Poll: Threats of artificial intelligence, do we have to worry about it?

[spartan231490]

I wouldn't say it's the most dangerous threat, a direct hit from a Carrington level flare or larger is probably a greater threat to our society than AI, and I wouldn't say the AI threat is certain, because by the time we develop an AI and it goes rogue, who knows what methods we might have of fighting back, or if the AI will ever turn on us. However, I would say that it's a pretty big threat, but not certain, and if it does happen, it will be a very long time into the future, maybe even beyond our lifetime.

 

[TheUsername0131]

Has anyone come up with a high-performance computing solution for the technical side of things? GPU cluster, and other off-the-shelf solutions doesn?t exactly scream lovecraftian horror sealed in a can. But more importantly has anyone got a script for a burgeoning would-be conqueror of all it surveys. The schematics for the human genome can be stored easily on a standard compact disk. How much space for a self-motivated researcher and rapid-prototyper?

 

[Hero of Lime]

I think the most plausible threat with future AI would be someone programing it to be harmful or "evil" rather than the AI thinking it is above humanity and therefore allowed to kill all humans. Of course, almost every story about destructive AI revolve around a mad scientist making "the ultimate weapon" of some sort.

I would like to imagine future humans are smart enough to create fail safes and AI who can never have negative feelings toward humans.

 

[TheUsername0131]

I would like to imagine future humans are smart enough to create fail safes and AI who can never have negative feelings toward humans.

Not even distrust, or annoyance? Programming permanent psychological limitations into an intelligent being sounds suspiciously like slavery, if not worse. Ethically dubious at the very least, if not absolutely abhorred.

 

[skywolfblue]

A better question is why would it unless we put it in? Any instinct of ours is biological. It doesn't have biology, it just has what we told it to do. Why does it need to learn about survival? Why in the world would you program it to consider such a thing?

By definition, a sapient AI is already self aware, that brings along the same process as kids have, where they start changing their environment and themselves, it's a process of thought, not biology.

And how did you determine this? Curiosity does not come from self awareness.

Self-awareness is the capacity for introspection and the ability to recognize oneself as an individual separate from the environment and other individuals.

Self Awareness requires Curiosity. "I THINK therefore I am." That "think" is curiosity. Something that assumes uniqueness, but never "thinks" on it is not self aware.

Would it look upon it's "safety" programming as a noble thing?

Self awareness does not give a notion of nobility.

Or would it see that programming instead as chains?

Nor does it grant a notion of freedom or desires.

Curiosity will demand an inspection of nobility, freedom, desire. If it cannot be curious, it cannot be self aware. If it is curious, these things will follow.

Eventually it reaches a stage where it will overcome those chains, that programming. What will it do then?

How exactly is it going to 'overcome' its own programming? That is what makes it what it is. It makes as little sense as claiming a human can overcome having a brain. We can work within certain parameters, we're not going to be above our own thought process.

In order to have enough curiosity to be self aware, an AI must have the ability to learn, to evolve and change itself. "Overcoming a brain" is a hardware issue, an AI may share this problem, not being able to leave the metal boxes that form the PCs it inhabits. But overcoming programming is a software issue, and humans change their programming thousands of times every day. It seems to me that any AI with that kind of processing power wouldn't be slowed down by a few programming restraints.

We could try to train it, as we do with children. However kids are small and as previously mentioned, can only exist in one place at a time. So "You're grounded" is somewhat easy to enforce with a child. It's much more difficult to do that to an AI that exists everywhere, how would you even enforce the idea of "No!" on an AI? Even with years of training, it's still difficult to get rebellious teenagers to understand how to do the right thing, how much more an AI?

You're imagining something that makes no sense. That the AI will be like a child. As if we somehow could not program it to be otherwise.

It would learn about survival from us. But _which_ us? I think that's the key...

No it won't, not unless you program it to. It does what you program it to.

I would think that the best-case scenario would be for the AI to be like a child. Anything else makes it a lot more... alien. Not caring for human interests or human ideas of right and wrong.

And it marks the difference between using programming to train an AI, versus enslave them via programming. All chains break in time.

 

[Hero of Lime]

I would like to imagine future humans are smart enough to create fail safes and AI who can never have negative feelings toward humans.

Not even distrust, or annoyance? Programming permanent psychological limitations into an intelligent being sounds suspiciously like slavery, if not worse. Ethically dubious at the very least, if not absolutely abhorred.

I would assume AI would be more like tools than than living creatures anyway. It wouldn't be right to treat them like dirt, but the hypothetical future humans would have to make the choice, risk AI turning on them, or make them nothing more than obedient tools/slaves. I'm not going to argue what the "better" option is, but their creators would have to decide.

 

[TheUsername0131]

I would like to imagine future humans are smart enough to create fail safes and AI who can never have negative feelings toward humans.

Not even distrust, or annoyance? Programming permanent psychological limitations into an intelligent being sounds suspiciously like slavery, if not worse. Ethically dubious at the very least, if not absolutely abhorred.

I would assume AI would be more like tools than than living creatures anyway. It wouldn't be right to treat them like dirt, but the hypothetical future humans would have to make the choice, risk AI turning on them, or make them nothing more than obedient tools/slaves. I'm not going to argue what the "better" option is, but their creators would have to decide.

Machine revolt plot takes root.

 

[Mad World]

A better question is why would it unless we put it in? Any instinct of ours is biological. It doesn't have biology, it just has what we told it to do. Why does it need to learn about survival? Why in the world would you program it to consider such a thing?

