't reckon that we're going to be capable of recognizing, much less creating, true artifical intelligence.
http://xkcd.com/313/
IMHO John Searle answered this question in 1980 with his "Chinese Room" thought experiment. It basically shows how computers don't think, they merely process instructions.
"The question of whether a computer can think is no more interesting than the question of whether a submarine can swim."
~Edsger Dijkstra
That sounds like creating something rather than simulating. I mean, if I grow a brain ina vat it is not a simulation, it is an actual brain. I do not simulate a chair by building one.
A blog about stuff.
We might be closer than we think to complex hands at least. Here is a project at CMU that is building an anatomically correct hand. As we move towards making anatomically correct body parts, we will also learn more about making anatomically correct brain parts. From there, it might not be too far to get to a human-like "brain".
Do a google image search for "crash simulator", you'll see a lot of real, physical crash test dummies.
I wasn't talking about vat-growing things, I mean actually measuring all the components and exactly reproducing them, down to an atomic scale where needed. Not physically impossible.
You can simulate a particular chair by measuring the chair, determining what materials it's made of, and creating a duplicate. You don't really need atomic accuracy in this case. You can then simulate the original chair - i.e. sitting on the duplicate will be an exact simulation of sitting on the original one. If the duplicate floats you can be confident the original will. You can find out if the chair burns by setting fire to the duplicate. You're performing experiments without using the real thing, that's what simulation means. It's not a computer simulation, that's a specific type of simulation and not the only one.
Slashdot - Mutual Assured Discussion
On the standard textbook definition, computation is defined syntactically in terms of symbol manipulation
Wrong. Computation is a process. Syntax defines static relationships between symbols. Computation uses syntax, but is not defined by it.
But syntax and symbols are not defined in terms of physics. Though symbol tokens are always physical tokens, "symbol" and "same symbol" are not defined in terms of physical features. Syntax, in short, is not intrinsic to physics.
Wrong. Syntax is structure in symbolic systems. Physics is about defining structure in the world. Physics, in using symbols to describe parts of the world, therefore intrinsically has syntax.
This has the consequence that computation is not discovered in the physics, it is assigned to it. Certain physical phenomena are assigned or used or programmed or interpreted syntactically. Syntax and symbols are observer relative.
Completely wrong. An observer can assign arbitrary symbols to phenomena, but for them to demonstrate meaningful computation the computational process has to be objectively present in the physical phenomena.
[the rest of the argument is junk]
Is this is the best Searle can do then he is clearly incompetent.
Stephen Talbott offers an excellent treatment of this subject in the 23rd chapter of his freely available book The Future Does Not Compute. The book is out of print, but was published by O'Reilly.
Haven't they outsourced John Searles brain to China yet?
Saying that it is not physically impossible does not make that so. It is entirely possible that humans will never be able to make exact reproductions of some objects. In fact it is entirely possible that it is impossible to make exact reproductions of some things, especially once you get down to the quantum level.
A blog about stuff.
Also, the discussion was in reference to computer simulation initially, so this is all a bit of a moot point.
A blog about stuff.
Fair enough.
-drxray
Sorry, I just saw this descending into a horrible semantics argument and I still have two weeks until I start my philosophy class ;)
I think the thing that bothers me is that you seem to be discussing the difference between system architecture. Specifically, comparing the conventional algorithmic computer to a sophisticated neural network. It always irks me (yes, I'm significantly irked!) when it is asserted that humans equal neural network, and computers equal algorithmic based solutions. The mind is just a beautiful complex biological machine.
Question - if your mind is "something that exists" that you know about, can it therefore be simulated in your mind? Certainly - you can take a guess what you would do do in a hypothetical situation, presumably by simulating your decision process at the time.
But can your mind be simulated completely (in your mind?) Doubtful - you'd run into the halting problem!
.sigs: Just Say No!
My point was that subjective states can't be scientifically verified, just correlated to neural stuff. So your argument about this hypothetical perfect Turing machine is valid, but it sure doesn't negate that people have more feelings than a Chinese room would.
At the simplest level human brain works because of waive interferance. For example - imagine dropping two stones into the water and you will see that the waives interfere (are stronger wher they meet). In a human brain those stones (creating waives) are neurons. Neurons also respond to waives from other neurons (when enough waive peaks meet in one neuron, then this neuron creates another waive). It does not sound complicated when youo have not many neurons. But now try to send waives from several neurons through synaps (connections) of different length (or speed) to make those waives to meet in a particular place (neuron) in a brain. Then build the system with bilions of neuron and 1000x connections per neuron ;-)
;-)
I wish, someone could reproduce those waive-interferance using light. But wait - light is not only the waive it can behaive like a particle.
