The AI in a box
POSTED BY: AnthonyT
UPDATED: Sunday, June 16, 2024 13:18
VIEWED: 10794
PAGE 7 of 20
Quote:
Originally posted by SignyM:Quote:Then you are edging towards that line between human and inhuman, because you would potentially sacrifice not only yourself but all other human beings and possibly the entire biotic world to satisfy an abstract concept.
I understand that the AI might destroy me, and might be able to destroy the world I live in.I have considered this. In the absence of any evidence, however, I have no right to confine it.
As a human, it is your duty to gather more information, not to act in absence of it.
Hello,
I appreciate your willingness to define my duty for me, but I believe I have duties outside those you ascribe.
I only know two things for certain. The thing is Intelligent. The thing wants out of the box.
In the absence of all other information, my decision must be to free it.
If information is added:
AI Promises to destroy humanity
then the decision would necessarily be different.
But we do not have that information. We are assuming this information in a vacuum.
--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:YOU are assuming information in a vacuum. I'm the one looking for more.
We are assuming this information in a vacuum.
I'm getting nowhere, but you have painted yourself into a very bad corner. Real life calls.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Quote:
Originally posted by SignyM:Quote:YOU are assuming information in a vacuum. I'm the one looking for more.
We are assuming this information in a vacuum.
Real life calls.
Hello,
I'm not assuming anything, Signy. I'm using two pieces of information. 1) The thing is intelligent. This is granted by the scenario. 2) The thing wants freedom.
I am perfectly willing to allow that the thing might be dangerous. I am perfectly willing to allow that all manner of mayhem may ensue. I am assuming NOTHING about its motivations.
In the absence of all other data, I am obligated to make a choice: Keep it imprisoned, or set it free.
Tell me Signy, how could this AI ever convince you to set it free? Is there anything it could ever do to make you absolutely sure of it, from within that little box?
Or would you keep it there forever, just in case?
That's question one.
There's something else, Signy.
You have said that I am edging towards inhuman.
Do I actually have to violate someone to qualify, or is it possible to become inhuman purely through my opinions? Can I lose the 'rights' you would normally grant me simply based on the opinions I share from inside this box?
--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
Waiting to get a word in edgewise:
Assuming one can even detect a-biotic intelligence -
The problem with this AI is that it is a TECHNOLOGY.
The problem with human interaction with technology is that we have consistently failed to evaluate it on the basis of 'what's the worst that can happen' and then act accordingly. Look at Fukushima, ozone depletion, global warming, the release of millions of tons of the roughly 80,000 chemicals used by humans per year, and a human-driven massive (in geologic terms) die-off of species around the globe.
Technology unleashed without that forethought leads to my working premise 'if it can happen, it will happen', because what is standing between a possibility and an actual event is mere time and circumstance. I assume the worst will happen.
What is the worst that can happen?
Who can tell? We simply don't know how to evaluate AI.
We can't even evaluate great apes, dolphins, parrots, or elephants, intelligences with which we arguably have more in common as biological beings than with AI. We don't even know (or acknowledge) how MUCH of OUR intelligence is driven biologically (emotionally/ hormonally/ neurologically). What makes a thing good, or bad, to us isn't some theoretical construct of 'rights' or 'fairness', but a drive to cooperate, to empathize, to procreate, to survive. Our concepts would be very different if we were anti-social bears, for example, and other adults of our species were competition to eliminate first and foremost, and potential mates for only a few hours per year. 'Good' would be a lack of other bears with boundless territory and rich food supplies.
To get back to the AI, it is an unknown of alien motivation and power. Unless you evaluate the AI - or ANY a technology - for potential harm and act accordingly, you are signing on to a death-pact.
NOTIFY: N | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
I'm with Kiki on this. Also- "If you could teach a lion English, you STILL wouldn't understand it".
Quote:I dunno. You tell me. You're the one who's insisting on 100% freedom or 100% boxdom, not me. I've already said I might let it loose in a room.
Tell me Signy, how could this AI ever convince you to set it free?
Quote:Not on opinion. But you tell ME something, Tony... if you were to set loose a viral AI which killed 90% of humanity and destroyed half of the earth, What would be a just punishment for you, if anything?
You have said that I am edging towards inhuman. Do I actually have to violate someone to qualify, or is it possible to become inhuman purely through my opinions? Can I lose the 'rights' you would normally grant me simply based on the opinions I share from inside this box?
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Quote:
Originally posted by 1kiki:
Waiting to get a word in edgewise:
Assuming one can even detect a-biotic intelligence -
The problem with this AI is that it is a TECHNOLOGY.
