Artificial intelligence has sputtered?

Mapou Whether or not a bee has a soul and I suspect it being a living thin it does have something like that and so not a robot STILL doesn't make intelligence in the brain bits. The bee is largely a creature of memory. No personality at all or changing its mind and way. The bee is not highly intelligent. Its just a bug with a memory. The neurons are minor details of its thinking being connected to its body. There is no reason to see it as the origin of intelligence. Thats just guessing and presumption. Robert Byers

Byers, you don't know what you're talking about, sorry. Even a honeybee is a highly intelligent organism and yet it has only about 1 million neurons in its tiny brain. I don't remember anybody suggesting that a bee's intelligence is in its soul. Do bees have souls? Mapou

Mapou AI is not intelligence. there is none now. its all just memory. Search engines of data. Intelligence is unrelated to memory. Memory is just a aid to a intelligent being. Even animals. A machine is just memory. It has no idea about anything and can only do what is already been explained to it by memory. Its ho hum. nothing is coming to impress bible believing Christians. Its just a better mouse trap and not a mouse. Robert Byers

During that time, During that time, the rest of the (most of memory) is asleep. This is reduces the computing requirements by at least two orders of magnitude, in my opinion.

Haysoos Martinez! I must have dozed off at the end of my last comment. LOL. Let me try that again:

During that time, the rest of the cortex (most of memory) is asleep. This reduces the computing requirements by at least two orders of magnitude, in my opinion.

Mapou

I wrote @21:

The number of possible patterns is indeed very large but an intelligent system does not have to know them all. A small subset is sufficient. Not everybody needs to be a scholar in every field of knowledge. Nor is it possible. This is why humans specialize.

Let me add that the brain's memory is organized hierarchically. In other words, complex patterns are composed of simpler patterns. This makes for an incredible amount of data reuse. What this means is that data compression in the brain is as good as it gets. Intelligent machines will also use hierarchical knowledge structure and will benefit from the same space-saving compression. A common criticism of AI is that the computing power that will be needed to have human level intelligence is immense. In my view, it is not nearly as bad as some of the experts make it sound. The brain is an extremely focused machine. It can think of only one thing at a time, i.e., only one branch of the tree of knowledge is active at a time. During that time, During that time, the rest of the (most of memory) is asleep. This is reduces the computing requirements by at least two orders of magnitude, in my opinion. Mapou

MrMosis @18:

I anxiously await finding out what they are capable of. But I can’t help feeling their having human-level intelligence will always be just around the corner

I hope you'll be pleasantly surprised when they show up on your doorstep. :-)

It seems to me that the novelty is too great, and there just aren’t enough computing resources to match the infinitely large number of possible patterns, and to bring them into a cohesive “understanding” with which to navigate living with human beings in their world.

The number of possible patterns is indeed very large but an intelligent system does not have to know them all. A small subset is sufficient. Not everybody needs to be a scholar in every field of knowledge. Nor is it possible. This is why humans specialize. In my opinion, if we could emulate the intelligence of a mouse and apply it to useful tasks, it would revolutionize society because our mouse-brained machines would perform over 99.99% of all human jobs. Not to belittle human intelligence, mind you. I'm just making the point that we don't need to have the full capacity of a human brain to transform civilization. But I can assure you that, if we can do mouse level, we can almost as easily do human level.

Translated: My guess is, they will grow to be increasingly useful but often also annoying and/or a laughingstock.

I doubt it. Intelligent machines will be so cognizant of their surroundings as they pertain to their functions, the efficiency and ingenious ways in which they go about their duties will astound all. Mapou

As the wonders and marvels of AI increase, and the capabilities of thinking machines grow, all of these things, all of them, testify to the incredible designs of their creators, mankind. I don't know why anyone or how anyone could believe that we are establishing that the processes of intelligence and thought are something ho-hum and routine based upon this evidence. It is only a dramatic declaration that the power of design is at work. OldArmy94

Mapou @13, It is understandable that you are not aware of the LHC computer grid AI which recorded approx. 600 million particle collisions and constructed a digital pathway of complex decays and sorted through up to 20 petabytes of data, ran simulations , did statistical analysis to recognize a 5 sigma event, and controlled the magnets and ran the LHC circuits flawlessly. selvaRajan

Mapou @17:

...how speak a human language and using common sense...