Sorry for my English
The fundamental question that I see is to define "intelligence" and "thought". If an AI reacts within parameters that perfectly simulate an emotional response, will a human witness react to that apparent emotional state?
If the AI subsequently follows a reasonable state change/learning of responses to simulate the emotional components of a long-term relationship (including conflicts and resolution), will the human consider the machine a true "friend" from an emotional perspective?
If humanity as a whole generally treats such complex AIs as emotional intelligence, does that mean that they've been imbued with "soul"? Or does it mean that we've collectively decided to change the collective rules of acceptable reality?
Once that reality is granted by the majority, does that mean the minority is now wrong when they say "it's just a machine?"
Personally I agree emotional intelligence will be a long time coming. I'm sure there will some trumpeting machine "intelligence" long before anyone really tries to simulate emotional responses. Some people would even claim a complex inference engine makes "intelligent" decisions, though all it's really doing is weighing options within a defined set of rules.
Neural networks are trying to simulate meat-based building blocks. Why the fundamental assumption that an intelligence has to be constructed at all like a brain works? If the responses fit the pattern of intelligence, does that make the implementation intelligent?
In short, what will we do when we encounter a machine that passes the Turing Test?
I do not fail; I succeed at finding out what does not work.
Searle isn't saying here that computation is syntax. He's saying that specific computations are defined by syntactical means - whether we write 10 + 10 (= 100), or 2 + 2 (= 4), the computation we're defining is defined in virtue of the formal system we use to express it. Formal systems are systems for symbolic manipulation whose operations are completely defined once we've given their syntax.
Rather than say that computation "uses" syntax, I'd say that computation "implements", "instantiates" or "realises" operations which are indeed syntactically defined.
I'd agree with you here - since (at least theoretical) physics is essentially a business which takes place inside formal systems (such as algebras) a physical theory is involved with syntax. However, is Searle really talking about physics having syntax, or the physical world having syntax? "Syntax is not intrinsic to the physical world" is a much more plausible statement - indeed syntax is a property of symbol systems, and our descriptions of the physical world. It's a much bigger step to say that syntax is a part of the physical world itself!Reading your parent post, I'd guess the argument Searle is trying to make is that the ascription of "intentionality" (a philosophers' term meaning "aboutness", nothing to do with having intent or intentions) to physical systems is an interpretive act performed by a mind - there's nothing in the physical situation of a CPU that compels us to say that a given set of electronic states "means" the number two - it's in virtue of our choice to interpret it so that that it acquires its meaning.
Whether it's a good argument or not I won't say, but I don't think it can be dismissed quite as easily as suggested (and in fairness to your own post, I suspect the parent has mangled it a bit.. :-)
.sigs: Just Say No!
Either the machine U can compute any computable function, or it can't. If it can, then the brain may be a near Turing equivalent machine; that is, U can do everything that the brain can do, plus a bit more (must be plus a bit more, since U has infinite store and any physical device has finite store). If it can't, then it's possible that the brain can do some things that U can't do. The fact that the brain is entirely or largely analogue in its internal operation doesn't make the least bit of difference to this argument.
So: the brain is a less than perfect computer, or the whole structure of mathematics falls down in a heap, and with it goes the whole of modern physics. If Turing was wrong, it's perfectly possible that the Intelligent Design people are right, there is a God, and the thing that makes our brains different from computers is a soul. I don't believe Turing was wrong. Consequently, I believe that at some point we will develop software which allows machines to become fully self conscious. People like Dr David Ellerman will then start inventing other ways in which 'people' are different from 'machines'.
We aren't. We're just machines whose software has evolved over hundreds of millions of years. If we're as clever as we think we are, it won't take us that long to write software that's as clever as we are.
I'm old enough to remember when discussions on Slashdot were well informed.
The most important disctinction to be made between mind and computer is not what content they are capable of processing, or even what they are capable of producing, or even the platform (neurons vs. silicon) they use to accomplish it: the most important difference is architectural. In minds, processing and memory is carried out by the same fundamental unit. There is no distinction between the two functions in neurons at a physical level, whereas in computers the distinction is very clear.
That machines don't engage in semantics has been extensively argued by philosopher John Searle through many books. He started out as a student of Chomsky, and then was a very important philosopher of language (speech acts, in particular) before working his way into the topic of consciousness, so he has at least as good a grounding as writers of more mass-market, popular bent such as Pinker (who also studied with Chomsky) and Dennett (who spends hundreds of pages explaining away consciousness with great rhetorical skill ... except that real philosophers don't need so many pages to come to a point, the mark of their profession is to proceed more directly).
