Theory about future technology

[Lee85]

As most of us will know technology is constantly getting better and better all the time, And what I was thinking is that there are scientists all over the world working on AI and synthetic brains, so in the future Would they be able to take say a virtual person with compleate ai and understanding into a simulation of earth and simply press fast forward and because the person will have unlimited knowledge would it be able to create new technology so if we fast forwarded millions of Years and see what it's created we can use it now? This means we could have technology millions or even billions of Years more advanced than we we should really have. I know this is only a thought but I don't see why this wouldn't be possible.

Don't call the men in white coats haha .

I look forward to hearing peoples opinions :-)

[iwatkins]

Thing is you need some very special computers to run a life scale simulation at real time never mind in fast forward. So you would need to have the system invent new faster computers first.

No wait...

Hang on...

(Deep Thought anyone?)

[Lee85]

Haha yeah I realise that we would need more than super computers to run this but in the future this type of computer maybe around,

Or deep thought lol I'll come back in seven and a half million years :-p

[Capricorn]

and because the person will have unlimited knowledge

How could they have the knowledge of what happens tomorrow?

Or what is discovered/developed tomorrow?

What is unlimited knowledge?

Evolution is random, if the random event is positive then it may remain. Which random event is chosen as the one that will happen?

Do we really have much new technology, or have we just refined what was around years ago. All we have done with the mobile phone is reduce it in size. When did Marconi start using radio?

Look at the advancement made in going to the moon in the last 40 years. Except we haven't.

Petrol engines, not exactly new.

TV - how many watched the coronation in 1953?

Hate to say it but I suspect the last leap was the transistor, which came out in the early 50's.

[ollypenrice]

I'm not convinced that computers and brains have much in common at all other that our brains invented computers.

Nor am I convinced that the phrase 'artificial intelligence' really means anything. So much language in the world of IT is anthropomorphic that it is no longer possible to think straight about the brain-computer debate since we have foolishly applied human terms to aspects of these machines. Memory. What memory? Does a piece of paper with 'E=MC squared' written on it remember Einsteins equations? It would be a very odd use of the term 'remember,' just as applying the term 'memory' to code stored in a machine is a very odd use of that word.

'Intelligence' in its early form meant 'awareness of.' What, in a computer, is aware? Now we may one day make machines (Turing machines) which respond as a human would respond and we may be unable to tell the difference (over the phone). But for me that is not proof of their equivalence.

I think computers might one day run amuck and 'take over' their own functions, ignoring our efforts to regain manual input. But to my mind that is simply analagous with a runaway train and is unconnected with intelligence. I think a word based on 'automation' rather than 'intelligence' would be more helpful and less likely to encourage us further to anthropomorphize computers.

Here's another thought about technology; we see it is a form of progress, but might it be the reverse? In the past environmental change has driven evolution by natural selection. Now we use technology to force our environment to evolve. Have we, perhaps, condemend ourselves to spend the rest of our days on a evolutionary plateau of our own making? Hmmmm...

Olly

[JulianO]

In popular culture such things are called singularities.

There is a rather disturbing novel which you can freely read about such an event, mankind taken over by machine for it's own good.

Be warned - it has strong language and concepts in it, but I found it fascinating none the less.

The Metamorphosis of Prime Intellect: Contents

[ollypenrice]

How could they have the knowledge of what happens tomorrow?

Or what is discovered/developed tomorrow?

By discovering that tomorrow doesn't exist? In other words by discovering things that lead to a better theory of time in which the notions of 'past, present and future' are no more than local conveniences, like all the theories of physics which became generalized by Einstein. If, as many think likely, our past-present-future theory of time is incomplete, who knows what a mastery of 'time' might bring? I use inverted commas because this would not, Jim, be time as we know it!

Olly

[phat tony]

i do think we could do what your saying but not without the right people and funding. i see huge problems with the human race, money is one but the more important is smarts. people judge intelligence on education and thats far from the truth. i am a great example. i am not any super brain but i am a lot smarter then i get credit for. i left school at 15 because my father had a breakdown and my family was starving. i had no choice i was doing great in school but we needed money in the house. now when i tell people i never finished school they think dumb dumb. but when they get to know me they are amazed with how wrong they where. i can learn and do the most complicated of tasks very fast. now enough about me. i am Irish and here they don't support inventors very much. the next big thing could have been invented 20 years ago and the person was ignored because of a lack of education, money or they didn't know any root to market. DA Vinci was an uneducated man (self educated)and in my mind one of the best inventors of all time. the computer tech you are talking about could have been invented years ago or maybe is in the planning or to soon be born in someones mind who will face the same problems.

one other thing is, computers are far more advanced then we know, the companies like Intel drip feed us better and faster equipment because of greed. if they released their grade A tech now the whole world of tech would move so much faster. if they done that from the word go we would be a huge amount of years ahead of where we are now. greed is slowing everything down. they want people to buy the new improved version this year but the truth is they had the tech years ago. profit will always beat the want for advancement.

