Ah, the exuberance of champagne singularitarianism! It’s entirely possible that one of the big names will make the first human-like AGI but I also have a sneeking suspicion that it might come from someone we’ve never heard of before, who is not part of the traditional “artificial intelligentsia”.
Also, the differences between Neanderthals and modern humans may not have been all that great, since they both had similar brain sizes. Probably the Neanderthals just lacked one small gadget, perhaps related to being able to produce language. They were also disadvantaged for reasons unrelated to intelligence. As the climate of Europe changed and large prey animals became more scarce the Neanderthals robust body design adapted to life on a cold climate would have required a higher calorie intake than the more gracile homo sapiens.
I agree that the developer of AGI could be a person unknown to us today. This is somewhat disconcerting because it means we don’t know which background they’ll be coming from (military, business, academia, etc.)
Maybe Homo erectus would have been a better species to use in the comparison.
Homo erectus may in fact be the better comparison, since we know they were around for a million or two years and never came particularly close to inventing the grilled cheese sandwich (much less the iPod.) However, if the difference between h. sapiens and neanderthals was any kind of “gadget” (genetic or man-made) then you have a great example of how much impact a very small difference can make. The secret sauce for AGI may end up being just a few lines of unassuming-looking code. But once that code gets written and implented, the potential for transformation is unleashed. So, yes, we would have intelligences as far beyond us as we are beyond h. erectus (or even as far as we are beyond dogs or chipmunks). Moreover, our world would be completely transformed. Plop your average homo erectus (or neanderthal, or even homo sapiens from 20,000 years ago) into Times Square and see what he makes of the experience. It wouldn’t take thousands or even hundreds of years — maybe just a matter of moonths or weeks before we became the cavemen in Times Square.
BTW, speaking as the filmmaker, I need to respond to the charges of “Champagne Singularitariansism.” Michael is clearly working on a nice California red, there.
@ 3 & 4: I don’t think AGI will be created by one person. The technologies that enable it will be created by thousands of people in little pieces here and there. And every time an advancement is made, we’ll see it being used in real applications. Just like the DARPA Urban Challenge today. Reaching full AGI will be gradual, I believe, not sudden.
Yah Waleed but it’ll take a single drop to make a lake flood over low-lying land in a torrent. And once it starts it self-accelerates, self-improves, the water gushing down becomes something like cement when it carries along stones, trees and mud.
AGI will attain moment from point X, somewhere between 2015 and 2035, improves itself, creates better conditions to improve itself and aquire resources, and spreads itself to survive. If it doesn’t it remains contained and containable. In fact we may be setting ourself up for a vicious selection event where only the AGI’s we do not want will be able to attain a take-off point.
October 31st, 2007 at 11:35 am
Nice to see what you’re like “alive”.
You seem to talk pretty fast at times. Do you always do that, or were you told not to be slow?
October 31st, 2007 at 11:44 am
Yes I usually talk relatively fast… also I was a bit buzzed and very excited about the event.
October 31st, 2007 at 11:59 am
Ah, the exuberance of champagne singularitarianism! It’s entirely possible that one of the big names will make the first human-like AGI but I also have a sneeking suspicion that it might come from someone we’ve never heard of before, who is not part of the traditional “artificial intelligentsia”.
Also, the differences between Neanderthals and modern humans may not have been all that great, since they both had similar brain sizes. Probably the Neanderthals just lacked one small gadget, perhaps related to being able to produce language. They were also disadvantaged for reasons unrelated to intelligence. As the climate of Europe changed and large prey animals became more scarce the Neanderthals robust body design adapted to life on a cold climate would have required a higher calorie intake than the more gracile homo sapiens.
October 31st, 2007 at 12:09 pm
I agree that the developer of AGI could be a person unknown to us today. This is somewhat disconcerting because it means we don’t know which background they’ll be coming from (military, business, academia, etc.)
Maybe Homo erectus would have been a better species to use in the comparison.
October 31st, 2007 at 5:39 pm
Great interview your enthusiasm made me excited. I never realized you talked with your hands so much. Keep up the good work.
November 1st, 2007 at 1:54 am
NO more wine for you before interviews, pretty boy.
November 1st, 2007 at 8:29 pm
OMG! You’re so hot!!!!
November 2nd, 2007 at 12:27 am
Whitey aren’t you a little young to flirt with Mike? You are only 15 ….
November 2nd, 2007 at 7:29 am
Homo erectus may in fact be the better comparison, since we know they were around for a million or two years and never came particularly close to inventing the grilled cheese sandwich (much less the iPod.) However, if the difference between h. sapiens and neanderthals was any kind of “gadget” (genetic or man-made) then you have a great example of how much impact a very small difference can make. The secret sauce for AGI may end up being just a few lines of unassuming-looking code. But once that code gets written and implented, the potential for transformation is unleashed. So, yes, we would have intelligences as far beyond us as we are beyond h. erectus (or even as far as we are beyond dogs or chipmunks). Moreover, our world would be completely transformed. Plop your average homo erectus (or neanderthal, or even homo sapiens from 20,000 years ago) into Times Square and see what he makes of the experience. It wouldn’t take thousands or even hundreds of years — maybe just a matter of moonths or weeks before we became the cavemen in Times Square.
BTW, speaking as the filmmaker, I need to respond to the charges of “Champagne Singularitariansism.” Michael is clearly working on a nice California red, there.
November 2nd, 2007 at 8:49 am
What are those structures in the background?
November 4th, 2007 at 3:37 am
@ 3 & 4: I don’t think AGI will be created by one person. The technologies that enable it will be created by thousands of people in little pieces here and there. And every time an advancement is made, we’ll see it being used in real applications. Just like the DARPA Urban Challenge today. Reaching full AGI will be gradual, I believe, not sudden.
November 4th, 2007 at 11:46 am
Yah Waleed but it’ll take a single drop to make a lake flood over low-lying land in a torrent. And once it starts it self-accelerates, self-improves, the water gushing down becomes something like cement when it carries along stones, trees and mud.
AGI will attain moment from point X, somewhere between 2015 and 2035, improves itself, creates better conditions to improve itself and aquire resources, and spreads itself to survive. If it doesn’t it remains contained and containable. In fact we may be setting ourself up for a vicious selection event where only the AGI’s we do not want will be able to attain a take-off point.
November 4th, 2007 at 7:29 pm
Nice interview, Michael! Awesome answers.
Michael G.R.: The Palace of Fine Arts: the Singularity Summit venue.