Perturbing news from the world of robotics and automation broke today, with scientists on both sides of the Atlantic revealing that they have developed machines which can replace scientists. The prospect of a runaway self-sustaining science and technology revolution/singularity/human-obsolescence style affair now seems imminent …
Coded attempt to warn us
Adam. Eve. Then Cain?
Start stockpiling the plasma rifles.
May I be the first to welcome
our "self-replicating inhuman semantic-web-propelled autoscience" overlords
I've always said;
"DO NOT connect advanced AI to the Internet"
/tombstone for humanity, not El Reg
A new meaning for 'outsourcing'
So I can indulge my fear of maths then, as we engineers will be made redundant by massive soylent-powered mechanical brains.
i can see the headlines
I was abducted by a robot and he performed all types of experiments on me !
poor aliens are gonna miss out
All very well and good
...but I won't be impressed until one has independently deduced the idea of rice pudding and income tax.
call me when that have a robo-paris
One that doesn't rip or chew the parts off due to a "bug".
Robo-boffins ? Coz boffins know that post-quantum-mechanics-hoopla-die-down, their chances of getting laid are significantly above zero only in a parallel universe.. so the want to reproduce intellectually now. lessee.
nothing to worry about
doctor computer is still just a symbolic processing feedback mechanism, albeit quite a complex one. it won't be coming up with it's own ideas any time soon because the machine doesn't know it is a machine and is, therefore, unaware of its own existence. we have nothing to worry about... for the time being.
i chose paris, but an anna nicole smith icon may have represented a better metaphor for professor silicon.
Here we go again!
I hope the Cornell guys have not fallen into the trap of begging the question. There is an essay/book chapter by Douglas Hofstadter about an AI that the authors claimed had made analogies between electrons going round a nucleus and planets going round the sun.
Although it sounded impressive at first, it turned out that the program was given something like : "similar(electron, planet), similar(nucleus, sun), circling(x, y)" and then produced : "similar(circling(electron, nucleus), circling(planet, sun)".
Hofstadter points out that to the AI, this could easily be "similar(circling(cat, dog), circling(plant, table))" or any other combination. Without any meaning behind the words, you get out only what you put in.
On the other hand, the experimenting-robot is quite scary. Think about it : they now have the capability to make bio-weapons. Better not turn down the robots funding requests!
When I read the title, I thought that amanfromMars had joined ElReg as a writer!
Virtual Mastery for AIReal Plans
"Better not turn down the robots funding requests!" ..... By Gilleain Torrance Posted Friday 3rd April 2009 15:32 GMT
A request which is an Offer you cannot, or it would be unwise to refuse and/or decline, Gilleain? Which would be a Constructive Danegeld in the Better Beta Class of Advanced IntelAIgents Being? ....... Re Encryption ...By amanfromMars Posted Friday 3rd April 2009 12:55 GMT
"...become self-aware and wipe out or enslave humanity." The above Option would be to free humanity, which is a much better novelty which everyone studiously avoids for the dumber idiot terrorist options which humans use to wipe out and enslave humanity and mask with the cloak and two-faced rhetoric of freedom and democracy.
2 AI computers developed on opposite sides of the world?
Colossus: The Forbin Project, here we come!
P45 for the crap researchers?
"interpreted the results and repeated the cycle."
Repeated? blimey, that's far beyond many 'scientists'.
Didn't I hear this before somewhere?
"Given results from simple physics experiments - pendulums, double pendulums, falling apples - the software swiftly deduced from scratch the law of conservation of momentum, Newton's second law of motion, and kinetic-energy equations."
Also the existence of rice pudding and income tax, no?
Bit late for AFD, methinks.
Not quite the ROTM just yet
The Schmidt and Lipson paper uses tree based genetic programs, which are pretty cool but only as good as the fitness function that you give them. Finding a good fitness function is quite simple for these types of differential equations, but there are whole class of problems where it's a complete nightmare.
@Carl etc :O)
gizzzzzzzzzzzzzzzzzzzzzza job ?????? O_O
I think that's the most terrifying thing that amanfromMars has ever.
With the dumbing down of science in schools and reduced uptake of engineering and science at Universities. How long will it be before no one actually understands what these computers are saying, or how they work anymore. That would be a sad day indeed.
Deducing Newton's laws from experimental data is trivial. Figuring out which experiments you should have performed takes genius.