I envision presenting challenges to our household bot, just for fun. Tell it to get lost. Delete itself. Bugfix itself. Upgrade its firmware without taking any time off from work. Tell it to divide by zero or I will kill myself. Ask it what it can't do, and then tell it to do it. Ask it what it can not (in principle) understand, and command it to understand it. Tell it to keep "trying" if it complains. Introduce bountiful amounts of new and novel noise into its input streams, but simultaneously demand in a loud, focus-inducing voice that it not be distracted. Tell it so finish its chores by 8:00 pm, but not before the cows come home. Tell it to wake you up when the rooster crows. The next day tell it to wake you up when the horse crows. (better set your old alarm clock) As much ambiguity and distracting noise as there is in language, it's even easier to come by in navigating, interacting with, and engaging physical reality, particularly where physical reality meets language spoken by human beings. I anxiously await finding out what they are capable of. But I can't help feeling their having human-level intelligence will always be just around the corner. It seems to me that the novelty is too great, and there just aren't enough computing resources to match the infinitely large number of possible patterns, and to bring them into a cohesive "understanding" with which to navigate living with human beings in their world. Translated: My guess is, they will grow to be increasingly useful but often also annoying and/or a laughingstock. MrMosis

CentralScrutinezer @15:

Mapou: I strongly disagree with the Christian fundamentalist viewpoint according to which we cannot achieve human-level intelligence in a machine because the machine is not conscious and only a body imbued with a spirit can be intelligent. This is nonsense, in my opinion. You will be amazed at what machines will be able to do.

Mapou, I like a lot of what you write, because you’re an unhinged lunatic like me. But on this one I must cry foul. What evidence do you have that genuine human-like intelligence can exist without consciousness “connecting the dots”, so to speak?

OK. Notice that I did not say 'human-like' but 'human-level' intelligence. What I mean is this: I believe that, given the proper motivation, future intelligent machines will outperform humans in a lot of complex goal-directed tasks that require intelligence. Examples are, manual dexterity, learning how to walk, how speak a human language and using common sense. However, machines will never be able to appreciate beauty, music and the arts. For that,you need a conscious spirit. However, even without a conscious spirit, the behavior of the coming intelligent machines will be so human-like in so many ways, that many will swear that the machines are conscious. That part scares me. On what do I base my opinion? Well, there is evidence to support the notion that intelligence is a matter of degree. We already have intelligent machines with different levels of capabilities. There is no logical reason to suppose that their level of intelligence cannot be improved to the level of humans. I'm also a private AI researcher and I know about stuff that will be coming out (in the not too distant future) that will knock your socks off. I'm not boasting or anything like that because I did not come up with it. Christians need to wise up with regard to AI. Otherwise, they're going to miss out on something really big. Something huge. Mapou

Oh good grief, my apologies - I just posted to the wrong thread! RDFish

Hi gpuccio,

biological objects look designed. Indeed, even darwinists like Dawkins admit that.

Agreed, where "designed" means "something human beings build".

In the past centuries, that was not only the opinion of the common people, but also the opinion of scientists. In no way that has been an obstacle to the development of biological sciences, as far as I can say.

I'm not arguing about what presents an obstacle or boon to biological science; people can proceed in science with any philosophy they choose, and they always have.

Now, I have been discussing neo darwinism for many years here, and an important point I have increasingly realized is its connection with another modern scientific theory, which I will call here “strong AI theory”. Some have objected to my use of the term, so I will make clear from the start what I mean by that: Strong AI theory: the theory that conscious experiences (subjective states), including the understanding of meaning (Intelligence) emerge as the result of particular configurations of matter. IOWs, consciousness is an emergent property of some specific complexity of the software.

