Artificial Intelligence at Human Level by 2029?

Gerard Boyers writes "Some members of the US National Academy of Engineering have predicted that Artificial Intelligence will reach the level of humans in around 20 years. Ray Kurzweil leads the charge: 'We will have both the hardware and the software to achieve human level artificial intelligence with the broad suppleness of human intelligence including our emotional intelligence by 2029. We're already a human machine civilization, we use our technology to expand our physical and mental horizons and this will be a further extension of that. We'll have intelligent nanobots go into our brains through the capillaries and interact directly with our biological neurons.' Mr Kurzweil is one of 18 influential thinkers, and a gentleman we've discussed previously. He was chosen to identify the great technological challenges facing humanity in the 21st century by the US National Academy of Engineering. The experts include Google founder Larry Page and genome pioneer Dr Craig Venter."

Comments

Oblig.

[Anonymous Coward]

I for one welcome our broadly supple, emotionally intelligent overlords.

Re:Oblig.

[fyngyrz]

Speaking as an engineer and a (~40-year) programmer:

Odds are extremely good for beyond human AI, given no restrictions on initial and early form factor. I say this because thus far, we've discovered nothing whatsoever that is non-reproducible about the brain's structure and function, all that has to happen here is for that trend to continue; and given that nowhere in nature, at any scale remotely similar to the range that includes particles, cells and animals, have we discovered anything that appears to follow an unknowable set of rules, the odds of finding anything like that in the brain, that is, something we can't simulate or emulate with 100% functional veracity, are just about zero.

Odds are downright terrible for "intelligent nanobots", we might have hardware that can do what a cell can do, that is, hunt for (possibly a series of) chemical cues and latch on to them, then deliver the payload -- perhaps repeatedly in the case of disease-fighting designs -- but putting intelligence into something on the nanoscale is a challenge of an entirely different sort that we have not even begun to move down the road on; if this is to be accomplished, the intelligence won't be "in" the nano bot, it'll be a telepresence for an external unit (and we're nowhere down *that* road, either -- nanoscale sensors and transceivers are the target, we're more at the level of Look, Martha, a GEAR! A Pseudo-Flagellum!)

The problem with hand-waving -- even when you're Ray Kurzweil, whom I respect enormously -- is that one wave out of many can include a technology that never develops, and your whole creation comes crashing down.

I love this discussion. :-)

Re:Oblig.

[Your.Master]

Let's imagine that computer "processing power" doubles every 2 years for the next 20 years, from a combination of hardware advances and software algorithm innovation. That's not quite Moore's law, and it's not really likely to work smoothly just like that, but just take it as one possibility. In that case, computers of 2029 will be 1024 times as powerful as today. So the question is, are human brains = 1000 times as powerful as a mouse's brain?

Maybe they aren't. But when you say a few centuries, I can't agree anymore. Let's imagine one century. Now we're hitting 1.12589991 × 10^15 times. A human brain is CERTAINLY within that complexity range. The caveat here is can we maintain the doubling rate for a full century? Well...Ray thinks we'll do far better than that (his "law of accelerating returns"), I'm not convinced we'll even be able to sustain the rate -- I think honestly we're looking at a plateau maybe 10, 20 years down the line, and will look back at computing as an S-curve until the next big breakthrough which nobody can predict. In my view the last couple "next big breakthrough"s happened at convenient times to make it look like we weren't following an S curve but we're just getting sharper and sharper, but I don't see any reason why the next one should happen just as conveniently. But since it's unpredictable, I could blindsided by it and it could happen next week.

Language isn't far off at all, we just about have it already. Emotions are nebulous and some people will move the goalposts forever, while some may prematurely be convinced by a video game character. I'm not necessarily convinced they are the hardest part of this. I don't know how to make them, but I don't know how to do the rest of this too. I just often see emotion being listed as the be all and end all most difficult task and I've never seen any reason to believe that to be so.

Re:Oblig.

[fyngyrz]

Wasn't there a simulation of a mouse's brain, or a few cells of it, for a few seconds with the help of a modern supercomputer, we can barley manage to do that.

Well, let's look at the rate of general progress in computing. In 1971, we were putting 2300 transistors on a chip. They ran at a few hundred KHz. In a fairly smooth progression, we've gotten to 3 GHz, where we're likely to stay, and today, we're at about two billion transistors on a chip, with no end in sight as to how far that can go. This is not Moore's law; Moore's law is about how many fit into a particular space; this is about how many can be integrated into a functional unit. That's 36 years. Thirty six years from now, that ability to "simulate a few cells" should grow just in the *normal* scheme of things into an ability to simulate a billion or so cells without any trouble. But there's more to this. Not everything in a cell needs to be simulated; for instance, metabolic processes such as waste generation and removal don't, nor do breakdown, aging, impacts by free radicals, all of that. Part of what needs to be done between here and the goal is streamline the simulation so that it is operating in the zone of mentation and not biological imperatives. I suspect, and yes indeed this is just my opinion, that the simulation will be much easier when we understand just what it is we need to simulate.

This all leaves out the issue of non-simulating intelligence, where the thinking is not patterned after human mechanisms; this could arise from evolutionary software or something along those lines. And of course, one of the reasons that all this is kind of a holy grail anyway, only the first intelligence is difficult; the second... Nth is just a matter of copying a machine state.

As for language, that's solved in the I/o sense -- synthesis and "listening" are both satisfactorily complete. Intelligent discussion can only be expected from an intelligent machine, so that's only as far away as machine intelligence is.

Even if an intelligent computer was somehow created it would be an enormous accomplishment to have it be as intelligent as a bug or a small animal.

Small animals, I'm of the opinion, are a lot more intelligent than most people give them credit for. They just have a different intelligence. I am sure that we will go through the small animal level on the way to our level, and beyond; the thing is, if you can do the one, you can do the other. There's no indication of a significant difference in the wetware, there's just more of it and it is arranged somewhat differently. No reason to expect anything different from hardware designed to do the same job.

Emotions and language seem very far off, I'd say such a thing is centuries away.

Why? Small animals do both. Those aren't even the hard things. The hard things are introspection and self-awareness. Those are the ones we have not even a theory for, today. In any case, your ideas are certainly in with a lot of good company; but not me. I think we're only one discovery - algorithmic in nature - from AI. Self-awareness may turn out to be a property that self-organizes and arises without any special prodding from us; that would be marvelous, not to mention fortuitous, but hardly impossible - again, that's how nature did it.

