Here's an idea - what if we ran some computer game with every possible sequence of inputs at once, and then selected the one that completes the fastest? Computer-generated speedruns.
Is quantum entanglement theoretical or has it been observed and if so how was it discovered and observed?
Dumb question, if with our current way of hardware growing exponentially every 1.5 years (or the other number if I have it wrong), how will Quantum Computers shape your living?
Because from the information I get from this video, it would basically make everything what we did know look like toddler work.
From the way I am seeing it, we currently have time lines that show what might come in the future, but aren't those time lines created from a point of view of what our technology is now?
So if this would become a real tangible thing for more ''regular'' usage, wouldn't that increase our development rate?
[QUOTE=Kaelnukem;49280305]Dumb question, if with our current way of hardware growing exponentially every 1.5 years (or the other number if I have it wrong), how will Quantum Computers shape your living?
Because from the information I get from this video, it would basically make everything what we did know look like toddler work.
From the way I am seeing it, we currently have time lines that show what might come in the future, but aren't those time lines created from a point of view of what our technology is now?
So if this would become a real tangible thing for more ''regular'' usage, wouldn't that increase our development rate?[/QUOTE]
[QUOTE=Elspin;49277379]I think the best thing they did in this video was admit that there's a massive amount of uncertainty over what we can reasonably do with quantum computers and to what extent they'll replace traditional ones, as opposed to the news articles written by someone who read 3 pages in a science text book once relating to quantum mechanics then decided to take their readers on a wild journey through their imagination :v:[/QUOTE]
Please god if you learned anything from this video don't let it be the misunderstanding that in 5 years all traditional computers will be made to look like they were designed by slack jawed retards because that's certainly not what they were saying at all :v:
[QUOTE=StrawberryClock;49279800]Is quantum entanglement theoretical or has it been observed and if so how was it discovered and observed?[/QUOTE]
It's been observed experimentally. You'd have to read up on it, but things like this are mind fucking. Even Einstein had troubles accepting quantum ideas.
[QUOTE=Elspin;49280491]Please god if you learned anything from this video don't let it be the misunderstanding that in 5 years all traditional computers will be made to look like they were designed by slack jawed retards because that's certainly not what they were saying at all :v:[/QUOTE]
Yeah - once again, Quantum computers are only better at solving very specific types of problems.
And, again, Quantum computers rely on a certain type of operating environment (shielded from electromagnetic radiation, and cooled to [I]extremely low[/I] temperatures) - they're just straight up not practical for consumer use at this time. And probably not in the foreseeable future. Or possibly ever, unless someone thinks up some ingenious way to do quantum computers without superconductors (which is probably impossible).
Are quantum computers better at rendering and high frequency trading?
[QUOTE=Jund;49276842]but can it run crysis[/QUOTE]
It can both, run it and not run it at the same time.
[QUOTE=RoboChimp;49285269]Are quantum computers better at rendering and high frequency trading?[/QUOTE]
Rendering as in drawing sprites or 3D triangles? No. This is a relatively cut-and-dry process of drawing sprites or triangles as quickly as possible; that is not an optimization problem.
High frequency trading? Probably, yes. My (limited) understanding of HFT is that a computer builds models of stocks to find the most promising ones, then trades them in high quantities very quickly; an optimization problem exists in finding the best stock(s) out of all possible stocks, which a quantum computer could potentially be much faster at.
Quantum computers will probably play a pivotal role in creation of a true A.I. IMO.
All this talk about hyper advanced computers makes me wonder. If a computer becomes smarter than any human being, how would we know? Since we would lack the intelligence to know if what it's stating is correct.
[QUOTE=Benjimon007;49287610]All this talk about hyper advanced computers makes me wonder. If a computer becomes smarter than any human being, how would we know? Since we would lack the intelligence to know if what it's stating is correct.[/QUOTE]
The computer takes an input and produces an output using instructions given to it by humans though?
[QUOTE=Benjimon007;49287610]All this talk about hyper advanced computers makes me wonder. If a computer becomes smarter than any human being, how would we know? Since we would lack the intelligence to know if what it's stating is correct.[/QUOTE]
Computers are already way smarter than human beings in the things we use them for.
No human can beat a computer in maths.
[QUOTE=paul simon;49289854]Computers are already way smarter than human beings in the things we use them for.
No human can beat a computer in maths.[/QUOTE]
That sorta depends on what you mean by that. Humans can't calculate a lot of things nearly as fast as computers can, but a human is still much more useful than a computer on the frontier of mathematics research, although computers have their place there. For instance, the computer-assisted proof of the four-color theorem is not completely human-verifiable.
[QUOTE=No_Excuses;49280929]It's been observed experimentally. You'd have to read up on it, but things like this are mind fucking. Even Einstein had troubles accepting quantum ideas.[/QUOTE]
In fact, Einstein fucking hated it due to the randomness and uncertainty of it.
[QUOTE=paul simon;49289854]Computers are already way smarter than human beings in the things we use them for.
No human can beat a computer in maths.[/QUOTE]
Computers are not smarter. They are [I]faster[/I].
The human brain is still much more efficient at anything above simple math operations.
[QUOTE=arbio22;49289809]The computer takes an input and produces an output using instructions given to it by humans though?[/QUOTE]
The instruction could be to learn, to change its own instructions based on certain findings.
[QUOTE=Benjimon007;49290722]The instruction could be to learn, to change its own instructions based on certain findings.[/QUOTE]
You are clearly not a programmer.
[QUOTE=Benjimon007;49287610]All this talk about hyper advanced computers makes me wonder. If a computer becomes smarter than any human being, how would we know? Since we would lack the intelligence to know if what it's stating is correct.[/QUOTE]
I'm far smarter than you, and you can definitely tell.