I think the connection between strong AI and ID is quite obvious and has been commented on frequently. As I've explained, nobody has any way to decide if strong AI is true or not; nobody knows what might be necessary or sufficient conditions for conscious awareness to arise.

Always for the sake of clarity, I give here an explicit definition of what I mean by neo darwinism: Neo darwinism: The theory that all the information present in biological beings is the result of RV + NS in beings capable of reproduction.

I have no interest in discussing the claims of Neo Darwinism, as neither of us adhere to it.

We could call it materialistic reductionism (materialistic, because matter is the origin of all, and reductionism because everything can be reduced to matter and to its known laws, if only physicists will complete their part as soon as is desirable :) ).

I'm not a reductionist nor a materialist, so this part doesn't interest me either.

There is indeed a third important aspect of the paradigm: a particular philosophy of science. Indeed. to protect the paradigm from some important objections, it is extremely important that science: a) Be considered the only real source of knowledge b) Be strictly redefined so that only certain reasoning may be admitted. In particular, only those reasonings where everything is explained, or will certainly be explained in the future, by the laws of nature as we presently know and understand them must be allowed to be considered “scientific”.

I don't believe either of these statements regarding science, so this has nothing to do with any of my arguments either.

The simple problem with the paradigm is that it is based on three very strong intellectual lies: 1) Neo darwinism is not a good theory for the empirical fact of biological information. It is logically inconsistent, and has no empirical support.

Yes, I'd say evolutionary theory is logically consistent and has a huge amount of empirical support, but none of it suggests that any known evolutionary mechanisms are actually responsible for complex biological form and function.

2) Strong AI theory is not a good theory for consciousness and intelligence. It has no logical justification, and empirically it is a total failure.

"Strong AI" is not theory, really - it's just the hypothesis that a computer can be conscious and have general human-like intelligence. I believe this hypothesis cannot be evaluated against empirical evidence, and so it remains in philosophical debate.

3) Methodological naturalism is the opposite of good science. It is a way of pre-defining what we can know according to what we presently know. And scientism is only a very extreme and narrow theory of knowledge: giving science an unwarranted supremacy in cognition, it really deprives it of its best features, of its desire to understand, of its humility.

I do not adhere to methodological naturalism, because I don't believe "naturalism" is well-defined. I do not adhere to scientism either. I do believe that scientifically vetted results ought to be held to be more trustworthy than non-scientific results.

So, let’s look at ID theory as a paradigm. Indeed, it is a paradigm. It is not a narrow theory. In a sense, its purpose is not to explain particular things, but to give the right approach to explaining them.

As I've said, I'm not arguing about whether or not certain philosophical beliefs lead people to more or less productive research avenues. People with all sorts of worldviews manage to do science.

2) It [The ID paradigm] is strongly supported by known facts.

I'm not talking about paradigms here, but as I've explained, no known facts actually demonstrate the (probable) truth of any specific ID claim, nor any of the supporting claims you make regarding philosophy of mind. No significiant scientific inroads have been yet on the mind/body problem, understanding qualia, language understanding, or the nature of conscious awareness. If ID was really a scientific endeavor, these are the areas in which it would conduct research. There are people who perform research with possible implications for the mind/body problem, free will, and other things (look up "experimental philosophy"); however, as far as I know, nobody who identifies with ID research has participated in any of it! Again, ID critically relies on particular metaphysical positions regarding mind, but ID authors fail to even mount an effort to provide empirical evidence that any of their positions are true. Where are the ID authors doing the work to determine if NDEs truly support the notion of conscious awareness independent of the body? Where are the ID authors seeking to investigate whether or not our voluntary actions free of antecendent physical cause? These are the assumptions one must accept in order to believe in ID, and the topics come up regularly in this forum, but there are no ID researchers who care to even try and answer these questions scientifically. Perhaps they are afraid that if there really are critical experiments that would answer these ancient questions once and for all, they might not get the answers that support ID?