Here's why I think we're just an algorithm away. If you left a question that absolutely required intelligence on a counter, and went back to pick it up the next day, and the answer was there -- you would agree that an intelligence had answered the question. If a human could answer it in one second, or an AI could answer it in 23 hours, it's still just as intelligent an answer when you pick it up. The point is that speed really isn't the issue. The issue is the process, that is, the algorithm. So it turns out that in terms of speed, number of transistors, etc, that's really not the limiting factor for developing intelligen

Re:Oblig.

[Walt Dismal]

I'll go Ray one better. We will have this before 2029. My company is working to bring to market synthetic intelligences that among other things have feelings, emotion, and mood, and understand human emotion. One may say, why try to do this? Because in order to understand people, a synthetic intelligence must understand these things, which serve real functions in people, and real functions in AIs trying to operate in a people world.

Note that it is not necessary to build 'perfect robots'. People think, and yet they are not perfect. They make mistakes, yet navigate through life. So we do not have to make flawless logic brains. The way people work is that we try to find good if not optimal solutions to problems, but we do not always exhaustively search for the perfect solution. Thus many problems in life can be solved in different ways than you would expect. We do not have to build a machine that finds the optimal solution to a traveling salesman problem in order to make a system that can walk from the kitchen to the front door. It just has to be able to get there reasonably optimally. Also, we do not have to replicate the human brain in order to think much like a human, we merely have to come up with functional systems that can provide similar functions. For instance, the human brain has the amygdala, which can be likened to an interrupt controller for emotional responses. Well, that functionality can be done in a hardware-software system that reasons about priorities of tasks and goals depending on their current 'value' of urgency to the 'brain'.

Many current researchers in many cases are missing the mark. For example, as good as it is, the widely-used AI textbook by Stuart Russell and Peter Norvig (who heads research at Google) has major omissions. It does not dwell on key things needed to bridge between AI and human psychology. Other things like the OCC model of emotion used in AI is incomplete and incorrect in parts. A new approach has been needed, and one I've been developing for decades in stealth mode. I'm writing a 5-volume book set on it. I want it to be the Knuth set of AI.

I'm in the process of patenting the mechanizations of my underlying technologies, and trying to cut deals with companies making multicore processors so their architectures support the thread swapping needed to make virtual neural nets practical. Once we get a 1024 simplified-core processor that supports virtual NNs, it'll be a lot easier to build a machine with many of these that does for NNs what disk swapping does for OSs, than to build a billion-neural processor hardwired machine. And easier to do visual perception systems properly too. So Ray is right. If I can drive certain companies to build the right silicon, we can get there by or before 2029. My current software does what I said, but it's too slow on current hardware. Needs new processors and new system architectures, and it will take 20 years to get the infrastructure all built up. But not a lot more.

Re:Oblig.

[fyngyrz]

But sadly, we still know jack shit about how the brain works

Most of us know jack about the algorithms that allow us to catch a baseball in flight, yet we can still do it. Furthermore, a person from 10000 BC with no math at all by today's standards could do it just as well as we can. Implementing solutions does not always require a complete understanding of what you've done. You can even be wrong and it'll still work for other reasons. So hard-pegging this to what we "know" could be a severe error.

And no, simply copying the brain structure will not the answer.

That's a very bold statement, especially since (a) that's the way nature does it for all its intelligences, high and low, so we know the process works in the general case, and (b) as you say, we don't know many things yet, so claiming that we "know" what won't work seems to be disingenuous or at the very least not well thought out.

I think it is important not to conflate the fact that we don't understand something with the idea that it will be difficult once figured out or discovered as a consequence of some fortuitous sequence of events. That's been shown again and again not to be the case. It *may* be so, but it is by no means certain to be so, and for that matter, it isn't indicated by the complexity of the brain's hardware. The brain is considerably more formidable as a mass of immensely complex moderated connectivity than it is as a collection of cellular-level mystery machines, and a good deal of the complexity at the cellular level is almost certainly irrelevant to the task of thought -- keeping the cell alive is probably in no way related to non-pathological mental operation, yet there's a lot of hardware and systems involved in the task.

Re:Oblig.

[fyngyrz]

My current software does what I said, but it's too slow on current hardware.

That's a *huge* claim; if it is true, you have AI now. Because -- as I explained in a previous post in this thread -- speed is absolutely irrelevant. If you can demonstrate your claim that your software operates now, no matter *how* slowly it operates, you are at the end of your funding issues, not to mention any other issues you may face in life. Which -- to be frank -- is why I doubt your claim. At the point you explicitly claim to be at, I'd already own a mega-yacht and be pulling up next to a lot of potential love.

But good luck, and I really mean that. I'd much rather be wrong and see you bring this right to the table, even if you have completely blown the financial potentials of the development process.

Re:Oblig.

[Kagura]

Wasn't there a simulation of a mouse's brain, or a few cells of it, for a few seconds with the help of a modern supercomputer, we can barley manage to do that.

We already know that it's possible to contain 100% of real-time human brain functions in a casing 10cm by 10cm by 10cm and weighing under five pounds. Now we have to build one from the ground up with potentially slower, yet better understood technology. The problem, unfortunately, isn't related to hardware. I have no doubt that processor power will soon be sufficient for our needs, but without software that can think on the level of a human, it's just another personal platform to play Duke Nukem Forever on.

Re:Oblig.

[flyingsquid]

I'll go Ray one better. We will have this before 2029.

Yes, but what do they mean by "human level intelligence", in particular, which human are we talking about? I mean, if "human level intelligence" means "as smart as George W. Bush", then I wouldn't trust that machine to handle my taxes, let alone any really critical tasks.

Re:Oblig.

[ralphbecket]

Speaking as someone with a PhD in AI, I'm very, very skeptical about having human-level AI by 2029.

Whatever definition of intelligence you choose, it probably includes learning and reasoning components. We have some effective learning algorithms, provided your domain is very specific and you have boat loads of training data. We have next to no good reasoning algorithms. Complete search is a dead duck and incomplete search is not very reliable. Worse, search algorithms get seriously confused when the data base is inconsistent (humans are good at maintaining several incompatible world models simultaneously). And that's all before you consider that we have no psychological models of human reasoning that are anywhere near being specific enough to guide an implementation project (please don't mention "Society of Mind"). Finally, there is precious little funding out there for this kind of research, which is a shame, but there you go.

Re:Oblig.

[fyngyrz]

However, to program a computer to simulate thought accurately, an accurate algorithm for thought (or the biological underpinnings of neural activity) IS implicitly required, as algorithms are the way the computer works.

No, you have missed my point. An algorithm or algorithms is certainly required, and I never meant to imply otherwise. Human understanding of said algorithm(s), however, is explicitly not required. And there are many paths that lead to such a situation. Whether one of those will take us to a form of AI remains to be seen, which is what I was saying.

I think it is impossible for any one brain to fathom how a brain works completely

It is one thing to understand the mechanism required for operation -- it is quite another to understand the state it is in. I think you are confusing the latter with the former; the former is relatively trivial, and the latter is not required any more than a complete understanding of the state of everything involved at NASA is required in order to create, launch and recover the space shuttle. Complex systems are holistic, mostly co-operative combinations of subsystems, and as long as someone, somewhere, understands (or understood at one time, or possessed an adequate analogy to, or approximation of) the subsystems, or even the subsystems that make up the subsystems, that's sufficient to develop a fully functional macro system. And -- most importantly -- it only has to be done once, because of the unusual copyable nature of the result.

Re:Oblig.

[hedwards]

Odds are extremely good for beyond human AI, given no restrictions on initial and early form factor. I say this because thus far, we've discovered nothing whatsoever that is non-reproducible about the brain's structure and function, all that has to happen here is for that trend to continue; and given that nowhere in nature, at any scale remotely similar to the range that includes particles, cells and animals, have we discovered anything that appears to follow an unknowable set of rules, the odds of finding anything like that in the brain, that is, something we can't simulate or emulate with 100% functional veracity, are just about zero.

Sure if you build in enough memory and processing power the bottle neck ends up being the designers, but it'll be a really long time before the hardware gets to the point where that's possible.

At present hardware will crash if a few bits get in the wrong places or if they're stored incorrectly, one of the things about organic lifeforms is that our consciousness doesn't cease to exist if one of our neurons misfires, at worst we get a seizure or possibly a hallucination. Any machine that's going to surpass us would have to turn those wrong bits into something meaningful without human intervention. Even if they are just unexpected rather than outright wrong.

It may very well be that computer technology will solve that problem, but quite a bit of what we are comes from these random misfirings and unpredictable unreliable results. Modeling what humans are presently like, or even modeling what humans are like at the point when this becomes realistic is far easier than creating something that will outdo us by intellect.

I'm somewhat skeptical when you say that nothing a person's brain can do which cannot be modeled by software, when it comes to talking, moving building, following instructions and things of that nature, I see no reason why a machine couldn't be taught to do those things as well as we do. But when it comes to more subtle things, things which require creativity, sometimes things which require for a deliberate violation of typical common sense, I'm skeptical that a machine could be taught to do so.

I'm especially skeptical about that considering that we don't even know most of the things which the human brain does or how it does it. We know many things, and we know enough to greatly benefit ourselves, but there are still a fair number of things which we don't understand about the brain. It is not a simple organ to understand, just in the last 10 years the amount of information gained about it is sufficient for me to suggest that you shouldn't suggest that there isn't a part of the brain which cannot be simulated.

I really don't want to suggest that it is impossible for us to create something that surpasses our own selves, but doing so would require things which we haven't even dreamed up yet, teaching a computer AI to be capable of meaningful creativity isn't something which is yet even on the most distant horizon, none of the programming languages or tool kits that are available presently offer that sort of capability in anything which resembles a reasonable number of lines of code.

Re:Oblig.

[Burnhard]

"The brain is considerably more formidable as a mass of immensely complex moderated connectivity than it is as a collection of cellular-level mystery machines, and a good deal of the complexity at the cellular level is almost certainly irrelevant to the task of thought -- keeping the cell alive is probably in no way related to non-pathological mental operation, yet there's a lot of hardware and systems involved in the task."

You (and most proponents of AI) have failed to answer any of the philosophical/metaphysical questions one inevitably becomes confronted with, by using the analogy of the brain as "software" and stating that the hardware is irrelevant. I suspect there are cellular-level mysteries yet to be discovered, including possibly quantum action at a low level, that would have a strong influence on the facts of the matter here. It is a rather simple-minded and arrogant "faith" that leads you to believe we have anywhere near a good understanding of how the brain works.

Re:Oblig.

[Walt Dismal]

It's okay to doubt it; until I demo pieces to people, they don't believe it either. I do have strong AI right now running on a Von Neuman class processor, but AI really needs a massively parallel architecture. Speed IS relevant. You can't just run big wide NNs simulated in a Von Neuman, it takes forever. The best architecture I find is where processing and recognition are the same as memory. That is, once the AI learns something, the pattern recognizer also serves as the memory and the belief logic. In a way, the NNs ARE the data storage, there is no separate RAM. And my AI is implemented in a connectionist architecture that embodies symbolic processing in a new way. The knowledge storage is synonymous with the logic. But this requires massive though simple logic units and parallelism. Further, it is not completely self-organizing from the ground up, I view that as a hopeless approach.

I define consciousness and awareness within a pre-determined architecture, not entirely self-organizing from scratch. The visual front-end in particular is very rigid, but I think it is okay because there is little need for an organism to self-evolve completely new architectures, but rather to be able to run cascaded pattern-recognizers. See the work of Biederman for examples of this. The VFE feeds deeper processing doing cognition, and there is feedback from that to the VFE for training. Just as a baby learns to see, and recognize shapes, and build up from that. The front end is trained as the cognitive end learns and grows too.

AI does not spontaneously rise alone from massive databases, either. I view that approach as useless and a false trail. However, human intelligence does depend on belief systems and knowledge, and those continually grow as we mature from infancy. But to create the equivalent of an 18 year old, you have to have what amounts to 18 years of accumulation of knowledge about the world, and draw upon that. And there is a key but proprietary subtlety about that I'm devoting an entire volume to, that is the key to humanlike AI. That volume is essentially a doctoral thesis about consciousness reworked for use by a design staff. As for funding, no yachts yet. But I'm real, sane, and not a charlatan, and have explained my technology to my patent attorney. I expect to be hiring staff within two years. I posted on Slashdot not for glory but to counter all the nay-sayers who haven't a clue what is achievable.

