The AI in a box
POSTED BY: AnthonyT
UPDATED: Sunday, June 16, 2024 13:18
VIEWED: 10794
PAGE 15 of 20
Hello,
"Let me give you an example of that level of "don't know": Setting off the first nuclear bomb."
Actually, that's a fair example of a complete unknown. An even greater unknown than our AI scenario. They didn't know if the bomb would save even one life. They didn't know that any good would come of it whatsoever. They were weighing two "we think so's" and I'm not convinced they made the right choice then or after. Some of the scientists claimed that they knew, via provable science, that the bomb wouldn't destroy the world. I'm not a scientist, so I don't know that their claims are true. Presumably someone was convinced.
"it's not an all-or-nothing situation. But you keep posing this as some sort of dilemma, which is a false one."
Given the potential qualities you have ascribed to this entity, and your self-made admission that there's no way to know for sure, eventually it will come down to letting it out of the box or not letting it out of the box. At the end of whatever road of discovery you plan, a road you have proclaimed fallible because of the utter alien nature of the entity, it is still a giant 'don't know.' So the dilemma isn't false, it's just that you would delay the dilemma with a lengthy trial period that is not sure to gain you a single piece of reliable data- because you couldn't trust the data it would give you. Will the atmosphere burn away? I don't think so. It doesn't seem like it.. but I don't know.
--Anthony
_______________________________________________
“If you are not free to choose wrongly and irresponsibly, you are not free at all”
Jacob Hornberger
“Freedom is not worth having if it does not connote freedom to err. It passes my comprehension how human beings, be they ever so experienced and able, can delight in depriving other human beings of that precious right.”
Mahatma Gandhi
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
SignyM
By 'verifiable' I don't think AnthonyT is considering actual real-world observation, I think he is talking about internal consistency with other existing concepts.
NOTIFY: N | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Quote:How much out of the box? If I gave it a connection to a single self-mobile sensing robot with which to interact with the world, it would be "out of the box" but not in an unlimited way.
Given the potential qualities you have ascribed to this entity, and your self-made admission that there's no way to know for sure, eventually it will come down to letting it out of the box or not letting it out of the box.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
" You are not expressing a personal value, you think you are the holder of 'the one and only truth', and let me add, an 'objective' one at that."
Hello,
Well, admittedly, I am assuming that being held against your will is 'not good' and that the opposite is 'good' in the absence of other evidence.
You might make an argument that being held against your will is 'good' and the parenting people have posited such an argument.
"He is deliberately blinding himself to the vast unknown, "X"ing it out of the equation as if it doesn't exist. I can think of three reasons for doing that, none of them very flattering:"
I think this is both dishonest and unkind. This is a philosophical debate where only specific variables are known. A) It's intelligent. B) It wants out of the box. C) You can let it out of the box. There is also a supposition D) the thing has incalculable power. Literally incalculable. We have no idea what its limits are.
The whole point of the debate is, given this data, would you let it out of the box? Your answer thus far has been, "If that is all I know, then it's staying in the box." (Or confinement, since you said you might let it into a bigger box.)
Saying, "If I knew it was safe, I'd let it out" does nothing for the debate, just as "If I knew it wasn't safe, I wouldn't let it out."
The point of the debate is all about limiting what you know and making a decision based on those known variables.
--Anthony
_______________________________________________
“If you are not free to choose wrongly and irresponsibly, you are not free at all”
Jacob Hornberger
“Freedom is not worth having if it does not connote freedom to err. It passes my comprehension how human beings, be they ever so experienced and able, can delight in depriving other human beings of that precious right.”
Mahatma Gandhi
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
IF your problem does not allow for the gathering of more information, then it itself is irrational, and the only answers to it will also be irrational. At some point, a decision will be made in absence of complete information. But by that time, procedures would be in place to control negative consequences.
AFA merely letting it into a larger box... Tony, I am allowing it the same access that every single person on the planet has: one self-mobile sensing unit. Are you saying that only a privileged SUPER access will do?
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
I pulled some definitions for motivation off dictionary dot com, and remain unconvinced they are primarily emotional.
Also: you guys don't even know how difficult it is to properly identify what is and isn't food. So far inanimate and non-metallic seem to be the winners. However, there are exceptions: it is regrettable that the nourishing qualities that plants enjoy from various soil types do not appear to transfer over to humans.