[gajjer]

Thats an interesting thought. Here is my take on it.

I think Olly is right that AI is largely automation. Getting a robot to move round a space with obstacles is a very limited form of inteligence. For advancement of technology you need new ideas. For example, on discovering electro-magnetism its not a simple deduction to invent a motor. Someone had to have a need and look for solutions to that requirement. Lasers didn't spark the idea of a printer or range finder or whatever. Someone wanted a better way to print and the laser was one of many ways of solving the problem. It also requred a whole heap of building prototypes and ironing out problems - all of which takes time and manpower.

Simulation software that I use simply calculates what something would do given the constraints of the components and with specified operating conditions. It doesn't predict - it simply calculates. I'm not sure we understand our own invention processes so automating it would be difficult.

There is then the question of what AI would determine as a need. Can a computer/robot have a need? I don't know. I didn't even know I needed an iPod until someone told me I did! And now I do!

I'm not saying it can't happen but so far I don't think I have seen any computer carry out a task that could be called inventive.

Interesting thought though.

[gajjer]

Here is another thought.

We can create electostatic forces. We can create electromagnetic forces. I don't believe we can create gravitational forces using electricity. How would a computer go about solving that problem?

[rfdesigner]

....This means we could have technology millions or even billions of Years more advanced than we we should really have. I know this is only a thought but I don't see why this wouldn't be possible.

The problem is that computers today.. and I'm not talking laptops, I mean the best of the best supercomputers with neural functionality are still not a patch on the brain.. and that's when they're running at full speed. Next imagine that we manage to create a machine brain equivelent to the human brain. You can be 100% sure it won't run any faster simply because it will be running at full whack already. Ok.. let's invoke moores law, and wait 20 years (equivelent to a 10000x improvement in processing power). well you could run a single human 'simulation' at 10000x speed, but.. last time I looked the world wasn't run by a single person but by a vast army of people across every part of the globe. So while I'm working on plasters with chips in for health monitoring in the UK some bright spark is working on new lossless transmissions for cars and someone else is thinking about how to use anti-matter at cerne and so on, if only 1% of the planet is really thinking at any one time, that's an awful lot of processing power.

So actually you'd need a whole army of the computers... anyone want to program them?

No that won't be the way development happens. What will happen is that computers become ever more part of the R&D process.

20 years ago a simple mobile phone needed a whole room of RF engineers to design just the radio bit. Now a team of maybe 3 or 4 people could do almost the whole thing, because the computing power available is so much better, the models are better, we can simulate the radio fields meaning we don't need to test things for real. In car design the vehicles are all crash tested in a computer 1000s of times so when they are formally crash tested by NCAP there should be few suprises. Having a real mega computer for this will simply shorten the R&D process further.

The development curve will just get steeper as it has done every year for as long as I can remember... unless we hit a wall on moores law which has been long promised and has yet to happen.

Derek

[Macavity]

It's clear that neural networks(?) have long been capable of "educating themselves". I sense we are already on (or exceeding?) the threshold, where machines are solving routine tasks, now beyond human understanding. Re. independent "machine inspiration", I kinda hope not? But / and, their input data is ever limited by OUR current experience of the world? We TRY to learn by (updated) experience, as we go along...

[phat tony]

we where the ones to teach computers how to play chess and now only the best can beat them, so they can pass the average people. their is so much that can only be done with computer programs. when we machine things we can do a good job but with computers involved they can do an almost perfect job. for years people have said that computers don't make mistakes its the programers who made them mistakes and they are right. now my pc is always making mistakes but for the reason i mentioned bad programs with too many flaws. can a computer create something based on what we can teach it, i don't think so. i was talking to my wife about this and she said what i was thinking. if we teach a computer to bake a cake and then every cake. now if we teach it what we like to eat and what we don't like to eat would it be able to CREATE a new type of cake???. to me that would prove they can create rather then copy or display what we teach them.