Re:Oblig.

[jdoeii]

Actually, Eugene Izhikevich run a successful simulation of the entire human brain in 2005:
http://vesicle.nsi.edu/users/izhikevich/human_brain_simulation/Blue_Brain.htm#Simulation%20of%20Large-Scale%20Brain%20Models
Simulation of 1 second took 50 days on a cluster of 27 PCs (~4.3M times slower than realtime). Eugene is a pretty smart guy, except not as prominent as Kurzweil. Here is Eugene's estimate for AI timeline:
http://vesicle.nsi.edu/users/izhikevich/human_brain_simulation/why.htm
You may also want to google for Henry Markram and his current project.

Re:Oblig.

[smallfries]

Although I have mod points at the moment it seems that your comments have already been modded up anyway so I guess I'll reply. We've actually argued about this before (probably last time Kurzweil was mentioned on slashdot), but as you say in a different comment this discussion is always fun :)

Slight nitpick, currently we are at one billion transistors on a chip, not two, but that doesn't really change the point you are making.

A bigger issue that I have with what you've described is that simulating a brain is not the same as "solving" AI. The problem that Kurzweil has is that he refuses to accept that there is a difference. Sure, if they are the same then strong AI is inevitable and it's merely a question of building fast enough hardware. But why assume that they are the same thing?

Twenty years from now we may have hardware that can simulate an entire human brain; and yet we may be no closer to solving any of the problems in understanding how to solve the many problems in AI. The mental sleight-of-hand that Kurzweil applies to this argument is: Once we can simulate a brain we have AI, therefore the AI can design he next generation, therefore we will reach the singularity. This argument is a logical fallacy because it assumes being able to run the system, and knowing how to design the system / how the system works are equivalent.

Everything that we know complex and dynamic systems tells us that this is not so; given a simulation of the brain it is reasonable to assume that intelligence is the ultimate emergent property of the system. Understanding this property and how to refine it is the hardest problem that mankind has ever undertaken. Currently we don't really know how to pose the question, let alone how to arrive at an answer. To assume that some kind of standard engineering methodology will solve this in 20 years is wild speculation.

As always with AI, the hardware will be available but nobody yets knows if we can write the software to run on it.

Re:Oblig.

[cp.tar]

but humans are the only creature that has ever been scientifically shown to have anything like language.

That is incorrect. Language is the ability to communicate feelings, goals, results. It is not "speech." Some birds do indeed have the capability of speech, that is, they can make the same sounds we can, closely enough as to make no difference. Apes, however, have demonstrated actual communications using symbols, and even dogs have recently been found to have a consistent, though very small, vocabulary. Elephants and other animals have demonstrated the ability to think in the abstract (the "recognize one's self in the mirror and operate on the information thus provided experiments.) Lemurs use calls to communicate safety and status. Don't confuse the lack of vocal apparatus with an inability to communicate. They're not the same thing at all.

As for the rest, I think you've got it, essentially, but we disagree on scales. We'll see.

Er... no.

Language is much more than that: it is a system of symbols that can even be used to describe any other symbolic system, and which can be extended at need and at will; animal communication shows little or no indication of that.

Nobody in their reight mind would deny that animals can communicate, and even that they can communicate very well.
However, that alone does not make them capable of using a language.

The cognitive leap a simple verbing of a noun requires is beyond any other animal.

No chance

[Kjella]

I mean it could happpen but this is so far from the current state of the art, I think we're talking 50-100 years forward in time. We have the brute powers of computers but nowhere near the sophistication in software or neural interfaces to do anything like this.

Hrmmmm

[BWJones]

I'll be meeting with Kurzweil in April.... Speaking as a neuroscientist who is doing complex neural reconstructions, I think he's off his timeline by at least two decades. Note that we (scientists) have yet to really reconstruct an actual neural system outside of an invertebrate and are finding that the model diagrams grossly under-predict the actual complexity present.

Re:Hrmmmm

[DynaSoar]

And as a cognitive neuroscientist, I say he's off the mark entirely. As per Minsky, a fish swims under water; would you say a submarine swims?

What exactly is the "level of humans"? Passing the Turing test? (Fatally flawed because it's not double blind, btw.) Part of human intelligence includes affective input; are we to expect intelligence to be like human intelligence because it includes artificial emotions, or are we supposed to accept a new definition of intelligence without affective input? Surely they're not going to wave the "consciousness" flag. Well, Kurzweil might. Venter might follow that flag because he doesn't know better and he's as big a media hog as Kurzweil.

I think it's a silly pursuit. Why hobble a perfectly good computer by making it pretend to be something that runs on an entirely different basis? We should concentrate on making computers be the best computers and leave being human to the billions of us who do it without massive hardware.

Re:Hrmmmm

[LurkerXXX]

What aircraft corner as fast as barn swallows?

There are still many things we can learn from biology that can be translated to machines. The translations don't have to be 1:1 for us to make use of them. The way birds as well as insects make use of different shapes in surfaces during wing beats have translated into changes in some aricraft designs. They weren't directly incorporated the same way, but they taught us important lessons that we could then implement in different ways but with a similar outcome.

I think Neuroscience does have a lot to teach us about how to do AI.

Re:Hrmmmm

[Flicker]

How many barn swallows can fly at 40,000 ft? Just what are you comparing?

Re:Hrmmmm

[Daffy Duck]

Maybe that's what Kurzweil is getting at: by the year 2029, AI will have achieved human-level abilities to make grossly inaccurate predictions of the future of AI.

Re:Hrmmmm

[teh moges]

What does a human do to read a bluff? He observes his opponent, takes inputs such as bet size and heart rate, applies them to known patterns of bluffers and looks for a match. Sure a human does this without realizing, but little of how this happens is a mystery. Also, how do humans bluff? They just bet at a negative EV play*, and bluffing properly is a matter of knowing the probability that the opponent will call. I am researching applying AI to poker (look out in June for a lot of high quality research from the AAAI Computer Poker Championship) and this sort of argument, "Computers can't bluff, they just run numbers", is both understating what has been achieved in AI in this field and also overstates what humans do. Yes, computer programs aren't quite up to the standard of world class players (Limit poker has achieved this, but not No-Limit), but this game has only a couple of years to go before this milestone is reached. I predict that by the end of the year, we will have high quality bots that can beat 99% of players, and by the end of 2010 No Limit Texas will be a computer dominated game.

The only thing that humans do that AI doesn't (well) is automatically follow a few paths, rather then look at the whole picture. As an example, it has been shown (sorry no reference right now) that some chess grandmasters look only at a couple of moves and then calculate all the possible combinations from there rather then examine every possible move. This drastically speeds up the calculation, however it does miss moves that could be considered the "best". So while this act of "feeling" which is the best move is a good approximation done by humans, it isn't an optimal or maximal play.

As for the article, I don't agree with all of what he says (the idea of nanobots doing what Kurzweil says scares me and I doubt it will be legal to do this), but I do agree with the 2029 prediction, that is if proper resources are given to that particular problem. Replicating humans is a goal in AI for some researchers, but not all of them. Personally, I couldn't care less if there exists a robot that perfectly resembles a human, as long as there are intelligent computers systems that can do the problems that humans find hard (such as finding patterns in very large sets of data or solving complex mathematical equations).

*Technically, it isn't a low EV play if there is a high probability of the opponent folding. In which case, playing the highest EV play naturally involves bluffing if it can be assumed that the opponent will fold to a bet.

Exponential AI?

[TheGoodSteven]

If artificial intelligence ever gets to the point where it is greater than humans, won't it be capable of producing even better AI, which would in turn create even better AI, and so on? If AI does reach the level of human intelligence, and eventually surpasses it, can we expect an explosion in technology and other sciences as a result?

Re:Exponential AI?

[psykocrime]

If artificial intelligence ever gets to the point where it is greater than humans, won't it be capable of producing even better AI, which would in turn create even better AI, and so on? If AI does reach the level of human intelligence, and eventually surpasses it, can we expect an explosion in technology and other sciences as a result?

That's the popular hypothesis.

Re:Exponential AI?

[wkitchen]

This positive feedback effect happens to a considerable extent even without machines that have superintelligence, or even what we'd usually consider intelligence at all. It's happening right now. And has been happening for as long as humans have been making tools. Every generation of technology allows us to build better tools, which in turn helps us develop more sophisticated technology. A great example from fairly recent history, and that is still ongoing, is the development of CAD/CAM/CAE tools, particularly those used for design of electronic hardware (schematic capture, PCB, HDL's, programmable logic compilers, etc.), and the parallel development of software development tools. Once computers became good enough to make usable development tools, those tools helped greatly with the creation of more sophisticated computer technology, which supported better development tools.

Superintelligence may speed this up, but the effect is quite dramatic already.

20 years is too long to predict

[Bill, Shooter of Bul]

The farther out you make a projection, the less likely it is to be true. With this one in particular, I just don't see it being a focus of research. Yes we will have increase levels of intelligence in cars toasters and ball point pens, but the intelligence will be in a supporting role to make the devices more useful to us. There isn't a need for a human like intelligence inside a computer. We have enough ones inside human bodies.

Also, I will not be ingesting nano bots to interact with my neurons, I'll be injecting them into my enemies to disrupt their thinking. Or possibly just threatening to do so to extract large sums of money from various governmental organisations.

Re:20 years is too long to predict

[Jugalator]

There isn't a need for a human like intelligence inside a computer. And even if there was (and I think this is key to the fallacy in this prediction), we wouldn't have the theories backing the hardware. We will most likely get some super fast hardware within these years, but what's much less certain is if AI theories will have advanced enough by then, and if the architecture will be naturally parallelized enough to take advantage of them. Because while we don't know much about how the human brain reasons, we do know that to make it at an as low temperature as 37 degrees Celsius in an as small area as our cranium (it's pretty damn amazing when you consider this!), it needs to be massively parallelized. And, again, we don't really even have the theories yet. We don't know how the software should best be written.

That's why we even in this day and age of 2008l, we're essentially running chatbots based on Eliza since 1966. Sure, there's been refinements and the new ones are slightly better, but not by much in a grand scheme. A sign of this problem is that they are giving their answers to your questions in a fraction of a second. That's not because they're amazingly well programmed; it's because the algorithms are still way too simple and based on theories from the sixties.

If the AI researches claiming "Oh, but we aren't there yet because we haven't got hardware nearly good enough yet", why aren't we even there halfway, with at least far more clever software than chatbots working on a reply to a single question for an hour? Sure, that would be impractical, but we don't even have the software for this that uses hard with even the boundaries of our current CPU's.

So at this point, if we'd make a leap to 2029 right now, all we'd get would be super fast Eliza's (I'm restricting my AI talk of "general AI" now, not in heuristic antispam algorithms, where the algorithms are very well understood and doesn't form a hurdle). The million dollar question here is: will we before 2029 have made breakthroughs in understanding the human brain well enough in how it reasons along with constructing the machines (biological or not as necessary) to approximate the structure and form the foundation on which the software can be built?

I mean, we can talk traditional transistor-based hardware all day and how fast it will be, but it will be near meaningless if we don't have the theories in place.

Some major assumptions

[Yartrebo]

How are we so sure that advances in computers will continue at such a rapid pace. Computer miniaturization is hitting against fundamental quantum-mechanical limits and it's crazy to expect 2008-2028 to have progress quit as rapid as 1988-2008.

Short of major breakthroughs on the software end, I don't expect AI to be able to pass a generalized Turing Test anytime soon, and I'm pretty certain the hardware end isn't going to advance enough to brute-force our way through.

To heck with Artificial Intelligence!

[RyanFenton]

Artificial intelligence would be a nice tool to use to reach towards, or to use to understand ourselves... but rare is there a circumstance that demands, or is worth the risks involved with making a truly intelligent agent.

The real implication to me, is that it will be possible to have machines capable of running the same 'software' that runs in our own minds. To be able to 'back up' people's states and memories, and all the implications behind that.

Artificial intelligence is a nice goal to reach for - but it is nothing compared the the siren's call of memories being able to survive the traditional end of existence, cellular death.

Ryan Fenton

This is good news and bad news...

[zappepcs]

Good news: This could herald a lot of good stuff, increased unemployment, greater reliance on computers, newer divides in the class strata of society, further confusion on what authority is and who controls it, as well as greater largess in the well meaning 'we are here to help' phrase department.