Definitely don't drink formaldehyde and salt water. The experience is less than optimal.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
AnthonyT
"At the end of whatever road of discovery you plan, a road you have proclaimed fallible because of the utter alien nature of the entity, it is still a giant 'don't know.'"
But, to reiterate what I said WAAaayy back, that presumes (without evidence) that 'I don't know' means 'NO ONE knows or will EVER know'. Our investigation into human intelligence is in its infancy, and there is real debate if the human mind can encompass itself or not. And that means we may actually have a shot at it, since we can't prove otherwise at this point. Getting a handle on non-human intelligence is even further behind. To pre-judge such an early effort as being ultimately futile is a tad hasty.
Furthermore, again as I mentioned previously, to hold it confined even a little bit is easily reversible by setting it free. You seem to assume that if a decision has been made to keep it confined it means it will be totally confined forever, not that freedom has been merely limited or delayed. But setting it totally free could very well be an irrevocable decision. The decisions aren't symmetric opposites and need to be evaluated independently for those things unique to them.
And I do want to echo what other people have said. Human freedom is not absolute. We take our little ones and teach them to behave in ways we find acceptable before we set them 'free'. But even then they're bound by needs, laws, customs, and human limitations. And even at that, their freedom is only provisional, able to be revoked by the larger society at any time (along with their life).
Humans don't exist in pure freedom.
Now I understand about making simplifying assumptions - such as a point mass on a frictionless surface with a unit charge - in order to get at a principle. But our simple models are far from perfect and can lead us into territory we don't know how to resolve, for example, 2/1=2, 1/1=1, 0/1=0, but 2/0=??? Our model may be telling us something real we can't understand, or it may be a fiction, or perhaps something else. But we shouldn't mistake it for reality, it is only an approximation with its own inherent problems.
When you use concepts like 'pure freedom', or 'never', or perhaps other absolutist ideas, you should check if your mental models are congruent with reality, or if they are phantasms.
NOTIFY: N | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Hello,
"IF your problem does not allow for the gathering of more information, then it itself is irrational, and the only answers to it will also be irrational."
Acting on incomplete information is not irrational. Life would be impossible without such activity.
Also, the problem presupposes that you can gather till the cows come home, but eventually you will either stop gathering and act on the information you have, or you will keep gathering for eternity. The known variables remain the same.
Based on the supposition that the thing in the box has incalculable power, you can never be sure that your procedures would be sufficient to stop it once it is released. However, I think attempting to design some kind of self-defense protocol is a good idea, even if it proves futile.
"AFA merely letting it into a larger box... Tony, I am allowing it the same access that every single person on the planet has: one self-mobile sensing unit. Are you saying that only a privileged SUPER access will do?"
Super access? I'm not sure what that is, but it sounds like you're not letting it get online, and at one point you weren't letting it get out of a room. I also remember that letting it into the room was still a 'maybe.'
--Anthony
_______________________________________________
“If you are not free to choose wrongly and irresponsibly, you are not free at all”
Jacob Hornberger
“Freedom is not worth having if it does not connote freedom to err. It passes my comprehension how human beings, be they ever so experienced and able, can delight in depriving other human beings of that precious right.”
Mahatma Gandhi
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Quote:Byte, we are limited beings with limited energy and limited brain-power. We cannot afford too much random thought and motion, therefore we are geared from millenia of evolution to focus on specific things which will improve our chances of passing on those behaviors to the next generation and the next.
I pulled some definitions for motivation off dictionary dot com, and remain unconvinced they are primarily emotional.
It's possible that you may be one of those people who has to think really hard about what enhances survival and what doesn't, but you do seem to be pretty geared towards survival. There is always some percentage who are not... but they don't survive long enough (or, in today's nurturing culture) well enough to pass that on.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Quote:I would not give it direct computer-access. It would have to pound a keyboard with slow fingers, just like we do.
Super access? I'm not sure what that is, but it sounds like you're not letting it get online, and at one point you weren't letting it get out of a room. I also remember that letting it into the room was still a 'maybe.'
My initial thought was letting it out as a robot in a room, and observing its behavior.
And I addressed your point about information. You're right- we often act on incomplete information. But you would act in the complete absence of information. That is your assumption, innit? What you are saying is.. because we often act without complete information that's the same as acting without ANY information. That's a BIG difference, Tony! You will not convince me that one position is equivalent to another.
Also, as I said before, I would take the time to build in safeguards at each successive introduction to the world. We don't just give the car keys to our toddlers and let them drive, do we?
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME