Kurzweil predicts that machines will match man by 2029 -- bring it on
By Ryan Block 
posted February 17th 2008 7:11PM
posted February 17th 2008 7:11PM
[Thanks to everyone who sent this in]
posted February 17th 2008 7:11PM
A look back on popular stories from today in a specific year.
Now that we've thrown 'em off the trail, use the form below to get in touch with the people at Engadget. Please fill in all of the required fields because they're required.
5. Aliens defeat both us and our machine overlords.
dam those pesky aliens
One word: Hogwash!
EMP > Hydraulic Pumps
There will be cake!
Surely we already have AI that is smarter than the average Engadget commenter? I know for a fact we already have machines smarter than the George Bush,
it's called a wrist watch.
Ha ha, douche.
Not only will they play doom, but they will finally be able to enjoy it.
George Bush vs. Miss South Carolina...
Either Fight of the Century, or Complete Waste of Time. You decide.
Cue Star Wars music:
-"Miss South Carolina. I am you father."
+"Nooooo. That isn't posi .. posibuh ... posibubble."
-"Search you beliefs. You know it true."
In all honesty, I don't believe we will ever achieve true artificial intelligence. I think we will reach something like the NS4s in iRobot (the dumber robots), but never real artificial inteligence. I just think that programing and human though are two different. Also, it is just too complicated to truely be able to replicate. Maybe we will get something similar, but we really know to little about our own selves to do something like that. It ain't happening.
Everything that appears in the nature can be replicated, solely because of the fact that there is atleast 1 mechanism that will end in that result, finding it is only matter of time
I agree. AI of even a primitive form is not happening anytime soon if at all. We are definitely not going to be able to program and understand it for sure. We may be able to simulate a mind by understanding the very basic pieces without understanding the organization but this too is highly unlikely.
I seem to remember a slashdot post about a year ago about how DARPA had successfully taken 10 CIA agents and rigged 5 of them with some sort of interface from their brain to a laptop (carried in a backpack). I have no idea how they did this (just read the slashdot post, not the article. I don't care about 40 page whitepapers when 1 sentance says all I'll understand anyway). They took all 10 agents to a local grocery store and the ones with the assistance of the laptop recognized and identified common household produce (fruit) something like 60% faster.
Personally, besides the obvious comment of "it's already here" I don't believe that the way of the future will be the construction of robots to do our bidding, but rather the addition of robotics to humans. Cyborgs sound MUCH more realistic to me than pure robotic AIs. I don't believe that you can compensate for a lack of emotion by simply quadrupling the CPU power, but I do believe you can give a 30 year old man who was born blind an ocular implant and 20/20 vision in an hour, and even if not yet, soon. Just the same, I don't see how it is that far fetched to create implans similar to those they're putting in deaf people now that give a person the capability to hear higher or lower frequencies than a normal person. I believe an artificial eye could eventually see ultraviolent and infrared light, and even be made to be turned on and off with a simple thought, but I don't believe a robot will ever be able to make the split second gut desicion to save a child in a burning building instead of the elderly woman when the elderly woman would be easier. A robot will never be able to factor in the lifespan of the two people, because though that may appear to be simple programming logic, it's not. It's emotion, and when the elderly woman is 10 times easier to save, the computer will do the rational, yet emotionally wrong action, and save her over the child. That's the problem with the entire theory of robot masters such as those in Terminator or The Matrix (both great films, BTW). The general theory behind both movies is that robots will eventually advance to a point where the concept of power (political, not electrical) or even the more basic instinct of happiness will become mere lines of code, but with a lack of endorphins or a feeling of reward, robots will never have the underlying reason to pursue world domination, and without motivation, nobody - or nothing - will do anything. I can, perhaps, see survival as basic logic, but that's a reason for robots to prevent the mas production of other robots or the blackening of the sky, not to accelerate it. From a basic survival point of view, the fewer other robots created to consume a limited supply of electricity, the better for each robot. That's the same reason humans fight wars - if there was infinite food, booze and land for everyone, nobody would care to own any of it. Strangely enough, it's basic supply and demand.
So though I do believe this may be a good post, and perhaps even an accurate estimate of whien certain events will take place - I do not believe that the basic concept behind an AI will ever match that of a human. At some point, yes, a spacecraft flown by an AI will dodge asteroids better than any human pilot possibly can, but if a firefighter robot still cannot choose the small child over the elderly woman - against rationality and odds, but emotionally the better choice - then it will never truly be an AI, and the simple fact is that irrational decisions cannot be coded in any language. Period.