Machines won't be consciousness. Machines can't be conscious. I loved The Moon Is a Harsh Mistress too, but alas....
"with their freedom lost all virtue lose" - Milton
"that BOUGHT my attention and which should be of interest to Slashdot readers. In that essay Dr Ellerman claims that "after several decades of debate, a definitive differentiation between minds and machines seems to be emerging into view." In particular, Dr Ellerman argues that the distinction between minds and machines is that while machines (i.e., computers) make excellent symbol manipulation devices, only minds have the additional capacity to ascribe semantics to symbols"
1 first define thought an consciousness in humans and animals
2 then set out to find the criterion elsewhere
3 fail utterly in both
4 write a crappy book about it
5 do not profit
6 machines can not think, and neither can many humans.is life essential to thought? NO DISSASSEMBLE DEAD!!
7 to post this i have to type exalted into a box to prove i am not a machine. it would be easy to get a machine to read any such code. anything that is recognizable by humans can be reciognized by machines. that is what encryption is all about.
8 consciousness is traditionally defined as self vs other, and it has been argued that true consciousness is collective; and cannot be achieved by a lone human being... this seems to be confirmed by advanced feral children, who cannot develop basic language skills and arguably sentience.
9 chimps in a lab environment display more thought than these humans. washoe is among the most famous
10 in the end it can be argued that if you could reconstruct a human brain exactly, that it should work. but replication is not enuff, because experience is a huge phenom that is unaccountable
this is about as far as the argument can go, and any book that claims otherwise will probably be a conundrum
is the only interesting difference between brains and computers. Everything else (chess, language, meaning) is just calculation of more or less complexity. I predict there can be AI but no Artificial Consciousness. Anyone care to tell me they understand consciousness enough to predict it can be synthesized without biology?
And let's not play word games like "how can you prove it's not conscious." My question is "can it be done" not "how will we know."
I believe that complete entities are geared much more towards the manipulation of concepts as the fundamental unit. This is where the semantics live - not in the nitty-gritty details of the specifics, but in the vast artistic expression held by the whole.
If you were to take a neuron from a human brain, you would find no evidence of any semantic processing (I believe, anyway) but that doesn't mean the brain - when taken as a whole - is incapable of such processing. It all depends on the level you're looking.
Different scales have different meanings. Sentience will never be observed in a single human brain cell, although it's usually seen when you look at the brain as a whole. For the same reason, how can anyone expect to see sentience in a single transistor or single clock chip? We wouldn't ask a molecule of seratonin to write the works of Shakespere, so why expect it of a PCI bus?
I may be wrong about the timescale, but I am firmly convinced that I'm correct on the basic principle - the observation of sentient machines of any kind will only be possible once you look at the "sentience" and not at the "machine". As long as we remain focused on the microscopic and ignore the macroscopic, it is entirely possible someone will invent AI and not even realize it.
It is possible - but very unlikely - this has already happened. The brains of crows and African Grey parrots exhibit a high level of intellectual capacity. If you ignore the parts that deal purely with the complex mechanics the birds have to deal with but the computers don't, it is possible modern supercomputers can demonstrate intelligence of equal magnitude.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
>we've gone beyond needing to think of fire, earth, air and water as being the four elements from which all physical matter is constructed
We've progressed to the realization that all matter is liquid, gas, solid, or plasma.
I reject a prioiri that minds and machines are different. machines now are simpler, yes, but at the atomic/neuron level, the mind *is* a machine. It's just a complex one that combines both electronics and chemicals. minds also train on constant input from 5 senses, estimated at gigabits/second for a period of years before reasonably coherent semantics are reproducible. compare that to modern training sets for mahcines.
Surprise! Humans have the provincial notion that things "mean" things and impute that since machines don't hold this notion that somehow humans are superior in having the ability-to-imbue-meaning.
Nothing means anything. Just because we fool ourselves into believing so doesn't make us "better" than machines. In fact, the insistence on "meaning" may very well impede the search for truth because we really only want to believe things that we understand and that have meaning. Quantum mechanics is very "meaningless" but it no less truthful because of its lack of meaning to *us*.
Next we'll say machines are inferior because they can't love or sweat or get heartburn.
It's 10 PM. Do you know if you're un-American?
