Computers in 30 years

[Suuper]

On the internet the digital trail you leave is probably already being read- and your personality/future marketing likelihoods- reconstructed from it.

The increasing ability to track people and obtain information like that also disturbs me.

[Chai]

Hopefully none of this lagging business.

[Cirmentia]

Hmm... That's pretty hard. Think about how much more advanced technology has increased in the past year alone, let alone five. In thirty years... I'll be blown away.

[御坂 美琴]

Hopefully none of this lagging business.

lagging will always happen.. software updates along with hardware

[Lonic]

lagging will always happen.. software updates along with hardware

...Unless, people have found a way to prevent it, 30 years in the future.

[御坂 美琴]

...Unless, people have found a way to prevent it, 30 years in the future.

`-`

highly unlikely

[PooZy]

...Unless, people have found a way to prevent it, 30 years in the future.

You can always create lag.
Even on a super powerful computer that can run minecraft with no lag, just keep on opening up new minecraft tabs and eventually the lag will come back.

[aaaaaa123456789]

Heh, I guess there's still some ideas brought by science-fiction.

I know computers will increase in power. Based on Moore's Law, the computing power increases every three years by a factor of 4, reducing costs to a quarter. Since 4^10 = 1,048,576, we would expect computers to increase a million times in power in 30 years. That's pretty much what happened in the past thirty years -- the increment will probably be much larger in the upcoming 30.

But there's still the problem of what computers cannot do. I think I said this before, that human behavior is infinite and uncountable, whilst computer states are countable and finite. Computers behave according to automata theory; a computer is, in essence, a finite deterministic automaton. There's a finite amount of states, and every state determines the next one. The finiteness of states is easily provable: there are only so many bits in a computer. Add up everything that holds information (memory, hard disks, even the data lines of the main board), and you'll have a certain amount of bits. Since every bit takes up physical space (tiny amounts of it, but it does take physical space), and the computer is physically constrained to a certain amount of space (i.e., it's not infinitely large -- at the very least, it's smaller than the Earth itself), the finiteness of states is proven: let T be the physical size of the computer, and h the minimum physical size of a bit; since every bit has only two discrete states, the total amount of states, call it s, is bound by s < 2^T/h; thus, it's finite.

Human behavior is infinite, as we all know. It's not hard to realize that humans have the potential to do an infinite and uncountable number of things. Therefore, a computer can't, and will never be able to, fully emulate human behavior. Computers are finite, and will always be; humans are not.

Maybe quantum computers (which have an infinite number of states) could somehow emulate human behavior, but I doubt it.

But as far as our common computers go, no matter how much they develop, they'll never have an AI capable of acting like a human being. They will never have something naturally infinite and incredibly complex as emotion. And they will never be able to read human thoughts with 100% accuracy.


What they can, and probably will do, is become more accessible. There's ways of determining human speech with decent accuracy (just so you know, not even humans can detect speech with 100% accuracy), so that's a first. Same comes from gestures, faces, and so on. Eventually, I do believe that computers will be able to (with lots of limitations) detect human behavior, and somehow read it. Of course, every reading a computer does of human behavior is a simplified model -- turning the uncountable infinity into a limited number of values. But still, it could be accurate enough for most purposes.
Long things short, I wouldn't be surprised if computers became controlled by common human interactions (like gestures or speech), without the need of any explicit interaction with peripherals (like typing).

Oh, and as for the 3D holograms, the technology is already being investigated, and I believe there's been a few successful prototypes. The only question is, when it will become cheap enough for a home user to have. That's something I expect to happen in much less than 30 years.

[nopedk]

You can always create lag.
Even on a super powerful computer that can run minecraft with no lag, just keep on opening up new minecraft tabs and eventually the lag will come back.

unless there's infinite ram

[[1/2]Lite]

New computers will be have IE 15 and be running Windows 49.

[Suuper]

But there's still the problem of what computers cannot do. I think I said this before, that human behavior is infinite and uncountable, whilst computer states are countable and finite.