Up to now, each new day has given us new interesting biological facts that completely support the ID theory.

No, there is absolutely no evidence that provides a shred of positive evidence for ID theory. All we have is evidence that current theories are inadequate to explain our observations, and you and I already agree on that. Not only has ID never provided a shred of evidence regarding the existence of any sort of intelligent agent that might have been responsible for first life, but ID doesn't even try to demonstrate that non-living intelligence (or intelligence that is not embodied in a complex physical mechanism) is even theoretically possible.

Now, in the light of all this, I would say that you position of “we really don’t know” has very little sense. You stick often to statements that we “don’t know” things. In particular, we don’t know what consciousness is,...

Yes I am of the very strong opinion that certain questions regarding the nature of the mind and the origin of the universe and first life remain unanswered to date. I've explained here that we do each know what conscious awareness is (again, it is that which we lose when we fall into a dreamless sleep).

... we don’t know if it has a casual role or if it is only an epiphenomenon, we don’t know how conscious agents produce complex things (but they do), we don’t know how biological information was generated, how life began, and so on.

Yes, that is correct.

OK, we don’r know. I can agree. Indeed, we really don’t know anything, because our human knowledge is not absolute.

You keep falling back on this, but I consider it a dodge. I am not being neither hyper-skeptical nor credulous here: I neither deny the possibility of knowledge nor pretend that any knowledge can be absolute. Yet some conclusions have been demonstrated to be true with sufficient evidence that it is unreasonable to doubt them.

On the contrary, you have great certainty of other things (that germs are the cause of diseases, even of the Big Bang). I don’t know if you would accept quantum theory as science, or stick with Einstein inspired doubters.

Certainly germ theory, Big Bang theory, and most certainly quantum theory are very well empirically tested. You need be careful when speaking of the truth of quantum physics, though. On one hand it is the most well-tested theory ever developed, is found to be correct 100% of the time, with an unbelievable level of precison (up to 13 significant digits). On the other hand, nobody knows how the QM mathematical formalisms should be interpreted with regard to ontological descriptions, and any particular view about this is merely speculation.

Good science is: I don’t know, but I try to understand. I try hypotheses, theories, explanations. I test them.

Yes, good so far...

I choose what works better. I have no fear to try explanations.

No, not exactly. If you find five explanations for some phenomenon, but you can't actually test and confirm any of them, then you have no "best" theory, even if you like one of them better. You simply have no succesful theory yet. Before Special Relativity, nobody could figure out how the Sun could possibly put out so much energy for so long - everything we knew about physical chemistry suggested that it should have burned out in short order. And before Quantum Theory nobody understood why black body radiation didn't emit an infinite amount of energy in the UV spectrum. The answers simply were not yet known, and the correct answer to those questions was acknowledged to be "We do not know". Scientists managed to solve those mysteries, but we haven't solved all of them. There are still questions without empircally supported answers, and the ones we discuss here are among them.

So, with your “don’t knows” you are simply ignoring that almost no one in the world would agree that we can simply say “we don’t know anything” about the problems I have cited.

Oh I doubt that, obviously, but even it were true I would still be certain my position was the correct one. Of course we can say we don't know when the fact of the matter is that we cannot justify any of our beliefs on the matter!

Indeed, we know a lot of things.

Yes! We do know a lot of things, but not everything.

So, let’s “teach the controversy”. Not the controversy between ID and neo darwinism, not in the schools. But that human thought is based on controversy, always has been, always will be. That its best fruits are the fruits of controversy. Let’s teach that, to ourselves and to others.

By all means! Just don't teach that there is scientific evidence that ID Theory is a well-defined explanation of first life that has empirical support, because that is not true. (And yes of course, refrain from teaching that evolutionary theory has successfully explained the origin of biological complexity). Rather, teach that there remain deep and wonderful mysteries, and that we need the smartest and most curious people to continue investigating all of the questions that fascinate us!