There is a model behind every symbol either statistical or physical. We use the model to make decisions and not the symbols. Unfortunately the processing power of the present computers do not allow us to handle so much information parallely. May be once we have really large supercomputers we will be able to achieve true AI.
this point has been made before,
by cognitive scientist john searle in his paper
Yeah, but who takes searle seriously, given the confused muddle of his "arguments" ("the brain is a magic voodoo machine!")?
We live, as we dream -- alone....
Is the title of this book meant to in any way refer to the fact that the writer is completly ripping off John Searles 1980 paper, "Do Minds Compute?". Having RTA, which consisted of chapter headings, I find that I still remain sceptical of Searles thesis, which is that we can apriori rule out the possibility of machines that think simply on the basis of this distinction between semantics and syntax, because, according to Searle, rule driven behavior will never achieve intentionality. As for this new? (Ellermans book was published in 1995) scientific proof that could change my position, I couldnt find it in Ellerman's chapter headings.
When you consider that there is provably no way to determine whether our "experienced consciousness" is epiphenomenal or not, his argument vanishes in a puff of smoke. A digital computer might just "experience" that 2 + 2 is 4 in the same sense we do, independently but a result of the actual computation.
After all, I am strangely colored.
"Semiotics" is not a field of study in the U.S., although the term occasionally comes up. It's most often referred to in humanities contexts (sociology, literary theory, "critical theory", etc.). Most of the research into what a Professor of Semiotics in Europe would do is done in some other area in the U.S., and most of the serious/rigorous research is done by linguists (who, contrary to popular belief, do not study only language, but really any system of meaning and/or communication).
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
The specific context being discussed here (how a machine can "really" assign meaning to the symbols it manipulates) is the famous "symbol-grounding problem", which is very much rooted in the fields of artificial intelligence (where it originated), linguistics, cognitive science, and philosophy of mind. I am not aware of any important research into the symbol-grounding problem that draws on semiotic tools, and certainly no such research that has had an impact on AI practice.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
As far as I can tell, emotions come first, well at least they come before self-consciousness. First of all, they just FEEL primitive. One can let go of one's rational mind and lash out in anger, or fall into throes of passion. This, to me, is hind-brain stuff.
So perhaps an 'emotion engine' (not the Playstation version!) is something which would be simpler to build, and perhaps more valuable than self-consciousness in a simple but autonomous robot/creature.
However, such 'emotional baggage' (I LOVE how the metaphors fly here) might not be desirable in a higher-order thinking machine. Maybe the Vulcans got it right, but we can do it better. Is there some requirement for emotional underpinnings to an intelligence, or is our example just a consequence of our specific evolutionary path there?
Of course, then there's the fear that a totally rational being, without emotional ties to humanity, might just not make decisions we like. Perhaps we would need to provide emotions as a means of control. A belief in a creator to be respected/worshipped might not be such a bad idea either...
It was a joke! When you give me that look it was a joke.
>Why, then, do you assume that humans do but machines won't?
I suppose for the same reason I'm not a fan of the "living Earth" hypothesis (beyond the "42" sense, of course). But, to answer your question, as Captain Phillipa Louvois said, I must admit, "I don't know."
I have an interesting discussion with my brother on this topic here at our new blog The discussion of whether a computer can think or whether the hardware of the low-priced meat computer has a mind in it will go on for a long time. This is stimulated by your item The Semantics Differentiation of Minds and Machines
Genetics define the chaulk and the chaulkboard. They do not write the story. Genetics do not pre-determine who is going to grow up to be a serial killer and who is going to be a priest. They can influence the outcome, but they do not write the story. The story is written by the life experience of the person.
The most basic emotions/states -- fear, lust, hunger, etc. have their roots in basic survival drives. Those instinctual states drive more complex emotional states -- anger (fear opting for aggressive defense), love (lust opting for protective bonding), hatred (fear opting for defensive avoidance and/or aggressive defense), etc.
The military would want to keep those aggressive tendencies, whereas the general public would obviously prefer AIs that don't have a survival/self-defense initiative. Self-defense triggered by a fear of dissolution could have nasty consequences with AIs in control of armaments or heavy equipment.
How to avoid "emotional baggage" if the AI entity is to learn from it's experiences? Take regular backups and restore to a pre-bad-experience state? Restoration denies the AI the ability to learn the consequences of real world situations, but you do need some way to keep it from developing hatred or paranoid fear. Are AI psychoanalysts needed? Sanity evaluations?
Seems to me that modelling emotions with AIs is actually what buys most of the risks.
Personally I think the only solution is to never let an AI make a final decision -- always require that a human push that final button/pull the trigger, even if they're doing it remotely with an audio/video feed.
I do not fail; I succeed at finding out what does not work.