Bad news: After reviewing the latest in the US political scene, getting machines smarter than humans isn't going to take so much as we thought. My toaster almost qualifies now. 'You have to be smarter than the door' insults are no longer funny. Geeks will no longer be lonely. Women will have an entire new group of things to compete with. If you think math is hard now, wait till your microwave tells you that you paid too much for groceries or that you really aren't saving money in a 2 for 1 sale of things you don't need. Married men will now be third smartest things in their own homes, but will never need a doctor (bad news for doctors) since when a man opens his mouth at home to say anything there will now be a wife AND a toaster to tell him what is wrong with him.

oh god, this list goes on and on.

AI may not get that far

[httpcolonslashslash]

As soon as they make robots that can have sex like humans...what's the point in inventing anything else? All scientists will be busy "researching" their robots.

wrong

[j0nb0y]

He obviously hasn't been paying attention to AI developments. The story of AI is largely a story of failure. There have been many dead ends and unfulfilled predictions. This will be another inaccurate prediction.

Computers can't even defeat humans at go, and go is a closed system. We are not twenty years away from a human level of machine intelligence. We may not even be *200 years* away from a human level of machine intelligence. The technology just isn't here yet. It's not even on the horizon. It's nonexistent.

We may break through the barrier someday, and I certainly believe the research is worthwhile, for what we have learned. Right now, however, computers are good in some areas and humans are good in others. We should spend more research dollars trying to find ways for humans and computers to efficiently work together.

Re:wrong

[Alaria Phrozen]

Excuse me.. who are you? You're saying RAY KURZWEIL hasn't been paying attention to AI developments? And you're modded insightful?

http://en.wikipedia.org/wiki/Ray_kurzweil

"Everybody promises that AI will hit super-human intelligence at 20XX and it hasn't happened yet! It never will!" ... well guess what? It'll be the last invention anybody ever has to make. Great organizations like the Singularity Institute http://en.wikipedia.org/wiki/Singularity_Institute really shouldn't be scraping along on such poor budgets, seriously if this ever worked, even a 0.001% chance of a friendly technological singularity occuring, isn't it worth investigating?

Re:Well I'm not holding my breath

[2.7182]

Yes, I remember well my youth, reading Goedel Escher Bach and Winograd, etc., thinking that the next scientific revolution was coming. Things never got any better than Eliza. Now as a hard scientist, I strongly feel that the problem is far far off.

Don't do it!

[magarity]

(most) People can go out to get more education to advance from a menial job to a more skilled one when taken over by a robot but wtf do we do if the machines are as smart as we are? Who is going to hire any people to do even the most advanced thinking jobs when the machine that works for electricity 24/7 can do it? This kind of thing will bring on the luddite revolution in a hurry.

Whatever Could They Mean?

[flyneye]

" Artificial Intelligence will reach the level of humans"
Buddy,I've been around more than four decades.I've yet to see more than a superficial level of intelligence in humans.
Send your coders back to the drawing board with a loftier goal.

The End of Intelligent Design

[denoir]

It is not too much of an overstatement to say that the field of AI has not significantly progressed since the 1980's. The advancements have been largely superficial with better and more efficient algorithms being created but without any major insights and much less a road map for the future. While methods that originated as AI research are more common in real-world applications, the research and development of new concepts has made a grinding halt - not that it was ever a question of smooth continuous progress.

It might seem like the lack of AI development is a temporary problem and altogether a peripheral issue. It is however neither - it is a fundamental problem and it affects all software development.

Early in the history of computing, software and hardware development progressed at a similar pace. Today there is a giant and growing gap between the rate of hardware improvements and software improvements. As most people involved in the study of the field of software engineering are aware of, software development is in a deep crisis.

The problem can be summarized in one word: complexity. The approach to building software has largely been based on traditional engineering principles and approaches. Traditional engineering projects never reached the level of complexity that software projects have. As it turns out humans are not very good at handling and predicting complex system.

A good example of the problems facing software developers is Microsoft's new operating system Windows Vista. It took half a decade to build and cost nearly 10 billion dollars. At two orders of magnitude higher costs than the previous incarnation it featured relatively minor improvements - almost every single new radical feature (such as a new file system) that was originally planned was abandoned. The reason for this is that the complexity of the code base had become unmanageable. Adequate testing and quality assurance proved to be impossible and the development cycle became painfully slow. Not even Microsoft with its virtually unlimited resources could handle it.

At this point, it is important to note that this remains an unsolved problem. It would have not been solved by a better structured development process or directly by better computer hardware. The number of free variables in such a system are simply too great to be handled manually. A structured process and standardized information transfer protocols won't do much good either. Complexity is not just a quantitative problem but at a certain level you'll get emergent phenomena in the system.

Sadly artificial intelligence research which is supposed to be the vanguard of software development is facing the same problems. Although complexity is not (yet) the primary problem there manual design has proved very inefficient. While there are clever ideas that move the field forward on occasion there is nothing to match the relentless progress of computer hardware. There exists no systematic recipe for progress.

Software engineering is intelligent design and AI is no exception. The fundamental idea persists that it takes a clever mind to produce a good design. The view, that it takes a very intelligent thing to design a less intelligent thing is deeply entrenched on every level. This clearly pre-Darwinian view of design isn't based on some form of dogma, but a pragmatism and common sense that aren't challenged where they should be. While intelligent design was a good approach while software was trivial enough to be manageable, it should have become blindingly obvious that it was an untenable approach in the long run. There are approaches that take the meta level - neural networks, genetic algorithms etc, but it is thoroughly insufficient. All these algorithms are still results of intelligent design.

So what Darwinian lessons should we have learned?

We have learned that a simple, dumb optimization algorithm can produce very clever designs. The important insight is that intelligence can be traded for time. In a short in

Re:The End of Intelligent Design

[Animats]

Can anyone name an important algorithm or representation from this decade?

There's been substantial progress in trainable computer vision systems in the last decade. Computer vision is finally starting to work on real-world scenes. SLAM algorithms work now. Texture matchers work. There really has been progress in those areas.

Predictions are useless in this case

[The One and Only]

It's one thing to predict when a building project will be finished or when we'll reach a certain level of raw processing power because these things proceed by predictable means. But strong AI requires us to make theoretical advances. Theoretical advances don't proceed like a building project--someone has to have a clever idea, fully develop and understand it himself and convince others of it. And it won't occur to someone all at once, so we'll need incremental advances, all of which will happen unpredictably.

I agree...

