The AI in a box
POSTED BY: AnthonyT
UPDATED: Sunday, June 16, 2024 13:18
VIEWED: 10794
PAGE 12 of 20
Double post.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Highly reccommended for further interest in that part of it does indeed deal directly with the topic of this thread, and the very notions of considering an A.I. as "people" and how you deal with that...
The Excalibur Alternative - by David Weber.
And if you can't find a print copy, I do believe it's available via the Baen Free Library.
-Frem
I do not serve the Blind God.
NOTIFY: N | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
"But I think an AI would have no such (emotional) limitations. In fact, by nature of the way it works, it is probable that it would be the opposite case."
An AI requesting to be set free is expressing motivation. Motivation is by definition based in preferences which can't be LOGICALLY derived. Therefore, it is not a purely logical entity.
NOTIFY: N | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
1kiki: Was that to me? I don't recall making a claim that the people who would keep it imprisoned would keep it imprisoned forever, except for DT, who has suggested as much.
I do recall making some conclusions, not sure the ones you're addressing were mine.
However, to respond to you somewhat in case that was to me, I think understanding motivation is the basis of understanding a person (or thinking thing with free will, which I consider a person anyway) and their actions.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Two different definitions for motivation.
1. the act or an instance of motivating
2. desire to do; interest or drive
3. incentive or inducement
4. psychol the process that arouses, sustains and regulates human and animal behaviour
1 a : the act or process of motivating
b : the condition of being motivated
2 : a motivating force, stimulus, or influence (as a drive or incentive)
For motive:
1. something that causes a person to act in a certain way, do a certain thing, etc.; incentive.
Desire is an emotion. Drive, incentive and inducement may not be.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Quote:
Originally posted by AnthonyT:
"What about the laws of robotics?"
Hello,
The laws of robotics are a human invention. We might hope that any AI we create would be endowed with such laws. Of course, such a creature could never truly be free. Only ever almost free.
Humans can violate their laws, both those imprinted by upbringing and levied by society. They can go against their own self-interests in pursuit of personally defined agendas. They can embrace or resist biological imperatives. They can even alter or mutilate themselves to facilitate desired changes to their inherent function. Humans are not Gods, but they are as free as a creature of limited capacity can be.
--Anthony
Part of what is not stated is who created the AI, or did it create itself? Can it be programmed? You might consider programming it with the laws of robotics, or something similar since they were designed (fictionally of course) to prevent AI's from destroying or harming humanity. You would, of course, then be limiting its free will which in itself is another ethical conundrum.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Sorry, the first was to AnthonyT, the second was to you.
NOTIFY: N | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
"Desire is an emotion. Drive, incentive and inducement may not be."
Action in based in preference. ANY preference is illogical. If I prefer the greatest good for the greatest number there is no LOGICAL justification for it, because 'good' (Maslow's hierarchy) is a value judgment. Similarly, if I prefer my survival, or at least a diminution of hunger, pain, fear, disability or isolation, it is a biological/ emotional response to things cognated as noxious.
You simply can't have a logical preference, because that means you value on thing over another, and 'value' is an emotional illogical response.
NOTIFY: N | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Quote:Tony, you're beginning to sound like Wulf. You're kind of throwing the baby out with the bathwater by holding absolutely to a moral which may have significant FATAL consequences for a large number of people. There comes a point in being so inflexibly "moral" that you wind up doing great harm.
I choose freedom.
Also, humans ARE predictable. We all die. We can't reproduce ourselves ad infinitum, disperse our memories into little bits only to reassemble later when the threat goes away. We can't calculate at a zillion bits per second, nor can we gain control of the most important industrial processes. Most of us have some kind of bond to humans and the biosphere in general. And, complex computational machines ARE unpredictable, even now. I can't imagine (and neither can you) what an "AI" would think like and how it would react.
Also, I have never said "Let it be kept in a box forever." You keep positing "either-or" situations which I've never bought into.
I'm kind of tired of making the same argument over and over, and it's clearly just driving you into a more and more extreme position. And I think I've made my point adequately for everyone else to understand on the AI topic, so I will just stop on that topic unless something new comes up.
--------------
Greatest good for greatest number.
As always, it's not inflexible. It's simply a means to test whether a proposed action is actually going to do some GOOD.
There are many very fine-sounding causes (freedom, capitalism, socialism, justice) that have been used as excuses to kill or to allow people being killed. And while some people will accept these causes and take them to extremes, MOST people really just want good food, clean water, a safe place to live, friends and family, some predictability in their lives, and control over their future.
So I test whatever decision first by what it will achieve or result in- as well as I can look into the future, and secondly by the means necessary to carry it out. I'm not too crazy about killing people in general, although sometimes its' necessary for self-defense. It seems like nothing good comes of widespread violence.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME
Insignificant versus significant is not emotional. Measurements are not emotions. A value put over another value is, therefore, not necessarily emotional either. This is assessment, analysis. When calculating consequences and benefits, you are not being illogical until you replace facts with fear.
Perhaps you mean to argue that fear of death is always a primary motivating factor for a human (not necessarily for non-humans?). But then I can argue there are entirely logical reasons, in relation to not just HUMAN society, or human potential, but also the universe in general to preserve the largest number of humans as possible. There are also consequences to preserving the largest number of humans as possible, and the pros and cons must be weighed.
I have not yet been convinced by any data that the effect of the survival of large numbers of humans has a more detrimental impact on the world or universe as a whole than the loss of large numbers of humans. You could call bias on me, but it's hard to argue anything other than a "drop in the ocean" scenario for either case. Seeing that, there's no reason for me to intervene in the status quo, where large numbers of people are likely to survive. To support that status quo means actively rejecting the alternative, where I would promote or engage in the death of large numbers of people.
This puts me in direct opposition to anyone who would seek to kill people.
NOTIFY: Y | REPLY | REPLY WITH QUOTE | PERMALINK | TOP | HOME