Transcending Complacency on Superintelligent Machines(huffingtonpost.com) |
Transcending Complacency on Superintelligent Machines(huffingtonpost.com) |
It looks to me, a layman, that the only approach that holds any water is the first one. But then again, it mostly looks like people are implementing software based on a flawed understanding of cognitive functions and basically hoping that something magic happens. How can a scattershot approach like this ever produce anything even remotely resembling human intelligence?
You're also crucially missing the possibility that someone comes up with an intelligent algorithm that doesn't mirror the human brain in much detail, but still manages to outperform it. Think of flight: inventing flapping machines didn't turn out to be very useful, but we figured out a workaround that was far more efficient. The most interesting (terrifying?) AI research is along these lines.
Are really we so special? Apart from being first (that we know of)? The history of science is a history of anthropocentrism iconoclasm!
Of course, copying what we know works seems like a reasonable starting point.
Secondly, who said it has to be done within a few decades? If it is done within centuries or millennia, it is still done. "Infeasible" would mean it can't be done, ever.
Personally, I tend towards 50 years off... as it has been for a long time (and probably will continue to be) <- this is a joke. I'm saying it's a looong way off.
Brain as a computer, in my opinion should be the default state for this discussions. Why? Consider the old and tired brain-made-of-matter argument. There's no reason to think there's something magical or supernatural inside the brain, so treat it as an organized collection of atoms doing cool stuff. The default state cannot be magic, it has to be something that can be disproved or ruled out.
Some parts of it seem to work, as fas as we know, in a (suspiciously) algorithmic way, or in other words, a highly abstract step-by-step chain of actions can be identify for a given part of the brain.
Why not start with the crazy assumption that the whole brain acts as a computer (the theoretical concept), and then identify which parts of it fail the analogy? The key part here is the word 'fail': it should not mean 'too complex for any computer we have built' nor 'we don't know any algorithm that does that', it should mean that there are parts that inherently cannot be modelled, under any circumstances, like the definition of algorithm. If some part is discovered not to hold the analogy, you should just then question if the analogy is question is apt or not.
Neurons are bit more general than our typical logic gates, but at a high level they have a lot in common.
Also, airplanes do imitate birds in a few fundamental ways.
There is another approach, what I would call the Airplane approach, since it is to the brain what airplanes are to birds. That is, to base machine intelligence on a new kind of mathematical logic that hasn't been invented yet.
As others have suggested, this is demonstrably wrong, trivially so.
Chess computers are better than humans, but I wouldn't trust them to manage the electricity grid. What if there was an equivalent quality of computer specialized for every significant area of society - electricity grid, packet routing, high speed trading, etc etc.
Approximate number of human neurons: 1.0e11
Approximate number synapses in a human: 1.0e14
Transistors in XBox One: 1.0e09
Approximate neural switching speed: 1.0e03/sec
Speed * synapses (brain ops per second): 1.0e17/sec
Minerscube 15 (hashes per second): 1.5e12/sec
AMD Radeon HD 6990 32-bit instructions: 2.6e12/sec
Bits of information in human genome: 1.0e10
Now, to address the argument about evolution taking millions of years. First, we can evolve programs much faster than nature can evolve humans. There have been, perhaps, 100 million generations of humans. Even if it takes six seconds of computing time to run an evolutionary computation for a single generation it will take no more than 20 years to run over 100 million generations.
Brute force evolution isn't the only way to build strong AI. A program can exhibit behavior that we don't anticipate. I've written simple programs that beat me easily at games such as Othello or Freecell.
Finally, once machines get smart enough to design other machines there may be a rapid acceleration of progress in this area as we employ them in designing subsequent generations.
I feel that strong AI may pose a significant risk to humans; consequently, we should proceed with caution. Here is a thought experiment. If a chimpanzee could be taught to drive, would you trust it to pick your kids up from school? What sort of value judgements would it make in the case of an impending emergency? Would you let an elephant baby sit for you? Even if was much "smarter" than a normal elephant?
Strong AI will not be like us. It will learn and develop without a human body, and it will not interact with the world and society as we do and may end up being very foreign to us. Will it be sociopathic? Or will it be like whales, intelligent, but mysterious, perhaps spending all its time singing AI songs to other AIs.
I also think it's a jump to go from "if brain isn't a computer - then magical". There's a lot of room in between. And there are plenty of reasons to think that what goes on inside the brain cannot be mimicked by a computer or algorithms as we currently know them. We don't even know what consciousness is! We should at least admit as much...
>And there are plenty of reasons to think that what goes on inside the brain cannot be mimicked by a computer or algorithms as we currently know them. We don't even know what consciousness is! We should at least admit as much...
I agree, it's a huge jump! And that's precisely my point. The brain as a computer paradigm has nothing to do with the idea that an i5 core can't recognize cats, is the theoretical aspect of a computing machine that is used when trying to argue in favour of the BaaC paradigm.