Same timeframe Ghost in the Shell takes place... I'm all in for it. Bring on my cyberbrain and wang.
His books are quite an interesting read.
this TOTALLY reminds me of an episode of the office season 4 where dwight tries to outsell a computer.
LMAO.
Unlock the door?
Lock the door?
It's funny Kurzweil is mentioning all this now. In "The Age of Spiritual Machines," which he wrote 9-10 years ago, he said the same thing, generally speaking. For a man to stick to his guns like that, especially in light of most of what's happened since 1999, can say a lot of things. Though I must admit, we should hold out till at least 2014 to see if any veracity to that statement is possible.
/And yes, I did put that reference on top in there intentionally. I'm made to heal. :D
Isn't global warming supposed to kill us by then?
That's where the robots come in...
If they do... I'm totally going to make sure we have arenas to have them fight each other. Unreal Tournament/Halo style...
hmm, lets hope i finish my plasma rifle by then, plasma should melt through most metals yes???? (help me out here guys)
ah, but if the machines got hold of the technology they could super mass produce it. man looks like i better get my time travelling boots on.
"Plasma" is just another word for "fire", so you'd at least need a rifle shooting a substance burning at over 500 degrees / C to melt most metals. Or you need thermite. And, even then, thermite takes some time to melt metal - it's not instant.
I guess a big frickin' laser on a shark's head would be a better choice. ($1 to Austin)
The problem isn't with Kurzweil's estimation of computing power, but rather with his incredibly limited view of a human brain. When you grossly underestimate the capabilities of the brain, it's easy to assume that computers will catch up in a few years!
In the real world, of course, we still probably won't have a realistic and objective definition of intelligence within the next 20 years, much less a "nanobot" that can increase it (whatever that would even mean!). Absolutely ridiculous. They thought we'd have natural language processing computers back in the 80s, too. There are only so many consecutive decades in which you can make the same predictions and still be taken seriously.
I hate you all, your ignorance on this topic is outstanding. Judgeing by Average Joe Engadget's opinion computers already exceed humanitys intelligence more than we realise. DOS is smarter than most of you.
2012 conspiracy theorys? THE END IS NIGH? Shit a brick and give me a break you fucks.
Hey......Fuck You. Seems like you can't take a joke
Kurzweil is certainly someone that has a good grasp on technology but as far as how accurate his time frame is, only time will tell. What people need to be concerned with right now is once robots are no longer perceived as cold, expendable machines but as creatures with their own identity and emotional capabilities, we will have to grapple with all the same legal and ethical issues that afflicts humans and animals today. In some sense, the threshold has already been surpassed with such toys as the Sony AIBO robot dog, which while not being anywhere close to what Kurzweil speaks of technologically, was advanced enough for people to establish strong emotional ties. It is critical that lawmakers establish forward looking regulations that covers robotics, or else we are going to end up with a whole big mess on our hands in the near future.
Ah...he's just saying 2029 because we all know the world comes to an end when the UNIX date problem rears it's head.
I suggest you join and donate b/c this is the only org that knows what the **** is really going on;
www.lifeboat.com/
"Futurist" I love that title. I'm gonna start putting it at the end of my name from now on.
-John, Futurist
Well, maybe they'll be smart enough to figure out all those biochemical processes we can't quite crack to cure the diseases we're taking forever to figure out.
But anyways, since we all gotta die someday...why not exterminate ourselves by manufacturing TERMINATOR units?
Its tough to define what is meant by 'exceeding' human intelligence. A lot of this is tied to the emotional and motivational side of things.
We get up in the morning and we are hungry. We perform actions so that we can earn food. We also seek clothing, shelter, procreation, etcetera. There are also human motivations regarding curiosity and entertainment. As a human we could just sit there all day, or we could do something. Our motivation to do something might be different for different people. It could be to achieve an ongoing existence, or it could be to relieve boredom or to address curiosity. Or even a need for recognition. Furthermore, each human may react differently and perform different acts.
What is the motivation for a robot? If they perform because they must then are they really as 'smart' as a human? If every robot is equally competent then acquisition of knowledge becomes predictable and linear. For any problem there is there only one 'proper' and best solution. Or is there room for creativity and uniqueness whereby one robot might develop a novel approach compared to another.
Anyway, this question doesn't make any sense to me unless you define things a bit further. Define robotic motivation. Define creativity versus 'optimal' approaches. Then I'll understand what is meant by equal to humans.
Michael