In all seriousness though, it's not like there's some hard limit to human knowledge that a computer can immediately exceed. Eventually computers will completely eclipse humans, yes. But it'll probably be a gradual progress triggered by some human intervention.
Also remember that if something is a definite fact, it'll be able to be proven somehow. There's no such thing as a fact that has no verifiable impact in the world.
[QUOTE=JohnnyMo1;49290202]That sorta depends on what you mean by that. Humans can't calculate a lot of things nearly as fast as computers can, but a human is still much more useful than a computer on the frontier of mathematics research, although computers have their place there. For instance, the computer-assisted proof of the four-color theorem is not completely human-verifiable.[/QUOTE]
Well, I meant "faster" I guess.
We built computers because they can solve the maths equations we invented faster than us.
Ok I have to ask. How would a quantum AI be different then a current AI?
[QUOTE=JohhnyCarson;49292062]Ok I have to ask. How would a quantum AI be different then a current AI?[/QUOTE]
In its most rudimentary form, an AI makes psuedo-intelligent decisions based on external stimulus. It decides the best course of action from all pheasible actions it could take, which is an optimization problem.
But that just means it can make decisions faster, rather than making better decisions. I heavily suspect that most talks of Quantum AI are done by people who don't know much about Quantum computers past "they're better than regular computers".
Well theoretically, how would it differ in behavior?
The difference I got from the video is current computers work with more regularity and certainty.
Quantum computers work on chances and probability. Im sure that would make things more interesting.
[QUOTE=JohhnyCarson;49292913]Well theoretically, how would it differ in behavior?
The difference I got from the video is current computers work with more regularity and certainty.
Quantum computers work on chances and probability. Im sure that would make things more interesting.[/QUOTE]
Unexpected behavior would be far more harmful than anything, especially in an extremely complex system like an AI meant to emulate humans.
It's a problem at the moment, but one that'll likely be ironed out in the future. Regular computers still make random mistakes too; that's why error-checking RAM exists. It just has to be brought down to an acceptable level of failure.
Quantum [I]could[/I] benefit AIs in that they may be able to process far more data than would be pheasible with a conventional computer.
The flip side is that quantum technology is not (and [I]probably[/I] won't ever be) portable, so quantum AIs would either have to wirelessly communicate with a mother brain, or be confined to a big box in a room.
[QUOTE=paul simon;49291975]Well, I meant "faster" I guess.
We built computers because they can solve the maths equations we invented faster than us.[/QUOTE]
Still only for some things. Not if it requires any sort of real cleverness, even if it's technically just computation. For instance, try to get a computer to integrate (1/x)*sqrt((1+x)/(1-x))*ln((2x^2+2x+1)/(2x^2-2x+1)) from -1 to 1 symbolically. Wolfram Alpha can't do it. Mathematica 9 can't do it. Sage has been working for several minutes (which makes me suspect it's not going to be able to do it at all). And yet...
[url]http://math.stackexchange.com/questions/562694/integral-int-11-frac1x-sqrt-frac1x1-x-ln-left-frac2-x22-x1/[/url]
[editline]11th December 2015[/editline]
Sage just spat the input back out at me, so evidently it can't do it either. If I understand correctly, any elementary function with a closed form solution in terms of elementary functions (for which this should qualify) [I]should[/I] be computable by the Risch algorithm, but none of the usual tools seem to be able to do it.
[QUOTE=paul simon;49289854]Computers are already way smarter than human beings in the things we use them for.
No human can beat a computer in maths.[/QUOTE]
computers are really really stupid, but very great at following the rules. currently and for the near future computers lack the abstract though needed to do math like a person or mathematician does
you put garbage into a computer, it will spit out perfectly logical garbage
[QUOTE=Sableye;49296403]you put garbage into a computer, it will spit out perfectly logical garbage[/QUOTE]
You put garbage in a computer, the computer crashes. Because it doesn't know what to do with the garbage.
[editline]11th December 2015[/editline]
After all, it's garbage. At best it can trigger a fault, at worst it can execute shellcode that will format the hard drive.
[QUOTE=Dr. Evilcop;49290774]You are clearly not a programmer.[/QUOTE]
You are neither, since he is not wrong. [URL="https://en.wikipedia.org/wiki/Evolutionary_programming"]Evolutionary programming[/URL] and [URL="https://en.wikipedia.org/wiki/Genetic_programming"]genetic programming[/URL] actually do what he describes, albeit in a very limited fashion. Obviously it still doesn't come close to the way humans think and learn.
[QUOTE=NoOneKnowsMe;49296633]You are neither, since he is not wrong. [URL="https://en.wikipedia.org/wiki/Evolutionary_programming"]Evolutionary programming[/URL] and [URL="https://en.wikipedia.org/wiki/Genetic_programming"]genetic programming[/URL] actually do what he describes, albeit in a very limited fashion. Obviously it still doesn't come close to the way humans think and learn.[/QUOTE]
I'm getting my Bachelor's in Computer Science, but thanks for the assumption anyway.
Evolutionary and genetic programming is not magic. It's a glorified form of trial and error. It's also already a major component in many AI systems; if you hadn't noticed, we don't have some ~super magical smarter than us~ AI yet, even with evolutionary/genetic programming in place.
And once again, the advantage AI's would gain from a quantum computer is the ability to process more data, and make decisions faster. That does not inherently make it choose better decisions, nor does it necessarily help achieve the illusion of sentience. Again, not magic.
[QUOTE=Twistai;49272086][video=youtube;JhHMJCUmq28]http://www.youtube.com/watch?v=JhHMJCUmq28[/video][/QUOTE]
I love in a nutshell.
Sorry, you need to Log In to post a reply to this thread.