Conciousness is precisely what doesn't fit in the BaaC paradigm. So the research should start from there. I'm curious if the definition conciousness will have to be changed in the near future. Exciting times!
And we are on HN, mostly smart people here who vastly over estimate 'normal people'. It's nice that we (me included) assume a human can be taught to be able to do anything other humans can (with some margins), but for now this is not true either. And if we want this empirical evidence thing going on; if a (kind of) Turing test would be done with a large part of the population who have not been told they are, for instance, we let a human with earplugs walking around a village in Arkansas and walk up to an average person and play the human interaction for Watson (or something like it), it would usually succeed. It would in my village for 100% sure; I could actually make a knowledge based script for talking to a lot of people and they would not see the difference. So I understand what you mean, but I don't think in a chinese room kind of way (and that experiment, as many have shown, doesn't matter) we are not that far off. When we reach your level of input/output you will 'see it' but still, because you don't have a definition, will deny it. I would wager that we are there are already in the 'fuzzy' sense of at least 40% (I think it's a lot more) of the population. My grandparents, bless them, definitely think they are talking to a human when they call they book a railway ticket (which has been a steadily improving AI for a little under 20 years now); for their 'fuzzy' it's been solved and strong AI exists.
But hubris - yes it is hubris. Because there is no scientific basis for the assertion that we will cross that chasm into 'true' AI, and thus it's based just as much on faith as any religious belief. And it's hubris because they claim a scientific basis where there is none.
When there is a scientific basis or proof that we've reached (or will reach) this 'singularity', you won't see me complaining. I'm not anti-science. I just don't think it's ever going to happen.
On a semi-related note, doesn't anyone find it kind of odd that Ray Kurzweil's calculations for when the singularity will occur happen to be just about the time his natural life will end (statistically speaking)? These projections are all driven by ego and faith, very little by science...
About religion and science; it is about definition where there difference is; IF you accept some definition X as being strong AI then when we reach that we have a scientific reality. The chasm and 'true' AI and what these are in scientific terms are vague, however in science we accept definitions of how nature works and if those definitions are things you hold true there is no reason why it won't be reached as there is not 'special' in the fabric of our brains which we couldn't copy given advanced enough ehm, take your pick; biology, nanotech, electronics, 3d printing etc. If you however cannot agree on definitions and have that (to me alien) quality of accepting mystery above all, then sure it's all believe or not. Not a good conversation maker as we are done after 2s, but he.
Seriously though, I don't think the Conversation needs to end there (conversation with a capital C - ours can end whenever we want). I do indeed believe in 'mystery above all'. I actually think that's a lovely way of putting it. Because mysteries are just unknowns, and without unknowns, what happens to scientific exploration? Do we just assume we know everything? And then the exploration stops. I'll be more explicit than that as well - I believe in God, and I am somewhat religious. I don't think that cancels me out of any interesting conversations.
I think you're making an assumption when you say that if you hold scientific definitions true then there is no reason why it won't be reached. Science says nothing about the future certainty. It is composed of models whose intent is to reflect reality, testable hypotheses to build and refine those models, and the results of the tests of those hypotheses to validate or disprove the hypotheses. We have no model (other than some vague calculations of processing power of the brain), no testable hypotheses and no results for these projections. It's not science.
But I would say I've proven you wrong that this isn't a good conversation!
Also like I mentioned before, I have no clue how AI would clash with the existence of a God or religion. And so I don't understand why religious people get so upset about it. There are many things we improved on where we don't try to take god's place (as I guess that's what it's all about) according to religious people; like when we made a wheel, did we better God's work or try to out-do His work by showing that wheels are more efficient for a lot of things than legs? I don't see the difference with copying or even improving on intelligence. So what that clash is I don't know but religious people seem to get downright aggressive when you talk about strong AI which gives me, and many others, even more incentive to just side them with the crazies.
I was born in the Dutch bible belt and I was raised with religion in school where I had to learn verses by heart and recite them every day; the people too stupid to learn them (small village with lot of inbred) were punished for not being able to learn them and I was punished for asking questions like didn't God create these stupid people too, why punish them for something they cannot do? My aunt used to give me physics books for my birthday written by religious professors; actual physics books with quantum mechanics and string theory. And although I did not believe in god at all from a very early age (mostly because none of the people who tried to push me into christianity wanted to answer any critical questions) and I don't and never will understand how someone can believe in most of the the things religion dictates, those books my aunt brought showed not everyone was a crackpot and actually there is no reason (and there isn't ffs) why physics, AI, evolution theory would not simply rhyme with religion. They are not mutually exclusive as so many (I cannot find another way of saying it) misguided individuals seem to think especially in the US. I assume you are not one of them as you don't mind critical discussion etc.