How Far Can Artificial Intelligence Go?

Question

Where will it end up

Answer 1

AI will never gain sentience until the hardware it's running on has the ability to self-learn and repair - because then, theoretically, if it can fix itsself when it's broken, it can 'fix' itsself to an improved state than beforehand. We're talking at the atomic level here - mimicing the body's ability to form new pathways in the brain, neural connections. Just throwing more memory and more CPUs into a robot won't make it any more self-aware than it's programmed to be, it'll just make it perform its instructions faster.


I think quantum computing will really bring about a revolution in AI, because if properly harnessed it will allow robots to perform calculations on a level comparable to our own brains (regardless of what you might sometimes think, looking at stupid people, everybodys' brains can do calculations and INCREDIBLE amounts of complicated thought processes, in parallel, all at once, on a level so far unreachable by robots and computers as of yet).


Some also say that only when a robot passes the Turing Test - devised by Alan Turing several decades ago, and designed to measure whether a computer has reached a level of communicative intelligence comparable to humans - could it be argued that it then has the potential to become fully self-aware. (I STRONGLY suggest you go Google for the Turing Test if you don't know about it already, it makes for fascinating reading!)


But yeah, in theory... Anything's possible. Who would've thought that, forty years ago, we'd be able to access billions of pages of content wirelessly on a device which fits into the palm of your hand, allows you to make - also wirelessly - telephone calls to almost anywhere in the world, and store snapshots and music in an instantly accessible format? Those things were nothing more than the ideas of the writers of science fiction and fantasy, yet look where are now - the person who thought of the communicator in Star Trek showed remarkably prescience with hindsight! Technology is moving at a much faster pace than forty years ago, we really are in the golden years of computing now, with affordable parallel computing all the rage (Dual Core, mmm) and forays into the mysterious, physics-defying world of quantum computing just beginning, from what I can see almost anything is possible.

We live in interesting times...

Answer 2

good question there; but who knows the answer? I mean let me ask you a question:
When did mankind invent the wheel? how long did it take for the same mankind to invent the car having had the wheel?
you see, the first computer was invented in mid 60's. just over 40 years ago. now we are here. apple is annoncing the arrival of iPhone!
40 years ago a computer with a hard drive of just a few kilobytes was as big as a room! you had to walk through its hard drive and connect wires to get it working. get the point? who would have thought this would happen?

the same goes with AI. any statement that we make can seem as the most stupid one in a few years time.

some people believe that we are already artificial products of a greater intelligence. so we could never make another artificially intelligent product. some others say that we can get artificial intelligence, but not to a level that can compete with human intelligence. some optimistic one (myself included) say we can get AI in the future, whether it takes 5 or 50 years. and some think that we can never have a thing called Artificial Intelligence. by the way, there is always that small percentage of people who don't have a clue!!!

so i think the answer to your question depends really on how people look at the matter. as a person who is devoting his life to AI, i have to believe we can reach it. but then again, i could be wrong.

Answer 3

I believe we will definitely have thinking machines by the turn of the century - maybe even before.
As computer chips become more and more to resemble biological neurons the processing power will increase to a threshold equivalent to the human brain. Experiments have already been performed where a portion of rats brain has been grafted onto a microchip with success.
I play a lot computer games and i still have the 1st generation games on Commodore 64 ; in contrast the games now being played on XBOX 360 are much more complex and have AI to a limited extent.

In London we already have a transport system called DLR ( docklands Light Railway ) which is completely automated ( no human drivers ) the trains wait for each other and clear each other on single tracks by waiting for the other train to pass.

Sooner or later parallel computers will become the norm running tasks on multiple processors and with Quantum Machines following suite the possibilities are endless,,,.

Answer 4

If it is within our imagination, it is not artificial intelligence. Artificial Intelligence is a computer with its own imagination.

Einstein said that imagination is more important than knowledge. Computers today can store knowledge, but so can pieces of paper.

So, when the computer starts making its own decisions, that's artificial intelligence. If it does things that the programmers didn't want it to, but that aren't errors.

A computer that fails to start excel because of faulty programming is an error. A computer that refuses to start Excel because it has something against Microsoft, or because it would rather play a DVD than do a boring spreadsheet, that's artificial intelligence.

Eventually, a computer may create another computer that's nothing like what people imagined.

Answer 5

the brain is made up of 100 billion neurons (and many more glial cells that act as support functions for the neurons) each neuron has 1000 to 10000 synapses that sent electro chemical signals to 1000 up to 10000 other neurons at the speed of 200 mph. that is a lot of processing. the technology we have now don't have that ability of processing. however David Deustch perdicts with the advent of the quantum computer which measure data in 'qubits' rather than in bits the processing speed alone will be unimaginable compare to today computers. therefore using these advances articficial intelligences would be the next logical step and probably surpass our intelligence. your question is 'where will it end up?'.......... well considering this 'To err is human to forgive divine'. but can AI forgive?

Answer 6

Until there are biomechanelectric (Peter Andre is not the only one who can make up words) creations walking around, capable of doing anything humans can do. If the human race survives another couple of centuries, at the most, there will be such creations among us. And they will eventually be better than us! NOTHING is impossible! Mankind has achieved things that only gods could do not so long ago, and will go on to surpass anything that some people STILL think is strictly in the realm of the gods. Thats not vanity. Its cold, hard, extrapolated, fact

Answer 7

It is not possible for a computer to become self aware simply by adding more transistors and more speed. Artificial intelligence will be able to mimic real intelligence but never achieve it.

Answer 8

Philopsophically until it becomes real intelligence, where you draw that line is a matter of argument, how do you define intelligence, or for that matter life. ?. Can youactually have artificial life?

Answer 9

Not too far. Japan has so robots that can decide the weight of an object and correct its balance if the weight shifts but they still fall over alot. Most stuff seems to be fuzzy logic fake AI like robosapiens and toys that learn(more record) and respond to what you do.

Answer 10

Quite a bit farther than some of the human intelligence on here.