Human behavior is infinite, as we all know. It's not hard to realize that humans have the potential to do an infinite and uncountable number of things. Therefore, a computer can't, and will never be able to, fully emulate human behavior. Computers are finite, and will always be; humans are not.

Maybe quantum computers (which have an infinite number of states) could somehow emulate human behavior, but I doubt it.

But as far as our common computers go, no matter how much they develop, they'll never have an AI capable of acting like a human being. They will never have something naturally infinite and incredibly complex as emotion. And they will never be able to read human thoughts with 100% accuracy.

Humans aren't infinite...we have quite limited abilities. We can do uncountable things, but not infinite.

Computers, I think, WILL be able to fully emulate human behavior at some time in the future. As computers become more advanced, they are getting closer and closer all the time. You can already have conversations with AI's. (Perhaps not good ones. Cleverbot isn't that smart, but AI's for supercomputers are far more advanced.)
Computers can be programed to learn. Maybe humans will never write an AI with human-like intelligence, but a computer might.

New computers will be have IE 15 and be running Windows 49.

Will Microsoft still be in business in 30 years? If so, they better have some good software by 30 years. (Windows 7 and the software that comes with it fail.)

[Mr Seven Up]

We'll be able to play games like Battlefield 3 on our cell phones

[aaaaaa123456789]

Humans aren't infinite...we have quite limited abilities. We can do uncountable things, but not infinite.

Anything uncountable is infinite, that's an easily provable mathematical fact. (Note: "uncountable" doesn't mean it takes too long to count. It means it cannot be counted at all.)

And humans have the potential to do infinite things, if you think of actions elementally. For starters, there's an infinite number of different things you could say at any given point. And that's just referring to speech.

[CPU]

I believe there will be computers this advanced in the near future but I doubt they'd be affordable. Only large corporations would be able to afford them (most likely). I personally would think that'd be cool if I were able to lock my laptop via eye laser verification.

[Suuper]

Anything uncountable is infinite, that's an easily provable mathematical fact. (Note: "uncountable" doesn't mean it takes too long to count. It means it cannot be counted at all.)

And humans have the potential to do infinite things, if you think of actions elementally. For starters, there's an infinite number of different things you could say at any given point. And that's just referring to speech.

Computers can produce more sounds than humans can.
If humans can say infinite things, computers can say even more.

Are there an infinite number of possible sounds?

[aaaaaa123456789]

Computers can produce more sounds than humans can.
If humans can say infinite things, computers can say even more.

Are there an infinite number of possible sounds?

Nothing a computer can produce is infinite. And yeah, there are infinite sounds.

But a computer can only produce a finite set of them.

[Suuper]

Nothing a computer can produce is infinite. And yeah, there are infinite sounds.

But a computer can only produce a finite set of them.

And a human can only produce a part of what a computer can.

[Wolframium]

Heh, I guess there's still some ideas brought by science-fiction.

I know computers will increase in power. Based on Moore's Law, the computing power increases every three years by a factor of 4, reducing costs to a quarter. Since 4^10 = 1,048,576, we would expect computers to increase a million times in power in 30 years. That's pretty much what happened in the past thirty years -- the increment will probably be much larger in the upcoming 30.

But there's still the problem of what computers cannot do. I think I said this before, that human behavior is infinite and uncountable, whilst computer states are countable and finite. Computers behave according to automata theory; a computer is, in essence, a finite deterministic automaton. There's a finite amount of states, and every state determines the next one. The finiteness of states is easily provable: there are only so many bits in a computer. Add up everything that holds information (memory, hard disks, even the data lines of the main board), and you'll have a certain amount of bits. Since every bit takes up physical space (tiny amounts of it, but it does take physical space), and the computer is physically constrained to a certain amount of space (i.e., it's not infinitely large -- at the very least, it's smaller than the Earth itself), the finiteness of states is proven: let T be the physical size of the computer, and h the minimum physical size of a bit; since every bit has only two discrete states, the total amount of states, call it s, is bound by s < 2^T/h; thus, it's finite.

Human behavior is infinite, as we all know. It's not hard to realize that humans have the potential to do an infinite and uncountable number of things. Therefore, a computer can't, and will never be able to, fully emulate human behavior. Computers are finite, and will always be; humans are not.

