Science What will computers be like?

[APDAF]

I hope there will be an AI inless than 100 years.

 

[T.Neo]

20 years - computers help us control our money, house, health, work, everything else.

Don't they do this already?

50 years - complete interactiveness, be able to surf the internet shops, not by clicking, but by immersing yourself in a fully interactive 3D environment, ebay will be like a real-life super market, all simulatored by a headset, goggles and sensors of some sort. and any material goods will be able to instantly materialise for use at the touch of the finger... (no mouse in a 3d interactive enviroment) whether it be food, clothing or accesories.

Other people have suggested this, and frankly I think it is pretty silly. People can navigate the 'net just fine with a screen, keyboard and mouse. In fact, they can probably do so far more efficiently than in some sort of silly VR environment. The only reason for such a VR environment would be for enjoyment of the environment itself (rather than actually using it for retail or research).

And I mean, are the people in this future all too agoraphobic to go to a shopping mall?

Materialising anything you want instantly is pretty... impossible, but you might have some sort of '3D printer' in that timeframe, that could create a lot of stuff using data downloaded or purchased off of the internet- including some foods.

100+ years - Something similar to the Holodeck or like that movie (gamer) i think its called?

Likely not something that works like the Holodeck... this was based on Star Trek physics of 'holograms' and 'forcefields', and we all know how... rubbish Star Trek science is.

We have the cellphone though, which some compare to the 'tricorder'...

500+ years - so our population doesnt have to suffer through what our fragile planet will probably become to by them, all people will abandon their material form for a life of data and variables, living completly inside a super-massivly multiplayer super server computer. we will have a choice, become a online data stream, or go into space and look for a new planet.

I have to agree with various other people here, I think that is complete rubbish.

There is no "abandon the physical form and live as data", there is only "kill yourself and create a likeness of yourself inside a simulated environment". This is exactly what it is, no matter how cheerfully people will avoid using this terminology.

Also, 500 years is not a likely time for a "collapse of a fragile Earth". The most critical time is probably the next 100 years, while population is rising and the necessary technology for sustainable living isn't mature yet. When we cross that point- if we cross that point- then we won't really have a problem on Earth.

But; good luck finding another planet. We're sitting on the richest one that we know of...

The current trend of lowering the main clock speed is NOT indicative of the computing technologies' development slowing down or even reversing. If you look at the transistor count, feature size and ANY performance benchmark, you'll see that computing technology is still advancing exponentially as it has been for the last 100 years, and it shows no sign of stopping.

At least Turbinator provided a graph that that clock speed is hitting a sort of plateau at least in some cases (and while clock speed is only one measure of performance, it is a measure of performance nontheless). Do you have a source that shows the growth of the other measures of performance you just mentioned?

By 2020, the computational equivalent of a human brain will be available at the price accessible to middle-sized corporations. By 2040, the computational equivalent of all of humanity will be available for the price of a today's personal computer.

I think it is pretty foolish to assume current growth trends into the future without looking at technological and demand limitations. You could take a lot of past technological capability growth rates for example, and carry them on into the modern day, and get pretty silly results.

The moment we construct a superhuman intelligence, we aren't needed for further development.

Wrong. You still need people to input commands and interpret results, as well as oversee the operation of the system. The alternative- having little to no control, or having the system control itself, would be a pretty bad descision in the world of anyone who does not have the worldview of an idealist dictatorial ideologue.

 

[RisingFury]

Its going down? D: if thats true i think the problem may be that more cores is the wrong approach to improved computing!

Let's let people who actually know a thing or two about computers and design them decide that, huh? :thumbup:

 

[Mojave]

Whoa, that video was amazing. I wish I could have a video wall. :hmm:

 

[n72.75]

Scary thought 8192: Computers only need to get as smart as humans before they can start making themselves smarter.

 

[PeriapsisPrograde]

If we made a computer the size of a room again, couldn't we do more or less whatever we want?

 

[Pyromaniac605]

If we made a computer the size of a room again, couldn't we do more or less whatever we want?

We have made computers the size of a room multiple times in the past few years, but to be honest, they're never really that spectacular. :dry:

 

[RisingFury]

This looks quite interesting too...

http://www.microsoft.com/surface/en/us/default.aspx

...Maybe my 50 year mark should be alot earlier :O

There's an awesome parody on YouTube about this. Just look for "Microsoft Surface Parody".