[Jane Q. Public]

As an party "outside" the field but interested, I agree with all of you here so far, except that of course you disagree on timelines. :o)

"Artificial Intelligence" in the last few decades has been a model of failure. The greatest hope during that time, neural nets, have gone virtually nowhere. Yes, they are good at learning, but they have only been good at learning exactly what they are taught, and not at all at putting it all together. Until something like that can be achieved (a "meta-awareness" of the data), they will remain little more than automated libraries. And of course at this time we have no idea how to achieve that.

"Genetic algorithms" have enormous potential for solving problems. Just for example, recently a genetic algorithm improved on something that humans had not improved in over 40 years... the Quicksort algorithm. We now have an improved Quicksort that is only marginally larger in code size, but runs consistently faster on datasets that are appropriate for Quicksort in the first place.

But genetic algorithms are not intelligent, either. In fact, they are something of the opposite: they must be carefully designed for very specific purposes, require constant supervision, and achieve their results through the application of "brute force" (i.e., pure trial and error).

I will start believing that something like this will happen in the near future, only when I see something that actually impresses me in terms of some kind of autonomous intelligence... even a little bit. So far, no go. Even those devices that were touted as being "as intelligent as a cockroach" are not. If one actually were, I might be marginally impressed.

Re:Probably false alarm ... again

[timeOday]

So ... I'd say: less claims, fewer predictions, and more work. Let me know when you've got anything worthwhile to show.

Has it occurred to you that all of us already work, to some extent, at the direction of computers? Think of the tens of thousands of pilots and flight attendants... what city they sleep in, and who they work with, is dictated by a computer which makes computations which cannot fit inside the human mind. An airline could not long survive without automated scheduling.

Next consider the stock market. Many trades are now automated, meaning, computers are deciding which companies have how much money. That ultimately influences where you live and work, and the management culture of the company you work for.

We are already living well above the standard that could be maintained without computers to make decisions for us. Of course as humans we will always take the credit and say the machines are "just" doing what we told them, but the fact is we could not could not carry out these computations manually in time for them to be useful.

The sacred brain and other myths

[denoir]

This is a sort of continuation of the parent post.

The comedian Emo Philips once remarked that "I used to think my brain was the most important organ in my body until I realized what was telling me this."

We have tendency to use human intelligence as a benchmark and as the ultimate example of intelligence. There is a mystery surrounding consciousness and many people, including prominent philosophers such as Roger Penrose, ardently try to keep it that way.

Given however what we through biological research actually know about the brain and the evolution of it there is essentially no justification for attributing mystical properties to our data processing wetware. Steadily with increased capabilities of brain scanning we have been developing functional models for describing many parts of the brain. For other parts that need still more investigation we do have a picture, even if rough.

The sacred consciousness has not been untouched by this research. Although far from a final understanding we have a fairly good idea, backed by solid empirical evidence that consciousness is a post-processing effect rather than being the first cause of decision. The quantity of desperation can be seen in attempts to explain away the delay between conscious response and the activations of other parts of the brain. Penrose for instance suggests that yes, there is an average 500 ms delay, but that is compensated by quantum effects that are time symmetric - that the brain actually sees into the future, which then is delayed to create a real-time decision process. While this is rejected as absurd by a majority of neuroscientists and physicists, it is a good example of how passionately some people feel about the role of the brain. It is however painstakingly clear that just like we were forced to abandon an Earth-centered universe we do need to abandon the myth of the special place of human consciousness. The important point here is that once we rid ourselves of the self-imposed veil of mystery of human intelligence we can have a sober view on what artificial intelligence could be. The brain has developed through an evolutionary optimization process and while getting a lot of benefits it has taken the full blow of the limitations and problems with this process and also its context.

Evolution through natural selection is far from the best optimizing method imaginable. One major problem with it is that it is a so called "greedy" algorithm - it does not have any look ahead or planning capabilities. Every improvement, every payoff needs to be immediate. This creates systems that carry a lot of historical baggage - an improvement isn't made as a stand-alone feature but as a continuation of the previous state. It is not a coincidence that a brain cell is a cell like any other - nucleus and all. Nor is it a cell because it is the optimal structure for information processing. It was what could be done by modifying the existing wetware. It is not hard to imagine how that structure could be improved upon if not limited by the biological building blocks that were available to the genetic machinery.

Another point worth making is that our brains are optimized not for the modern type of information processing that humans engage in - such as writing software for instance. Humans have changed little in the last 50,000 years in terms of intellectual capacity but our societies have changed greatly. Our technological progress is a side effect of the capabilities we evolved that increased survivability when we roamed the plains of Africa in small family hunter-gatherer groups. To assume the resulting information processing system (the brain) would the ultimately optimal solution for anything else is not justifiable.

There has been since the 1950's ongoing research to create biologically inspired computer algorithms and methods. Some of the research has been very successful with simplified models that actually did do something useful (artificial neural networks for instance). Progress has however been agonizi

Re:Don't think so

[bnenning]

Predictions like this have been made in past, and not even come close. This one is no different.

The difference is that in 20 years we may have sufficiently powerful hardware that the software can be "dumb", that is, just simulating the entire physical brain.

The bottom line is that humans process some information in a non-representational way, while computers must operate representationally.

What prevents a computer from emulating this "non-representational" processing? Or is the human brain not subject to the laws of physics?

Kurzweil's rebuttal from his book...

[doug141]

The Singularity is Near has a rebuttal of your first paragraph. Any sucessful part of AI research spins off into its own well-functioning discipline... optical character recognition, dictation software, text-to-speech, etc... they were sci-fi "AI" in 1980 and now they are working technologies. AI research is the umbrella under which only the unsolved problems still lie, and thus is always undone.

Projection length

[Myria]

The farther out you make a projection, the less likely it is to be true.

I predict that the Sun will become a white dwarf within 10,000,000,000 years. Predicting 10 billion years instead of 5 billion years actually makes it more likely to be true.

There is no such thing as Artificial Intelligence

[glwtta]

At least not yet. I can't believe that the sort of bullshit that Ray Kurzweil keeps peddling gets taken so seriously.

There is a lot of talk about computers surpassing, or not surpassing, humans at various tasks - does it not bother anyone that computers don't actually posses any intelligence? By any definition of intelligence you'd like? Every problem that a computer can "solve" is in reality solved by a human using that computer as a tool. I feel like I'm losing my mind reading these discussions. Did I miss something? Has someone actually produced a sentient machine? You'd think I would have seen that in the papers!