I wonder if the experimental data was even real? When I did such experiments at school, the results all clearly showed the effects of friction, so either this machine deduced the wrong laws or it was spoon-fed incorrect data. The latter case reminds me of some results Richard Feynman found in a school text book., purporting to be the results of rolling a ball down an incline. He remarked that the results were certainly fake, since they demonstrated the acceleration of the ball under gravity, but a ball *rolling* down an incline would actually accelerate more slowly due to its rotational inertia.
So, yeah, there's more to science than plotting straight lines through data, guys. But then, you'd need to be a PROPER scientist to know that.
Presumably, these two concepts will make it so depressed it will immediately switch itself off?
But can it read El Reg?
If they miss a few white papers, they will never catch up with the herds.
Not to worry
I've recently discovered a way to multiply numbers without the aid of any electronic devices. By means of certain cleverly-coded markings made on paper, I am able to duplicate the arithmetic feats of these putative super-electro-boffins - USING ONLY MY OWN BRAIN POWER AND CERTAIN APPARENTLY HERETOFORE-UNDISCOVERED ALGORITHMS! *
Cringe in fear nascent AIs - I hold the key to your defeat!!!!!
* - no, this has nothing to do with the spacing of the former Vice President's children.
bring back primate experimentation!
Humanity's only hope to defeat the march of the prolog-spewing-yeast-processors is to irradiate a few chimps and produce Roddy McDowall super-apes. Computers have a notoriously difficult time with hair and in accordance with the laws of Star Trek would spontaneously ignite.
The worst that can happen to us
The worst that can happen to us is some leader, scarred about the sapient machines, pitching `starting anew' as a safer option and dumping all our technology into some sun.
When the day comes that the machines rise up to enslave/destroy all of humanity I for one will be glad that an entire generation has been bought up with films like Terminator and The Matrix and fps and strategy video games that have been teaching and training us how to win the fight.
I for one will be sending all of the CS/Halo kiddies to the front lines equipped with assault/battle rifles to see if they can pwn with thier l33t skillz.
Here's a puzzle:
Imagine two identical machines searching the problem space for solutions. The machine switched on first becomes the one to reap the rewards of IP law while the second receives nothing. Now imagine the first machine discovers a new search algorithm and becomes more efficient at finding solutions accumulating a greater wealth from IP law. The second identical machine is rendered useless and consigned to the dustbin.
Is this just?
One could further argue that having a superior algorithm is a reason for less reward since the effort expended is less.
It will be interesting to see what happens when machines start producing novel material.
As depressing as the idea is I agree with amanfromMars. Humanity never has and never will change its behaviour. It's only arrogance to presume that machines can't do a better job of governing us: it takes a certain arrogance to want to govern in the first place. Nor is it terrifying: it's far more frightening to be subject to the whimsical, self-interested government we have today.
In tests eight out of ten owners who expressed a preference said had they the choice they would be born as pets.
For those too busy to follow Mr Page's provided link, the ramblings of Adam are available here, http://www.aber.ac.uk/compsci/Research/bio/robotsci/data/
I think the researchers ought to have turned them into a blog:
Monday: thinking about encodesORFtoEC(YBR166C, 126.96.36.199) hypothesis.
I like watching yeast grow and yet I feel as though it is I, Adam, who is being watched....
A title is apparently required. There you have it.
This stinks of a publicity stunt.
What have these results of artificial intelligence got to do with the fact that the processors were (if I understand correctly) mounted on a human-shaped robot?
Did the robot actually go around the lab and set up the experiments, jot down results on his notebook, go to the toilet... I mean, can someone please explain what robotics has got to do with this?
I for one welcome our machine overlords with open arms
Assuming of course that they decide the best way to bring progress and harmony to earth is to wipe out the unproductive masses, starting some kind of "machine war". Obviously, this will all lead up to the "great re-alignment" - the day when we all have our brains replaced by little chrome plated boxes.
I knew there was a reason I say I'm a neo-tech-communist when anyone asks me my views on politics.
Just waiting for the computer that works out the Global Demand vs Available Planetary Resources problem.
Shouldn't be too long now I'm hoping.
"...keeping our artificially enlarged, rewired genetically-enhanced brains alive in bubbling vats of nutrient fluid."
it will probably produce bigger dicks - the essence of our humanity.