Maybe quantum computers (which have an infinite number of states) could somehow emulate human behavior, but I doubt it.

But as far as our common computers go, no matter how much they develop, they'll never have an AI capable of acting like a human being. They will never have something naturally infinite and incredibly complex as emotion. And they will never be able to read human thoughts with 100% accuracy.


What they can, and probably will do, is become more accessible. There's ways of determining human speech with decent accuracy (just so you know, not even humans can detect speech with 100% accuracy), so that's a first. Same comes from gestures, faces, and so on. Eventually, I do believe that computers will be able to (with lots of limitations) detect human behavior, and somehow read it. Of course, every reading a computer does of human behavior is a simplified model -- turning the uncountable infinity into a limited number of values. But still, it could be accurate enough for most purposes.
Long things short, I wouldn't be surprised if computers became controlled by common human interactions (like gestures or speech), without the need of any explicit interaction with peripherals (like typing).

Oh, and as for the 3D holograms, the technology is already being investigated, and I believe there's been a few successful prototypes. The only question is, when it will become cheap enough for a home user to have. That's something I expect to happen in much less than 30 years.

I disagree, in general, about what binary computers can and cannot do. I believe that is it fully possible for a binary computer to accurately model the human brain. When you say that humans have an infinite amount of actions and computers do not, you are pretty much wrong. At this point, a computer pretty much has an infinite amount of states to the human brain. If you take a modern computer and you look at its RAM, hard disk, and the state of its CPU, you will pretty much never see two times where those are the same without purposefully trying to achieve that goal. Second, while the brain is complex, and computationally expensive to model, it isn't impossible. According to the best models I've seen, it will be between 10 and 25 years before we have enough computing power to make a perfect computer model of the human brain. Of course, it will probably take another 10-20 years after that to effectively manage it, but I see no real reason why we can't. We can model the behavior of individual neurons, and we can model networks of them, and there isn't really much more to the brain than that.

As for computers reading human behavior, I don't think that is the real problem must be overcome. We have had decent speech recognition and facial recognition and even gesture recognition for a while, but the key problem is figuring out what to do with the data that the computer is being given. It is much easier for a computer to figure out the text of what you are telling it than it is to actually figure out what it should do in response to your query. While things like wolfram alpha and siri are certainly pretty big evolutions in that field, they are still nowhere close to just being able to find the proper answer to a generic query. Further, the ability to do actions based on generic user input is still quite far away. Being able to tell a computer to simply "make a program that will take an image as input and turn it upside down" is quite far away.

As for 3D holograms, I don't think it will be here in 30 years. There are various approaches, but its one of those things that is perpetually 30 years away. Every few years they find a new way of doing it, and there are lots of news articles, but it never seems to scale.

[御坂 美琴]

this discussion about organic computers (brain) vs digital computers is quite interesting.. i know there is a set number out there for how much knowledge the brain can hold, but from what i know about the brain, iirc, its good at compressing info via hippocampus and also wiring its self to access things faster and easier.. i dont know how a non-organic computer would be able to mimic that, but they might just achieve or surpass the brain through other methods like quantum computing (assuming they ever get past the whole observer effect and quantum entanglement)

[Rammjet]

Heh, I guess there's still some ideas brought by science-fiction.

I know computers will increase in power. Based on Moore's Law, the computing power increases every three years by a factor of 4, reducing costs to a quarter. Since 4^10 = 1,048,576, we would expect computers to increase a million times in power in 30 years. That's pretty much what happened in the past thirty years -- the increment will probably be much larger in the upcoming 30.

But there's still the problem of what computers cannot do. I think I said this before, that human behavior is infinite and uncountable, whilst computer states are countable and finite. Computers behave according to automata theory; a computer is, in essence, a finite deterministic automaton. There's a finite amount of states, and every state determines the next one. The finiteness of states is easily provable: there are only so many bits in a computer. Add up everything that holds information (memory, hard disks, even the data lines of the main board), and you'll have a certain amount of bits. Since every bit takes up physical space (tiny amounts of it, but it does take physical space), and the computer is physically constrained to a certain amount of space (i.e., it's not infinitely large -- at the very least, it's smaller than the Earth itself), the finiteness of states is proven: let T be the physical size of the computer, and h the minimum physical size of a bit; since every bit has only two discrete states, the total amount of states, call it s, is bound by s < 2^T/h; thus, it's finite.