 

[T.Neo]

We have made computers the size of a room multiple times in the past few years, but to be honest, they're never really that spectacular.

They sure are, compared to the room-sized supercomputers of yesteryear...

Scary thought 8192: Computers only need to get as smart as humans before they can start making themselves smarter.

Huh? That makes no sense... I'm as smart as a human, and I haven't suceeded in making myself smarter yet. :uhh:

 

[Artlav]

Huh? That makes no sense... I'm as smart as a human, and I haven't suceeded in making myself smarter yet. :uhh:

That's because you have no access to your source code, hardware design and equipment/knowledge on how to make more of you out of it, with directed alterations.

An AI would have that.

 

[APDAF]

An AI that can make it self smarter?
Lets hope it does not become GLoDOS.

 

[jangofett287]

An AI that can make it self smarter?
Lets hope it does not become GLoDOS.

There's plenty of stories about AI deciding to kill/imprison the humans that created them. It is defiantly something that a sane person would hard-code a safeguard against. Oh, and by the way, its GLaDOS not GLoDOS.

 

[APDAF]

That was a typo.

 

[T.Neo]

That's because you have no access to your source code, hardware design and equipment/knowledge on how to make more of you out of it, with directed alterations.

An AI would have that.

There are various AI programs that don't have that, so it certainly isn't something that just pops up automatically whenever you have an AI.

Also, imagine the sort of lag that would go on if you thought about every single thought you had. What would stop you from thinking about thinking about thinking about thinking about thinking about thinking and crashing in a sort of recursive loop?

I don't think something like the human mind is particularly is particularly suited to that sort of mode of operation, to be honest.

There's plenty of stories about AI deciding to kill/imprison the humans that created them. It is defiantly something that a sane person would hard-code a safeguard against.

I think the assumption that a generalist AI would be some sort of "artificial person" existing inside a computer is a really silly cliche, that a lot of people seem to like, but it really, really doesn't make sense. It actually defeats the purpose of having an AI in the first place.

The real generalist AIs that we'll see will be nothing like the human mind; they'll be very powerful programs, and they'll be able to do a huge number of things in an amazing way, that trying to emulate a human would just be a waste of memory and processing power.

Of course, that way, you also prevent the moral concerns of creating a sapient being solely for the purpose of being a slave. Or the concern of a powerful sapient being becoming uncooperative, unconstructive, or worse. That does make it far harder for a sci-fi writer to write interesting stories about it, but it's also pretty hard to write an interesting story about a safely operating nuclear reactor, or a launch vehicle that actually flies well...

It would be superhuman, but this is nothing special. Superhuman abilities have been the staple of our technology ever since our ancestors knapped flint tools more efficient at cutting through flesh than their own bare hands.

 

[Jarvitä]

Huh? That makes no sense... I'm as smart as a human, and I haven't suceeded in making myself smarter yet. :uhh:

Are you a neurobiologist with perfect understanding of the brain down to the synapse level, and a tool to move neuron connections around at will? An AI would have the equivalent of that once you give it its source code and the specifications of the hardware it's running on.

Of course, that way, you also prevent the moral concerns of creating a sapient being solely for the purpose of being a slave. Or the concern of a powerful sapient being becoming uncooperative, unconstructive, or worse. That does make it far harder for a sci-fi writer to write interesting stories about it, but it's also pretty hard to write an interesting story about a safely operating nuclear reactor, or a launch vehicle that actually flies well...

It would be superhuman, but this is nothing special. Superhuman abilities have been the staple of our technology ever since our ancestors knapped flint tools more efficient at cutting through flesh than their own bare hands.

A true artifical general intelligence wouldn't need to be designed to be a slave, it shouldn't need any inherent rational biases in the form of hard-coded blocks. If it truly is superhuman, there's nothing preventing it from subverting those anyway. The AGI should be designed and programmed from the start to be friendly to humanity, to realise in a perfectly rational manner that we are the reason and purpose of its existence. That way, even when it achieves far superhuman intelligence, it can act as a transition guide/matter operating system to move humanity to the posthuman world safely, avoiding any existential risks such as nano grey goo, unfriendly AI and any number of others we can't even comprehend at this level of intelligence.