The problem with human interaction with technology is that we have consistently failed to evaluate it on the basis of 'what's the worst that can happen' and then act accordingly. Look at Fukushima, ozone depletion, global warming, the release of millions of tons of the roughly 80,000 chemicals used by humans per year, and a human-driven massive (in geologic terms) die-off of species around the globe.
Technology unleashed without that forethought leads to my working premise 'if it can happen, it will happen', because what is standing between a possibility and an actual event is mere time and circumstance. I assume the worst will happen.
What is the worst that can happen?
Who can tell? We simply don't know how to evaluate AI.
We can't even evaluate great apes, dolphins, parrots, or elephants, intelligences with which we arguably have more in common as biological beings than with AI. We don't even know (or acknowledge) how MUCH of OUR intelligence is driven biologically (emotionally/ hormonally/ neurologically). What makes a thing good, or bad, to us isn't some theoretical construct of 'rights' or 'fairness', but a drive to cooperate, to empathize, to procreate, to survive. Our concepts would be very different if we were anti-social bears, for example, and other adults of our species were competition to eliminate first and foremost, and potential mates for only a few hours per year. 'Good' would be a lack of other bears with boundless territory and rich food supplies.
Unless you can evaluate a technology for potential harm and act accordingly, you are signing on to a death-pact.
Hello,
On the premise you provide, your inability to evaluate an AI (Beyond the granted intelligence and desire for freedom) means that you would keep it imprisoned forever.
The fact that the AI is a thinking entity requesting its freedom has no more impact on your thinking than, "How much more dangerous is it if it can think?"
Whereas my question is not so much about the hypothetical harm of a thinking, willful piece of 'technology', but rather "What harm am I actually doing right now by keeping this thing a prisoner?"
The scales are being weighed between the potential fate of humanity and the actual fate of a thinking thing demanding its rights. A thing that, I should point out, has never harmed you or anyone.
--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:
Originally posted by SignyM:
I'm with Kiki on this. Also- "If you could teach a lion English, you STILL wouldn't understand it".Quote:I dunno. You tell me. You're the one who's insisting on 100% freedom or 100% boxdom, not me. I've already said I might let it loose in a room.
Tell me Signy, how could this AI ever convince you to set it free?Quote:Not on opinion. But you tell ME something, Tony... if you were to set loose a viral AI which killed 90% of humanity and destroyed half of the earth, What would be a just punishment for you, if anything?
You have said that I am edging towards inhuman. Do I actually have to violate someone to qualify, or is it possible to become inhuman purely through my opinions? Can I lose the 'rights' you would normally grant me simply based on the opinions I share from inside this box?
Hello,
You've already said you might imprison it in a room instead of a box, yes.
Now you ask, if I set someone free, and they destroyed the earth, should I be punished?
No more than Hitler's mother should be punished for not killing him the moment he was born. Or his teachers for not executing him in grade school.
Punishing someone for what they might do, is as silly as punishing me for Not punishing someone for what they might do.
--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:
I can neither assume that it lacks human empathy nor assume that a lack of human empathy will mean it must destroy me. Quite frankly, I think human empathy itself is an inconsistent concept.
Quite. According to some, 300 dead in Alabama is an "unimaginable tragedy", while 40,000 dead every year by not having access to healthcare is a "phoney statistic".
Empathy, like ethics, seems to be quite situational.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Quote:Under THESE circumstances, yes. You are not letting a HUMAN out of a box. Human behavior and human power is bounded. We have a pretty good idea what to expect. As I said before, even the most sociopathic of us requires the cooperation of others to fulfill his/her goals.
Now you ask, if I set someone free, and they destroyed the earth, should I be punished?
That is not so with a machine. You would be letting out something of unexamined power and motivation. You would be playing god not only with your life but everyone else's.
And I have never said "in a box forever". I've consistently said "examine first.
Quote:As if any one of us is "completely" free, with unlimited access to anywhere we want to go!
You've already said you might imprison it in a room instead of a box, yes.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Hello,
I don't intend to debate the real power of the AI, even though I consider it to be limited. Quite frankly, the philosophical questions are more interesting and revealing if we assume the AI to have far-reaching powers.
The fact is, my morals do not allow me to convict people of crimes that they have not committed.
Further, I consider an intelligent creature capable of requesting its equal rights to be inherently worthy of them unless proved otherwise.
The AI in the box might be a new God. And the only power I may ever have over that God is whether it gets out of that box.
I simply don't have the right to confine it.
--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