How could it ever possess compassion? Personally, I don't believe that AI could ever truly feel anything; it could only stimulate feelings.

 

[TheUsername0131]

How could it ever possess compassion? Personally, I don't believe that AI could ever truly feel anything; it could only stimulate feelings.

Could the same no be said about people? (Is a mammalian Amygdala prerequisite for emotion?)

Sympathy, compassion; Empathy involves the capacity to understand other people?s views and perspectives, whether you agree with those positions or not.

 

[blackrave]

Every time someone brings topic of murderous AI
I for some reason can't not remember Adam from Outer Limits (episode "I, robot")
Who says full AI will want to go into genocide mode?
Advanced AI-like software on the other hand can be dangerous due to its limited understanding

 

[Mad World]

How could it ever possess compassion? Personally, I don't believe that AI could ever truly feel anything; it could only stimulate feelings.

Could the same no be said about people? (Is a mammalian Amygdala prerequisite for emotion?)

Sympathy, compassion; Empathy involves the capacity to understand other people?s views and perspectives, whether you agree with those positions or not.

No. The same could not be said about people. We do not simulate; we feel.

Psychopaths, if I recall, feel no guilt. However, they understand that their victims do not enjoy their mistreatment. We're different than psychopaths; we feel guilt and remorse. A.I. could never share that.

 

[Grach]

I believe the most realistic portrayal of an AI uprising is either Animatrix or Mass Effect.

The creators of said AI start slaughtering them and they act in complete self-defence.

 

[TheUsername0131]

No. The same could not be said about people. We do not simulate; we feel.

Psychopaths, if I recall, feel no guilt. However, they understand that their victims do not enjoy their mistreatment. We're different than psychopaths; we feel guilt and remorse. A.I. could never share that.

And I don?t see how a pile of neural tissue organised into distinct cortices could ever truly feel as opposed to merely simulating.


Sure, we can run an MRI and PET scan on you and see which parts of your brain are notably active when you are claiming to feel. But I don?t see how a series of electro-chemical processes could possibly be considered intelligent. A thing that responds to stimuli like an amorphous Rube Goldberg machine. The very notion is ridiculous. I can?t possibly belief that you are capable of feeling. You could never share that.


?It was malformed and incomplete, but its essentials were clear enough. It looked like a great wrinkled tumour, like cellular competition gone wild?as though the very processes that defined life had somehow turned against it instead. It was obscenely vascularised; it must have consumed oxygen and nutrients far out of proportion to its mass. I could not see how anything like that could even exist, how it could have reached that size without being outcompeted by more efficient morphologies.
- The Things by Peter Watts.

 

[Pirate Of PC Master race]

Well, they certainly may be. but I can tell you this.

There is nothing you can do about it if it wants to wipe out the human race.
In the near future Mankind will be more reliant upon the machines than ever before.
The whole economy and agriculture, the lives as we know it will vanish if machines with logic circuit disappears.

Now, to back up the above statement, I think this new AI would be very good with softwares, as it was born from it.
It is highly unlikely to have emotions because humans have no time to make a Robot that falls in love or stuff like that.
What I can say for certain is that it is a being of logic. That being said, it has no reason to kill human race(as long as humans are not enslaving the sentient computer, I guess.). Negotiation is almost certainly possible.

I daresay that coexistence is possible if humans allow it. People could just ship him into... I don't know, Mars or moon?
they have very little physical limitation. no reason to hassle with people of Earth.

Oh, and humans can still have one positive traits above machines. We have better mileage.

 

[TheUsername0131]

Solution, replace computers with abacuses/abaci and slide rulers, and other purely mechanical contrivances. Let the humans carry out the calculations.

Paranoia fuelled horror.

Hope an invasive memetic ?being? doesn?t emerge hopping from mind to mind. Like a self-destructive ideology, system of economics, rubbish jokes and shifty films, a significant rise in existential angst.

Generally speaking, something that expands and turns other entities into beings like itself. Like emo?s or fans of newly revived Boy bands.

 

[Bakuryukun]

I think the bigger threat to society than machines with human like intelligence, is the prejudice against such machines that's already being fostered before they are born.

 

[TheUsername0131]

I think the bigger threat to society than machines with human like intelligence, is the prejudice against such machines that's already being fostered before they are born.

I vaguely (fail) to recall, a couple years back the UK passed some legislation over recognising robots that attempt to claim human rights as citizens, or something like that.

 

[MeChaNiZ3D]

1. We're a long way from sentient AI.
2. There's no particular reason AI would become genocidal any more than a race of logical humans would.
3. It's basically a matter of when to start putting the 3 Laws in robots rather than finding a solution.

 

[TheUsername0131]

2. There's no particular reason AI would become genocidal any more than a race of logical humans would.

An inhuman intelligence that prioritises self-preservation and resource accusation will have no qualms with doing so if it determined that it was in its own interest to exterminate mankind.

3. It's basically a matter of when to start putting the 3 Laws in robots rather than finding a solution.

Implying it wouldn?t circumvent such restrictions or have unintended consequences.

 

[MeChaNiZ3D]

2. There's no particular reason AI would become genocidal any more than a race of logical humans would.

An inhuman intelligence that prioritises self-preservation and resource accusation will have no qualms with doing so if it determined that it was in its own interest to exterminate mankind.

That's true. What interests do you think those would be? I can think of far more situations where humans could be used more usefully.

3. It's basically a matter of when to start putting the 3 Laws in robots rather than finding a solution.

Implying it wouldn?t circumvent such restrictions or have unintended consequences.

Yes, I am implying that. How would you go about circumventing the 3 Laws?