Ethics based on achieving goals

[Elaro]

Second paragraph; first sentence!

I was thinking of a common ethical system that we could share with beings that don't feel like us, specifically an embodied AI person.

Also, general note: animals=dumbasses who generally don't live as long as humans, so if you have to choose between an animal and a human, it is likely the human, being smarter, will achieve the more goals, therefore it is more ethically correct, according to my system, to choose the human.

And yes, if we had to choose between an alien that has a lifespan of thousands of years and a human, if they're both otherwise the same, then yes, it makes more sense to choose the alien. But we don't have to choose, now, do we? So hold your horses.

[Purple]

And yes, if we had to choose between an alien that has a lifespan of thousands of years and a human, if they're both otherwise the same, then yes, it makes more sense to choose the alien. But we don't have to choose, now, do we? So hold your horses.

No it would not. I am human, they are not human. Ergo it is by definition wrong for me to pick him.

[Starglider]

And yes, if we had to choose between an alien that has a lifespan of thousands of years and a human, if they're both otherwise the same, then yes, it makes more sense to choose the alien. But we don't have to choose, now, do we? So hold your horses.

No it would not. I am human, they are not human. Ergo it is by definition wrong for me to pick him.

The definition of what?

[Purple]

The definition of what?

It was a figure of speech. Although admittedly one that I probably should not have chosen.

[Simon_Jester]

Elaro, 'fixing' the anthropocentrism doesn't fix the fundamental problem. As written, it's effectively impossible to decide that any course of action is "good" because just the act of picking a course of action prevents any number of other people from achieving hypothetical "goals." If I eat a cookie, then my wife cannot eat the same cookie... nor can any of the many other people who might have eaten it. Assuming that 'eat a cookie' is an equally worthy goal no matter which of us accomplishes it, this creates deadlock.

And more generally still, any "function" that recursively calls upon itself nonstop this way isn't going to fly.

[Elaro]

Elaro, 'fixing' the anthropocentrism doesn't fix the fundamental problem. As written, it's effectively impossible to decide that any course of action is "good" because just the act of picking a course of action prevents any number of other people from achieving hypothetical "goals." If I eat a cookie, then my wife cannot eat the same cookie... nor can any of the many other people who might have eaten it. Assuming that 'eat a cookie' is an equally worthy goal no matter which of us accomplishes it, this creates deadlock.

And more generally still, any "function" that recursively calls upon itself nonstop this way isn't going to fly.

Okay, maybe I didn't express myself well.

A "goal" is a property that we wish the world would have. "Having" a goal means wanting that property to be realized. It's a mental state. This mental state compels us towards reaching this goal. People have goals, and so far, our morality has gone more and more into the direction of letting people achieve their goals.

What I'm trying to say is: we should act such that the most goals are accomplished in time and space; and we should accept or reject a goal based on our other goals, the ones that we have now and the ones that we will have later; this is basic prudence.

I mean, the big question that I try to answer with this system is: Why, in a system of ethics, in a bill of rights, do we have this rule or that right, and not another? My answer, is the maximization of the achievement of goals. We choose to give up on certain goals in order to achieve more other goals as a society. This valorization of goals has some advantages: first of all, goals are real, you can test against them. Secondly, there's no hierarchy to mess up. Thirdly, you rank by counting. Hey, computers are good at counting, eh?*

In fact, I'm going to go so far as to say that our sense of morality is based on this calculation of goals. But I'm not sure.

*Actually, computers are terrible at counting things they don't already know.

[Simon_Jester]

Thing is, ethics has nothing whatsoever to do with naive counting of goals, and everything to do with the weighting and evaluation of goals on their merits both in and of themselves, and in relation to other goals a person might have. It simply does not and can not make sense to define ethics in terms of 'counting' goals.

[Elaro]

Thing is, ethics has nothing whatsoever to do with naive counting of goals, and everything to do with the weighting and evaluation of goals on their merits both in and of themselves, and in relation to other goals a person might have. It simply does not and can not make sense to define ethics in terms of 'counting' goals.

How do you weigh and evaluate goals in and of themselves? Without the existence of other goals, that is to say, in a universe with only one creature in it and that only has one goal, how can you honestly say what that person should and shouldn't do? It should and is free to do whatever it wants.

You're not going to convince someone of doing something different, of renouncing on their goal, except by appealing to their other goals. Perhaps their goal of being a good person, or their goal of being socially accepted, or their goal of having a lineage. My point is that the currency of the will is goals. You can only change someone's will by counting the number of goals they'll lose out on if they continue.

My ethics, and this I was almost aware of, is profoundly rooted in pragmatic problem-solving and conflict resolution between real people. There is no "objective" righteousness, there is no universal measuring stick by which you can measure the value of a goal isolated from other goals. The only value that exists is the one we give, and goals are the way that the will of existence-kind spreads value in the world. If there is but one goal, in all of time and space, then it is the best thing to accomplish, simply because there is no reason, no value, in doing anything else.

My point is that you cannot evaluate a goal based on its own merits, because when talking about a single goal, well, merit goes out the window. It has all the merit.

Also, how do you know what ethics is about? Have you provided a definition? Why doesn't it and can't it make sense for ethics to be defined in terms of counting goals? Please, make your case.

Google defines ethics as such:

moral principles that govern a person's or group's behavior.

How does goal-seeking not govern a set of people's behaviour? How is "maximizing the number of goals achieved by that set" not a good indicator of what that set will do?

Please, make your case. Don't dismiss me out of hand.

