Poll: Will there ever be true AI?

[Jumplion]

I don't think so, these days good AI is just the enemy taking cover and shooting you perfectly while you can only take 3 hits on the foot and still die without ever knowing where the hell the shots came from.

People say "This game has awesome AI!" when the only way the AI isn't retarted is that they take cover and the only reason why the AI is challenging is because they can kill you with perfect aim in only 3 hits.

But eventually AIs would be very smart but I don't htink they will still be distinguishable as computers.

VR Porn, here we come.

 

[SteinFaust]

they will, but they shouldn't. think about it. AI enemies designed to outsmart and kill human players? that will screw the pooch once the navy starts using them to control automated artillery and gunships and all those new toys they're always going on about.

 

[DougNotDougie]

u have to think about that someone made an ai the cant make something that can build memory onto itself the human mind can hold more information then any computer so it obvious it wont exactley happen completely

 

[fulano]

Ahem, actually, it IS possible that one day, we will be able to design an AI that truly gains some form of sentience, mainly, by the time the processes of the brain are fully understood which will take looong time. Considering the fact that our brain is just a bunch of material stuff stuff anyway, it can be argued that consciousness is just the process of billions of neural nodes working parallel to each other, exchanging information pretty fast. Our "minds", can be said, are nothing but the result of a pretty neat circuit.

It would be surprising, in turn, if we never actually developed AI since, in theory, all we would be doing, really, would be creating a process similar to ours. Of course, it won't happen from one day to another, and we will have to go through phases of truly stupid AI, and I don't mean videogame stupid, but truly stupid, as in making a judgement which turned out to be BAD judgement, like kids...or bugs, for that matter.

 

[vede]

Ahem, actually, it IS possible that one day, we will be able to design an AI that truly gains some form of sentience, mainly, by the time the processes of the brain are fully understood which will take looong time. Considering the fact that our brain is just a bunch of material stuff stuff anyway, it can be argued that consciousness is just the process of billions of neural nodes working parallel to each other, exchanging information pretty fast. Our "minds", can be said, are nothing but the result of a pretty neat circuit.

It would be surprising, in turn, if we never actually developed AI since, in theory, all we would be doing, really, would be creating a process similar to ours. Of course, it won't happen from one day to another, and we will have to go through phases of truly stupid AI, and I don't mean videogame stupid, but truly stupid, as in making a judgement which turned out to be BAD judgement, like kids...or bugs, for that matter.

Wow, seriously, you just convinced me that computers would be able to form a sentient AI. I have never though of the brain like that. And I usually do think about stuff like that.

It sucks that I've already voted...

 

[NinjaDwarf]

Well, there's always the possibility that all of humanity is the bio-engineered construct of some vastly advanced alien race. I don't know why they'd bother though seeing as all they seem to do is probe people...

 

[John Galt]

I, Robot. (Going by the movie, the 3 Laws are inherently flawed.)

But going by the book, the 3 Laws made things quite awesome, even starting a robot religion akin to sun-worship.

 

[FranicalFrazical]

I think there is actually a kind of AI robot in development called COG or something similar and its learnt like a child, I believe its 7 years old now

 

[fat american]

I think we will all be dead before we even get advanced to make AI or VI. Even if we get far enough there will not be true AI.

We already have VI in the form of telemarketing and other crap. Call X-Box customer support. You'll get some guy named Mac or Max or something like that and he responds to voice commands and all that crap. He is, in essence, a VI because all a VI does is respond to queries asked and delves into it's data bases for answers. I read all the Codex entries I got because I thought they were interesting.

On the Sentient AI though. Last year there was a soccer tournament between robot dogs and they did it entirely by themselves. There was no human interevetion what so ever. Yes they were programmed to play only soccer but they were free to run about the field and be stupid as expected but they actually played soccer. If a goalie dog let a ball get past him he realized it and he tried to hide his face under his paws like a real dog does. So I think some sort of sentient AI isn't far off.

 

[Zemalac]

I think the only real answer is a shrug and a "maybe." Because while it may seem impossible, even frightening now, what will happen in the next three hundred years or so is totally beyond our comprehension.

On a related note, true AI in a video game would be scary. Think about it...all those guys you kill? They're thinking beings just like you now. Then what happens if you tell one of them that he's actually in a game that you're playing? Scary stuff.

 

[Treblaine]

Well, asking if something fantastical is possible in the future then the answer is usually YES as the future is a very long time. I'm quite sure that given enough time and effort we will figure out some ingenious way of breaking the light barrier and flitting across the universe in weeks but that is so far beyond current technology and knowledge that it is pointless pondering about it when we SHOULD be thinking about the NEXT STEP.

A far better question is; will we have a computer that can pass itself off as human in the our lifetime. There exist "sentience tests" where a computer is tested to see if it can pass itself off as a human such as having a long conversation about varying subjects, engage in several complex games. This test was invented in the 50's, I forgot the name and basically involves a human communicating to someone in another room through something simple like typing though the person he is talking or playing a game with could be a human or a computer.

The human testers can ask whatever they like to tell who is human and who is robot and in over 50 years the machine is always caught out. It's just like how in an FPS deathmatch you can eventually spot which players are bots and which are real players.

As to the next step in Artificial Intelligence, What is becoming very popular and is being driven by consumer demand is for a more accessible human interface. Ask any hotel manager and he'll tell you how important it is to have a helpful and intelligent person at the lobby desk to welcome guests and answer any question clearly and insightfully as well as suitably informing them.

This human interface is needed in the millions of technological gadgets that people get frustrated with as they don't "speak machine", this goes from automated helplines to parking meters to new phones and even ticket machines.