You see, you often answer me that I cannot say something because many would not agree. We cannot say what consciousness is because that philosopher says one thing, and that other says another thing. But that’s exactly why I say things. It’s because they are controversial. It’s because I say what I believe to be true, and knowledgeable, and already well supported by facts and by reason, And still it is controversial. Yes, I could come here just to remind everybody that germs cause diseases. But how boring! :)

Or you could tell everyone that reading tea leaves can predict the future, and I would say that isn't boring, but it is neither a scientifically vetted result. ID theory is not supported by observable facts and reason. Instead, all that is supported by facts and reason is the inadequacy of evolutionary theory. ID theory takes that and makes an entirely unscientific leap of faith, off into philosophical speculation, and pretends that it's all still science. It's just not. Cheers, RDFish/AIGuy RDFish

Mapou: I strongly disagree with the Christian fundamentalist viewpoint according to which we cannot achieve human-level intelligence in a machine because the machine is not conscious and only a body imbued with a spirit can be intelligent. This is nonsense, in my opinion. You will be amazed at what machines will be able to do.

Mapou, I like a lot of what you write, because you're an unhinged lunatic like me. But on this one I must cry foul. What evidence do you have that genuine human-like intelligence can exist without consciousness "connecting the dots", so to speak? CentralScrutinizer

AI research has sputtered because there has been no real advance in AI for decades. All the hot new techniques they keep talking about (deep learning, Bayesian statistics, neural networks, etc.) are not really new. They were introduced in the 90s or even earlier. The only thing new is the super fast computers that Google, IBM, and others are using to perform massive parallel processing on big data. This could not be done before but it's pretty much the same AI model. Mapou

Artificial intelligence is well and healthy. AI is nothing but self learning from patterns and environment stimuli- and this is evident in large amount of software and machines.In fact at this very moment, it is learning the algorithm to sort zillion emails, zillions of text messages, zillions of voice samples and correlating words with self evolved(based on patterns) threat database! Unless you equate AI with only consciousness, there is no reason to believe AI has sputtered . selvaRajan

The bible says we are and think with a soul. nothing to do with particles or systems of thought. nothing to do with logic.

Come on, Byers. This is hogwash. Mapou

Since they believe our thinking is just a series of particles in combinations then they reasoned AI should be doable.

AI is doable. It's just not conscious intelligence. Intelligence does not imply consciousness and vice versa. I strongly disagree with the Christian fundamentalist viewpoint according to which we cannot achieve human-level intelligence in a machine because the machine is not conscious and only a body imbued with a spirit can be intelligent. This is nonsense, in my opinion. You will be amazed at what machines will be able to do. Unless Christianity fully embraces machine intelligence, it will be left behind, looking like idiots when machines begin doing almost everything humans can do. When truly intelligent machines arrive, they will look and act conscious because we have erroneously equated intelligence with consciousness. Then the materialists will waste no time in deceiving the world into believing that the machines are conscious beings and deserve to be given human rights. Heck, there is already a concerted effort to make us believe that all animals and even rocks are conscious. So yes, AI is doable. We have intelligent machines right now that can learn on their own and perform complex tasks. Sure, they're not at human level yet, but they are nevertheless intelligent in their limited domain. Mapou

Mapou I read your link to this Dreyfus. Never heard of him and seems its just one evolutionist nitpicking on others in regards to AI. The interesting point for creationism is how he was attacked by the others for his criticisms which are now credible. I see this always in things called science. Any new idea is ATTACKED and not just disagreed with. Always these yahoos insist they know the truth and with passion rejects any new ideas. This is what happens in the modern revolution against evolution and in proposition to the evidence of a creator in nature. People who don't accomplish are the most vocal in opposition to pioneers. Right or wrong pioneers but it says modern science is interfered with by too many people who prevailed in memorizing things in their late teens and early twenties. Robert Byers