 

[RS-232]

50 years - complete interactiveness, be able to surf the internet shops, not by clicking, but by immersing yourself in a fully interactive 3D environment, ebay will be like a real-life super market, all simulatored by a headset, goggles and sensors of some sort. and any material goods will be able to instantly materialise for use at the touch of the finger... (no mouse in a 3d interactive enviroment) whether it be food, clothing or accesories.

100+ years - Something similar to the Holodeck or like that movie (gamer) i think its called?

In 50 to 100 years as computers get faster and hardware integration becomes smaller and when we reach the quantum level and have the ability to take a snapshot of the human body ,store it, then shoot that stored info to be reassembled at another location ,we will then be able to "Beam" ourselves from one place to another. This is already being done with particles now. Will teleportation lead to quantum computing or will quantum computing lead to teleportation? It should be fun to watch evolve either way.
http://news.discovery.com/tech/teleportation-quantum-mechanics.html

 

[T.Neo]

Are you a neurobiologist with perfect understanding of the brain down to the synapse level, and a tool to move neuron connections around at will? An AI would have the equivalent of that once you give it its source code and the specifications of the hardware it's running on.

Er... yeah, except it isn't that simple either. Even if you did give me all the details to how my brain worked, it doesn't mean that I could change it- likewise an AI would require the ability to usefully interpret that information and modify it, in order to improve on its software.

And of course, it would have to know what to do, where. You don't just have that ability instantly.

A true artifical general intelligence wouldn't need to be designed to be a slave, it shouldn't need any inherent rational biases in the form of hard-coded blocks. If it truly is superhuman, there's nothing preventing it from subverting those anyway.

I may be not superhuman, but I am human. And I can't really subvert my evolutionary hardwiring. I'm not so ultra-special that I can do that... even suicide which can be said to go against those rules, actually goes along with it, especially in cases of depression (the mind can work itself into a bad place, that you have huge difficulty getting out of, even if you try).

"AI knows better", is part of the highly disturbing "Superintelligent AI God Cult" that seems to have popped up in recent years...

The AGI should be designed and programmed from the start to be friendly to humanity, to realise in a perfectly rational manner that we are the reason and purpose of its existence.

A human can realise in a perfectly rational manner that its parents are the reason and purpose for its existence.

A human can realise a lot of things in a perfectly rational manner.

And it can still go against all those things.

You say that the superhuman AI can magically subvert its own programming rules, yet you go on to say that they can be programmed to be 'friendly'. Why don't they subvert that programming as well? I'm friendly to a lot of people, that doesn't prevent me from going against them if I wish.

Of course if the programming means that this AI is so 'friendly to humanity', and that is so hardcoded into its being that it can't decide otherwise at all, how different is that to programming it to be a slave?

That way, even when it achieves far superhuman intelligence, it can act as a transition guide/matter operating system to move humanity to the posthuman world safely,

You mean murder billions of people and run a gigantic computer game?

One thing I find particularly disturbing about transhumanism is it seems to give itself the self-ascribed ability of determining who or what is "better" or "superior". In the human world, doing so is generally a recipe for very bad events.

I remember a warm and fuzzy little piece of fiction, that had a lot of posthumans in it- miniscule worm-people, blind cavedwellers, and sapient quadrupeds with no way of usefully interacting with their environment, the results of radical genetic engineering following an alien invasion. The label of 'posthuman' was rightly applied, but very different from the ubermensch idealism that a lot of people seem to have.

avoiding any existential risks such as nano grey goo,

Didn't nanotechnologists such as Drexler himself declare that grey goo isn't a real threat?

unfriendly AI

Who determines what is 'friendly' and what is 'unfriendly'? It could certainly be totally subjective.

I sure hope you don't place your trust in people based on how 'friendly' they appear...

and any number of others we can't even comprehend at this level of intelligence.

Concepts such as "uncomprehensible idea" and "this level of intelligence" are at best absurd, and at worst totally unscientific.

There is no evidence whatsoever for anything like "intelligence levels". Even sapience is not a magical 'intelligence singularity', it is a set of traits that allow the manipulation and creation of information as subject the same logic that governs all reality. There's no evidence or reason to believe that there is another "sapience level" above us, or indeed, that sapience is a level of any sort at all.