Human behavior is infinite, as we all know. It's not hard to realize that humans have the potential to do an infinite and uncountable number of things. Therefore, a computer can't, and will never be able to, fully emulate human behavior. Computers are finite, and will always be; humans are not.

Maybe quantum computers (which have an infinite number of states) could somehow emulate human behavior, but I doubt it.

But as far as our common computers go, no matter how much they develop, they'll never have an AI capable of acting like a human being. They will never have something naturally infinite and incredibly complex as emotion. And they will never be able to read human thoughts with 100% accuracy.


What they can, and probably will do, is become more accessible. There's ways of determining human speech with decent accuracy (just so you know, not even humans can detect speech with 100% accuracy), so that's a first. Same comes from gestures, faces, and so on. Eventually, I do believe that computers will be able to (with lots of limitations) detect human behavior, and somehow read it. Of course, every reading a computer does of human behavior is a simplified model -- turning the uncountable infinity into a limited number of values. But still, it could be accurate enough for most purposes.
Long things short, I wouldn't be surprised if computers became controlled by common human interactions (like gestures or speech), without the need of any explicit interaction with peripherals (like typing).

Oh, and as for the 3D holograms, the technology is already being investigated, and I believe there's been a few successful prototypes. The only question is, when it will become cheap enough for a home user to have. That's something I expect to happen in much less than 30 years.

Human behaviour is not infinite. The reason computers can't fully emulate human behaviour is the same reason you can never know what it's like to be a bat, because to know what it's like to be a bat you wouldn't ask that question...because you'd already be a bat. A computer wouldn't be able to pretend to be a human entirely because to emulate is to approximate anyway, and there are some logical paradoxes that can't be stepped around. Other than that a future computer might put on a convincing show.

Human emotion isn't infinite, most of the time we respond just like automatons with the 'big six' emotions, in a way which certainly doesn't show infinite complexity. The human brain has a limited capacity like any computer, its processes deterministic just like a computer, which is why it quite reliably throws up the same output if you chuck the same input in. 700nm wavelength light falling on your retina sends an input message that is determined by predictable electrochemical processes, and your brain produces a determined response, of reporting the colour red.
If you take the same human brain and subject it to the same conditions it will produce the same response, if you take a human brain and provide it with more input than its cognitive functions are capable of dealing with, it will slow down or fail just like a computer etcetera
complexity is not infinity.

I agree that this will not happen in the next 30 years, but I think you're making naive assumptions about what humans are.

Anything uncountable is infinite, that's an easily provable mathematical fact. (Note: "uncountable" doesn't mean it takes too long to count. It means it cannot be counted at all.)

And humans have the potential to do infinite things, if you think of actions elementally. For starters, there's an infinite number of different things you could say at any given point. And that's just referring to speech.

No, this is an argument from ignorance. It's currently beyond our capability to model human brains, but is that because they're infinitely complex? No, it's because their complexity is currently beyond our understanding.

Humans don't have the potential to say infinite things, because even given their entire lifetime the human mouth, a mechanical instrument, has a finite number of possible contortions with which it can produce sounds. Humans can't do infinite things, even if you said 'I can stand 1mm away from the desk, or 'nmm and n/2 mm' and so on you'd soon find your model was inappropriate because you'd get down to trying to claim humans could willfully move between half a quantum state. etc

Nothing a computer can produce is infinite. And yeah, there are infinite sounds.

But a computer can only produce a finite set of them.

Are there really infinite sounds?

I think that's another assumption. Represent a 'sound' as a wave carried through a finite number of particles [even if we're using all the particles in the universe], the positions these particles can occupy as they transmit a wave function is not infinite, even if it's an extremely extremely high number.