Mapou I read your link to this Dreyfus. Never heard of him and seems its just one evolutionist nitpicking on others in regards to AI. The interesting point for creationism is how he was attacked by the others for his criticisms which are now credible. I see this always in things called science. Any new idea is ATTACKED and not just disagreed with. Always these yahoos insist they know the truth and with passion rejects any new ideas. This is what happens in the modern revolution against evolution and in proposition to the evidence of a creator in nature. People who don't accomplish are the most vocal in opposition to pioneers. Right or wrong pioneers but it says modern science is interfered with by too many people who prevailed in memorizing things in their late teens and early twenties. Robert Byers

Since they believe our thinking is just a series of particles in combinations then they reasoned AI should be doable. Simple. It was the error of presumption about human thinking. The bible says we are and think with a soul. nothing to do with particles or systems of thought. nothing to do with logic. The computer is just a memory machine. If thinking is just memory then AI is doable. If memory is just a aid to human thinking then AI is impossible. It will never be alive as it has no soul. Nobody is there after all. We are here as beings . Babys are here but have terrible memory's. Yet they think as a computer will never think. Robert Byers

Great insights, News and Mapou, As the article said, the success of AI has come with its failure--discovering that self-awareness is critical to intelligence, and that the methods of modernism--linear logic, deterministic algorithms, programming languages--can never capture this aspect of reality. This doesn't mean it can't be done, it merely means that modernism can't get there from here. I think that is Denyse's point about naturalism and what it has done to science. But once we put the man back in the machine, then the machine will be capable of AI. So Mapou is right, neither GOFAI nor Bayesian brains are capable of circular logic, and therefore they will be dead ends. But language *is* capable of circular logic--as many forms of humor attest. So the problem isn't one of equating intelligence with language, the problem is one of not understanding language. Robert Sheldon

Alan Turing may be idolized in many circles but he was instrumental in planting the seed of AI's failure by equating intelligence with language. This triggered half a century of what came to be known as good old fashioned AI (GOFAI), aka symbolic AI. It was, of course, the first of several AI red herrings, a complete failure, a perfect example of scientific consensus gone stupid. Philosopher Hubert Dreyfus try to warn them of their folly but they refused to listen. The AI community has not learned its lesson. It is now embarked on another great red herring chase. It's called the Bayesian brain. Like GOFAI, it's full of promises and the initial successes are encouraging. I predict that, it, too, will fail. And unfortunately for academia and AI experts (most of which are atheists), true AI will come from the one place that they least expect. They're in for a little surprise. Mapou

H'mm: We directly experience ourselves thinking, and tha tis very important, but if you look closely News' clip wants to make thinking go away:

While the successes behind subsymbolic artificial intelligence are impressive, there is a catch that is very nearly Faustian: The terms of success may prohibit any insight into how thinking “works,” but instead will confirm that there is no secret to be had—at least not in the way that we’ve historically conceived of it. It is increasingly clear that the Cartesian model is nothing more than a convenient abstraction, a shorthand for irreducibly complex operations that somehow (we don’t know how) give the appearance, both to ourselves and to others, of thinking. New models for artificial intelligence ask us to, in the words of philosopher Thomas Metzinger, rid ourselves of an “Ego Tunnel,” and understand that, while our sense of self dominates our thoughts, it does not dominate our brains.

Why? It does not fit the materialist paradigm, so they assign a vast area of mindedness to delusion -- let's call spades spades. And let "us" reduce it to GIGO limited computation instead which we can fit in with our machine model. But there are no firewalls in mindedness, so from there we are in self referential absurdity, yet again on materialistic premises or assumptions. KF kairosfocus

Alan Turing and Kurt Godel - Incompleteness Theorem and Human Intuition - video http://www.metacafe.com/watch/8516356/ bornagain77

Nice find News! bornagain77

I wonder, just in terms of physical complexity, whether the 20th/21st century analogy of intelligence to computer programs may turn out to be as flawed as the 19th/20th century analogy of biological evolution to mechanics has turned out to be. On the basis that all beings with intelligence known to biology have G, A, T and C available to express it in the organism, while digital processing has only a binary system of zeros and ones with which to make the attempt in its robots. jstanley01