Realistically, the first Artificial Intelligence we will see will be superficial. I say 15 to 20 years from now we will go to a train station and you will not have to navigate a touch screen on a ticket machine but will talk to a talkative and helpful person on a 3D monitor who I will actually have a fluid conversation with as I ask about train times, services and after walking away I will have to remind myself that she may seem friendly and nice but she is not real. Even the most useless technophobe could use the service while they would get frustrated with a touch screen computer, especially about finding basic information.

 

[huntedannoyed]

Thats all we need, machines to become self aware. Anybody remember 2001 or Terminator? Hey, I replied without even bringing up the Matrix (Damn!)!

 

[sidhe3141]

Treblaine
There exist "sentience tests" where a computer is tested to see if it can pass itself off as a human such as having a long conversation about varying subjects, engage in several complex games. This test was invented in the 50's, I forgot the name and basically involves a human communicating to someone in another room through something simple like typing though the person he is talking or playing a game with could be a human or a computer.

It's called the Turing test, and it has some problems (like that it completely ignores Option C, and the famous Chinese Room argument).

Realistically, the first Artificial Intelligence we will see will be superficial. I say 15 to 20 years from now we will go to a train station and you will not have to navigate a touch screen on a ticket machine but will talk to a talkative and helpful person on a 3D monitor who I will actually have a fluid conversation with as I ask about train times, services and after walking away I will have to remind myself that she may seem friendly and nice but she is not real.

Whether or not that AI would be superficial kind of depends on exactly what is involved in making an AI capable of fluid conversation, because to have a fluid conversation, you need to understand what another person is likely to say, and doing that involves understanding how people in general think, and by creating a program that can figure that out in realtime, you would essentially be creating an artificial human that the AI uses to determine its responses.

 

[SilentHunter7]

It doesnt matter how much processing power a computer has, a computer can only do exactly what it's programmed to do. My first Programming teacher had a very effective way of teaching the very basics on how computers work:

Mr. B: "Okay, how would you go about programming a humanoid computer to make an omlette?"

Person: "Uhh...first you get a bunch of eggs, and beat them together."

Mr. B: "The way you phrased that, the computer would smash a bunch of egg cartons on the floor, and start smearing them together with its hands."

Person: "Oh, um, put them in a bowl"

Mr B.: "Now it'll just toss them in, shell and all."

Person: "Okay, then I tell it to break the eggs and dump the yokes into the bowl."

Mr. B.: "Better, except it doesn't know how to break the eggs, or when to stop putting eggs into the bowl, and still has no idea what to do with the shells."

Person: *grumble*

Computers only do what they're programmed to do, regardless of power. They can't think on their own. You write software that makes a robot clean your kitchen floor, that's all it'll ever do, unless someone changes the software. Unless someone can write a program that can somehow modify it's own source code without potentially killing the program, then software, and by extension, AI, will never be able to 'naturally' evolve. And even then, it took the Earth over 4 billion years to make a creature that looked up at the night sky and wondered what its purpose was. So even with software evolution, a program that will question its own existence is very far off.

 

[Gamer137]

I dont think so because you have to program whatever you want them to do. When they make choices, its based on pre-programmed human choices. Computer are capable of learning information by themselves(like those little robot dog toys or whatever), but they have to be told how to use it. Electronic circuts do not work like brain chemicals.

 

[Copter400]

I, Robot. (Going by the movie, the 3 Laws are inherently flawed.)

But going by the book, the 3 Laws made things quite awesome, even starting a robot religion akin to sun-worship.

Hmm. Must read the book sometime.

 

[Erana]

Well, we're already doing better than the vast majority of organisms on the planet. Just try to get a fern to shoot at enemy soldiers.

 

[Dommyboy]

If a computer ever turned against me, I would just delete C/system32. Hah! Take that society!

 

[Sejs Cube]

"Ever" is a dangerous term.

Eventually, someday? Yes, absolutely there will be true self-aware AI.

It may just take a very long time before that happens.

 

[Treblaine]

Following SilentHunter7's omelette-bot example, he illustrates how hard it is to "tell" a robot what to do as this robot doesn't KNOW what an egg is and cannot see, feel, hear or smell/taste, only be activated where it can follow a rigid routine but any deviation in egg size would throw it off and it could not recognise that an egg was rotten or already cracked as some eggs in the box. Imagine a deaf-blind-anosmic man with no sensation in his hands is asked to fix a ramjet engine though he has never even heard of one, he possesses an incredibly intelligent brain yet it is useless as he can't interpret what he is doing and he has never been taught (programmed) what a ramjet engine is or how it works.

For this omelette-bot to make a decent breakfast it must be able to sense to world through sight or touch. It needs confirmation that it is doing it right. It must also know what the materiel properties of an egg is so it knows how to hold it and crack it into a bowl even if it is larger or smaller than expected. Now the tricky part, every egg is slightly different and no hob can cook an omelette exactly the same every time so the robot will have to judge when to flip the omelette and when to serve. Most people go by the consistency and colour of the egg while a great chef may even use the smell of the food and subtle subconscious clues to know when to serve it.

I think the key to artificial intelligence is the machine 'knowing' what it is doing and knowing whether it did it right, whereas a simple machine has little input or knowledge of what it is doing.

The only machines I can think of that know what they are doing are multiplayer bots that can navigate a large map, shooting opponents and dodging fire while picking up ammunition. Whereas a sentry gun is simple, it senses the enemy and will shoot it with no idea what the environments is like, a smart sentry gun would keep its gun pointed at the one entrance rather than mindlessly scanning.