What's the point of projecting that A will surpass B in X if the current level of X possessed by A is zero? There seems to be an underlying assumption that merely increasing the complexity of a computational device will somehow automatically produce intelligence. "If only we could wire together a billion Deep Blues," the argument seems to go "it would surpass human intelligence." By that logic, if computers are more complex than cars, does wiring together a billion cars produce a computer?

Repeat after me - The current state of the art in artificial intelligence research is: fuck all. We have not produced any artificial intelligence. We have not begun to approach the problems which would allow us to start on the road to producing artificial intelligence.

Before you can create something that surpasses human levels of intelligence, one would think you'd need to be able to precisely define and quantify human intelligence. Unless I missed something else fairly major, that has not been done by anyone yet.

Eliza and the sad state of expert systems

[StreetStealth]

Every time I try out a new expert system, it gets more depressing -- it honestly feels like no progress is happening in that market at all. I have yet to have a conversation with a computer that has been any more compelling than my first round with WinEliza on Windows 3.1 in 1995.

There's still no semblance of a short-term memory, even so much as continuity between responses. It always quickly becomes obvious that each response has been prepared verbatim beforehand by a human, that the system is still performing only a keyword-canned response routine, perhaps feeding in a few variable strings.

Today we have the same stone wheels we've had for decades, and the article suggests we'll have an internal combustion engine with antilock brakes and a hood ornament in another 20 years. We'll see.

How about world abandonment?

[slew]

Why on earth would this advanced AI want to stay on little old earth?

Seems to me that any crazy smart AI would just beam themselves out into space to avoid us and maybe watch us from a distance occasionally for amusement.

Think of this way, when you see an anthill, it's rather curious for a while, then you get bored and go on your merry way. Unless of course you are a sociopath and want to destroy the ant hill and all the ants for fighting with other ants, or you are insane and you want to teach the ants to get along with other ants or spiders their mortal enemy or perhaps you are psychotic and want to train the ants to do your bidding. More likely you would just leave and go on to something more interesting (unless you are not that intelligent to begin with).

I fail to understand why people seem to insist that any really smart AI would want to have anything to do with us except on an occasional basis. Humans and earth aren't really that important in the bigger scheme of things (just important to us humans of course) and we'd probably not have much in common with any really advanced AI anyhow.

If humans would ever create such an AI, it would be like a bunch of ordinary joes giving birth to a super einstien. Eventually, the 'kid' would stop listening to us, go do their own thing which we would be too dumb to understand or appreciate and occasionally we'd invite it to visit to help us fix the settings on our computer because we got it messed up. It would explain to us in excruciating detail how we were using the wrong type of computer and how we needed to get up to date on technology and we'd just tell them a story about how it was in the old days, it would roll it's virtual eyes and say thanks for the tip, and go back to it's own business of which we would be blissfully ignorant...

Just think about it for a second.

He must know something I don't

[melted]

And I work on AI and machine learning day in and day out. I'd put the goal post at 50 years, and that's an optimistic estimate. There are scant few research centers that do "general AI" research. Even fewer actually talk to neuroscientists, thus dismissing one viable (though extremely complex and costly) avenue of research. The fact remains, however, that at this point we don't have the required sophistication in any of the areas that presumably would be required to build a "thinking" machine. We can't process human language well enough (and therefore speech recognition and textual information sources are pretty much useless), we can't process visual information well enough either (segmentation, recognition, prediction, handling a continuous visual stream), we don't know the cognitive mechanisms below high level abstract reasoning, and even at a high level our abilities are weak (try to build a classifier that will recognize sarcasm, for example), finally even if we could do all that, we wouldn't be able to store the resulting data efficiently enough (in terms of required space and retrieval speed), because we have no idea how to do it.

That said, a lot of stuff can happen in 50 years, and I bet that once some of the major problems get solved, there will be an insane stream of money pouring into this field to accelerate the research. Just imagine the benefits an "omniscient" AI trader would bring to a bank. The question is, do we want this to happen? This will be far more disruptive a technology than anything you've ever seen.

What aircraft corner as fast as barn swallows?

[YetAnotherBob]

Any aircraft the size of a barn swallow.

Your question displays a lack of understanding. Not of biology, but of physics. Square cube law specifically. Aircraft don't corner as fast as small birds. the reason isn't any magic of biology, it's simple momentum.

The larger any object is, the more it weighs. Make it twice as big, it weighs eight times as much. packs eight times as much momentum. A large bird doesn't turn s fast as a small bird. Same is true of planes. Same is true of ships. A buss won't corner as fast as sports cars either.

A typical aircraft is 1000 times bigger than a swallow. It's a million times heavier. It packs a million times the momentum. It's not that the swallows design is better, or that there is some biological magic. It's just a question of size. It's true the other way too. A mosquito can turn a lot quicker than a barn swallow. Barn swallows catch mosquitoes because they can fly faster. Guess what, the aircraft you were so dismissive of can fly a lot faster than that barn swallow too. Visit a large airport. Swallows get killed by aircraft every day. They can't get out of the way in time. A barn swallow that was as large as a chicken would be ripped apart by the stresses if it were able to corner as fast as a real barn swallow. That's the real reason that chickens don't turn well in flight. (Yes, chickens can fly for short distances.) Momentum.

Your problem appears to be that you just don't understand scale. It is a wonderful thing when you do. You see reasons all around us, for all kinds of things.

So, yes, we should study biology. But, we should also remember the physics. The tricks the mosquito uses just won't work for a passenger jet. Nor will the barn swallows turns be good for the passengers on that jumbo jet. Still, some things will be useful. We just don't know what. Who would have thought that studying a sharks skin would help racing yachts. Personally, I hope that we get a lot of surprises. That's where the fun in science is.

I don't expect AI research to give us human type intelligence in a machine. Ever. That doesn't mean we shouldn't try. We don't know what we will get, or what it will make possible. We can't know before the fact. Studying birds didn't give us aircraft that can corner in a second or two, it did give us jumbo jets that can take us half way around the world in an easy chair. That took a lot of other things too.

The Wright brothers succeeded where Lilenthal failed. Not because they understood birds better, but because in the meantime the internal combustion engine was developed. AI will be the same. Right now, we don't even know what we need in order to make this work. There will be surprises.