Furthermore while there are many things that can't be fully or completely comprehended at once by a human mind, that doesn't mean we don't try. For example a googolplex is impossible to write down in the entire visible universe, yet we managed to come up with the idea of it anyway. And while the distance even to the nearest star is difficult, if not impossible for a human to imagine, I work in such distances as often a hobby. While no human can calculate the paths and interactions in their own mind of molecules within a rocket combustion chamber, we are still able to devise calculations to determine such performance and implement them in simulators for greater accuracy.

There is no suggestion of any "intelligence level beyond our own". You can be more intelligent than a human, sure- I'm not trying to refute that. My point is that you don't get magical "intelligence levels" at which point you start to operate on a whole different logic than everything else in the universe. The whole concept makes no sense, but it is of course a nice idea for sci-fi writers... and the Superintelligent AI God Cult.

Also, not once do you answer my question of why it is always assumed that a generalist AI would be an 'artificial person'. I presume this assumption is so heavily ingrained culturally or ideologically, that people don't question it... but it also makes little sense. The whole 'create an artificial person' mindset is just limited and limiting.

I find it pretty ironic that people complain about people assuming that the future will be like Star Trek, for example... only to replace that assumption, with their own shaky assumption.

In 50 to 100 years as computers get faster and hardware integration becomes smaller and when we reach the quantum level and have the ability to take a snapshot of the human body ,store it, then shoot that stored info to be reassembled at another location ,we will then be able to "Beam" ourselves from one place to another. This is already being done with particles now. Will teleportation lead to quantum computing or will quantum computing lead to teleportation? It should be fun to watch evolve either way.

Er... yeah. Aside from the... continuity issues (read: murder) with teleportation, it isn't that simple. That is a lot of data (I'd imagine even for a quantum computer), the structure is constantly changing, and scanning it and reconstructing it is not easy (unless you have the futurist staple of magic, of course).

The sort of quantum effects that are at play on the pico-scale don't apply in the macro-scale, as far as I know. Or at least, they are not at play in the same way.

Some of the problems of quantum computing have already been described here. Quantum computing likely isn't anything that could be easily achieved, either...

 

[RS-232]

Er... yeah. Aside from the... continuity issues (read: murder) with teleportation, it isn't that simple. That is a lot of data (I'd imagine even for a quantum computer), the structure is constantly changing, and scanning it and reconstructing it is not easy (unless you have the futurist staple of magic, of course).

The sort of quantum effects that are at play on the pico-scale don't apply in the macro-scale, as far as I know. Or at least, they are not at play in the same way.

Some of the problems of quantum computing have already been described here. Quantum computing likely isn't anything that could be easily achieved, either...

They are already achieving alot. It is well on it's way.
(The last paragraph I found very intresting)
http://www.nytimes.com/2010/11/09/science/09compute.html

 

[T.Neo]

They are already achieving alot.

Really? Where? Didn't someone in this thread say that programming for a quantum computer is purely hypothetical at this point?

It is well on it's way.

Interesting experiments do not correlate with a usable technology on the verge of becoming widespread. There's a big difference.

And a quantum computer doesn't suddenly translate into a Star Trek transporter, either. Obviously.

 

[FADEC]

What will computers be like?

I think that computers will rule everything. More than today. Even power production and distribution will be fully computer controlled as demanded by households, factories etc. Just everything will be digitally controlled. Real money even might disappear. And the power of computers will be beyond our current imagination I guess. Certainly also the way hardware works today.

Just watch a 1960s James Bond movie. What appeared to be science fiction back then already is outdated. Even the "small" spionage cameras back then. A modern mobile phone can do way more ever expected by the Bond makers in the old days.

And look at the giant leap from the 1980s until today. Although it's not always a true gain...

:lol:

---------- Post added at 07:30 PM ---------- Previous post was at 07:21 PM ----------

By the way, while I'm reading "16 bit" on the Mac in the video above: too bad that Commodore crashed in the early 1990s. They already offered a 32 bit multi tasking OS in the mid 1980s. MS presented 32 bit as an innovation in the mid 1990s :lol: Seriously: if Commodore would not have failed and crashed due to the bad management in the early 1990s, the computer marked would look different today. Those computers had a "soul". Something a usual PC never had..., less than ever M$ Windows.