Venture capitalist: We need to prepare for artificial intelligence
mercurynews.com
mercurynews.com
http://www.longnow.org/projects/seminars/ (search for him near the bottom of the page)
Talk of the singularity reminds me of the furor over human cloning - it's an excellent generator of fantasy-lit plot coupons, late-night drunken dorm-room philosophy, and paranoid fundamentalist ravings, but in real life it's much more boring. I know lots of human clones - they're called "identical twins", and they are little different from anyone else. And, speaking as a strong AI that is composed of trillions of microscopic robots working in tandem, it's hard to be impressed by yet another strong AI like me. There's well over six billion of us already, and that's being awfully ungenerous to chimps, gorillas, dogs, and my pet parrot. If I really wanted another strong AI, wouldn't it be cheaper and easier just to stop using birth control? Or to go to the pound and adopt a friendly but unwanted puppy?
To pick on a different example: it is the height of human egotism to believe that our scientists can create a "grey goo" that would scare a bacterium, let alone me. Bacteria have been dividing every few minutes for several billion years. They cover every inch of the earth, and compose large portions of the soil. As I type this, billions of bacteria are actively trying to digest me. It's no big deal. We grew up with this problem and have coped with it all our lives, as did our ancestors - and I'm not talking about our African ancestors, our shrewlike ancestors, or our fishlike ancestors, but about our algal mat ancestors. We've got a lot of experience, and the bacteria have even more. In a battle between a human-designed replicator and a cubic foot of soil bacteria, I'm going to bet on the machine that's had a longer design cycle.
Y'know, that would be an awesome first sentence for a sci-fi story.
"it's hard to be impressed by yet another strong AI like me" - I think the point is that it won't be like you but orders of magnitude smarter.
"As I type this, billions of bacteria are actively trying to digest me. It's no big deal" - there are several bacterium that are very successful at digesting people which are a big deal and worth worrying about.
There's a large gap in between the development of an exciting new algorithmic technology and use of it in the real world, if anyone ever takes notice at all.
The silver lining to this is that if you find an interesting but obscure paper and you implement what the authors implemented, you can be pretty sure you and the authors are the only people who ever have or ever will actually do that. It's a nice change of pace from writing CRUD web apps.
Yoda: "That is why you fail."
See http://www.ted.com/index.php/talks/view/id/38
Cf. heavier-than-air flying machines, walking on the Moon, the Internet, sub-4-minute-mile, summiting Everest, Deep Blue, the integrated circuit, Cold Fusion [1]
I think what's much more realistic and almost certainly inevitable is things like brain implants and other hybrid human-computer stuff. I'd be the first to sign up for something like that. And by the time it's ready it'll be able to store most if not all of recorded human knowledge - the implications of implanting that in a child's head in lieu of putting them through school for 10-15 years is in itself amazing.
"Reasonable men adapt themselves to their environment; unreasonable men try to adapt their environment to themselves. Thus all progress is the result of the efforts of unreasonable men." -- George Bernard Shaw
If you can give a solid reason, with any actual logic to it rather than referring to past events, I'd be delighted. Just putting quotes and referring to totally different technological developments which many said were impossible indicates that you probably have nothing because you don't know any better than I do. Blind optimism is as unreasonable as blind pessimism.
I have some reasons: a lot of people who should care about machine learning and other fields related to AI who could benefit from it don't; research, even in a hot field like machine learning, moves slowly and it takes years to determine in retrospect that a given body of work made a significant, noteworthy impact. Papers go through months of review/revising and then get published months after that; by the time you read a journal paper the original work behind it might have been done 2-3 years ago. Things move on the scale of _years,_ and I don't see that changing anytime soon. That doesn't sound very singularity-ish to me.
Now, can you give anything based on actual reason/logic/evidence, or are you just going to give me another unrelated quote?
-- Arthur C. Clarke
If you're looking at academic papers, no wonder you don't believe it.
http://groups.google.com/group/Hutter-Prize/browse_thread/th...
http://blogoscoped.com/archive/2005-05-22-n83.html
> Blind optimism is as unreasonable as blind pessimism.
Oh no.
Guess which one is more useful.
I hope my competitors are all very reasonable people.
If you believe the estimates for the computing power of the human brain, we only need a few more years of growth in the dollar/flops ratio to make the required computational horsepower available to smaller organizations. In fact, supercomputers being designed today already touch the lowest estimates of brain computational capacity. So if these estimates hold, we only lack the knowledge of how to copy this process artificially. Which certainly doesn't mean that it is right around the corner, but not looking into it would be stupid.
Then, of course, the regular pro-singularity lament goes like this: Once we have created one machine that faithfully copies what the human brain does best, the exponential growth in processor capability will ensure that we soon have an army of intelligent machines. This sentiment is easy to attack. But even if processors stop getting better overnight (which is a grossly conservative estimate) we can still find improvements to our algorithms to make our hypothetical artificial intelligences better. In nature, brain mass doesn't equate with capability. (This last sentence is anecdotal evidence, but I won't bother to dig up the papers to see whether it holds).
You cite the sorry state of machine learning/artificial intelligence to show that we aren't anywhere near what is required to implement strong AI. I agree that the fields of cognitive science and artificial intelligence are in a sorry state that barely make any effort at copying what biological brains do. But didn't you just warn against extrapolating future change from history? It is obvious to any young, ambitious scientist that our parents have been banging away at the wrong problems for the last 30 years.
This isn't a pro-singularity post, because I believe that what Ray Kurzweil and his cronies preach carries too much resemblance to a cult. They may or may not have ulterior motives, but as you say, things always look better on paper than they do in the real world. What I am trying to say is that is would be incredibly stupid not to investigate these things closer (not to _try_, throw our full weight on the problem), because the implications of successfully pulling this off would be profound.
I am totally positive that most people don't give a shit about machine learning or data mining even though it's amazing technology. I am sure I am right about that.
I share your frustration that Machine Learning isn't part of the /core curriculum/ of every CS program, but I bet it'll become more widely available at least. The job market for ML professors is no doubt going to be very, very good shortly.
As for people not caring: most people don't know how the world ticks, they take for granted that electricity comes from the power socket and food comes from the super market. They will have to care when everything changes (ie their jobs are being taken over by machines).
Oh wait... :~(.