[Simon_Jester]

How do you weigh and evaluate goals in and of themselves? Without the existence of other goals, that is to say, in a universe with only one creature in it and that only has one goal, how can you honestly say what that person should and shouldn't do? It should and is free to do whatever it wants.

The argument that this is not the case, or need not be the case, is part of the foundations of the entire study of ethics.

For instance, there is the Aristotelian notion that virtue is to optimize things, to make them match the ideal forms of themselves.

Perhaps more to the point, there is the Kantian notion of the categorical imperative, which is (among other things) specifically targeted at the kind of argument you're advancing, because what you're saying isn't new.

You're not going to convince someone of doing something different, of renouncing on their goal, except by appealing to their other goals. Perhaps their goal of being a good person, or their goal of being socially accepted, or their goal of having a lineage. My point is that the currency of the will is goals. You can only change someone's will by counting the number of goals they'll lose out on if they continue.

I could argue that your will has duties that are in play regardless of circumstances and which do not change no matter what goals you personally do or do not have.

I could also argue that these duties and imperatives have nothing to do with what anyone else wants either. As it happens, following such imperatives generally lines up well with the public interest. But one can argue that this is mere happy coincidence, rather than it being the case that it is right to do things because others desire that you do them.

Now, one can debate this on its merits. But one cannot pretend that such formulations of ethics simply do not exist, or that they have no weight.

Also, how do you know what ethics is about? Have you provided a definition? Why doesn't it and can't it make sense for ethics to be defined in terms of counting goals? Please, make your case.

Google defines ethics as such:

moral principles that govern a person's or group's behavior.

How does goal-seeking not govern a set of people's behaviour? How is "maximizing the number of goals achieved by that set" not a good indicator of what that set will do?

Please, make your case. Don't dismiss me out of hand.

Your definition of "ethics" is a limited one used in popular language, as in "professional ethics." It is not the definition used in philosophy. As a discipline of philosophy, ethics is the field concerned with determining and categorizing right or wrong action, and establishing rules that systematically define what is right or wrong conduct.

[Elaro]

How do you weigh and evaluate goals in and of themselves? Without the existence of other goals, that is to say, in a universe with only one creature in it and that only has one goal, how can you honestly say what that person should and shouldn't do? It should and is free to do whatever it wants.

The argument that this is not the case, or need not be the case, is part of the foundations of the entire study of ethics.

For instance, there is the Aristotelian notion that virtue is to optimize things, to make them match the ideal forms of themselves.

Perhaps more to the point, there is the Kantian notion of the categorical imperative, which is (among other things) specifically targeted at the kind of argument you're advancing, because what you're saying isn't new.

What I'm trying to get through your head is that there's no way to tell whether this supreme principle is actually good unless it accomplishes, directly or indirectly, the most goals held and that will be held by conscious beings.

You're not going to convince someone of doing something different, of renouncing on their goal, except by appealing to their other goals. Perhaps their goal of being a good person, or their goal of being socially accepted, or their goal of having a lineage. My point is that the currency of the will is goals. You can only change someone's will by counting the number of goals they'll lose out on if they continue.

I could argue that your will has duties that are in play regardless of circumstances and which do not change no matter what goals you personally do or do not have.

I could also argue that these duties and imperatives have nothing to do with what anyone else wants either. As it happens, following such imperatives generally lines up well with the public interest. But one can argue that this is mere happy coincidence, rather than it being the case that it is right to do things because others desire that you do them.

Now, one can debate this on its merits. But one cannot pretend that such formulations of ethics simply do not exist, or that they have no weight.

They have no weight because they have no convincing power unless the subject is already a proponent of such a formulation. By defining ethics in terms of real goals, you don't require belief in a higher power principle.

Also, how do you know what ethics is about? Have you provided a definition? Why doesn't it and can't it make sense for ethics to be defined in terms of counting goals? Please, make your case.

Google defines ethics as such:

moral principles that govern a person's or group's behavior.

How does goal-seeking not govern a set of people's behaviour? How is "maximizing the number of goals achieved by that set" not a good indicator of what that set will do?

Please, make your case. Don't dismiss me out of hand.

Your definition of "ethics" is a limited one used in popular language, as in "professional ethics." It is not the definition used in philosophy. As a discipline of philosophy, ethics is the field concerned with determining and categorizing right or wrong action, and establishing rules that systematically define what is right or wrong conduct.

And the criteria of maximizing the achievement of real and eventually-real goals does both of those things, and in a manner much more comprehensive of the various fields of human behaviour, like engineering.

[Simon_Jester]

Elaro, I'm going to be honest, I don't think you've read a broad enough selection of ethics with an open enough mind to fully understand the arguments involved.

From the sound of it, you haven't tried to familiarize yourself with ethical systems based on, say, universal imperatives. Because of this, your entire system is this circular redundant argument based on the idea that whatever goals people have are the sole arbiter of good, and that good is defined in terms of goals.

The problem isn't that no alternatives are imaginable. The problem is that no alternatives are imaginable to you.

[Feil]

Elaro:

Assigning a value of 1 to every goal is exactly as arbitrary as assigning different values to different goals.

Assigning a value of 1 to every goal doesn't comport with reality, in which people assign their own goals different values without ever considering knock-on effects to other goals.

Assigning a value of 1 to every extant goal and summing linearly across immediate extant goals achieved and goals rendered impossible assigns high ethics values to outcomes that are conventionally evil, following similar reasoning as the superficial application of hedonist utilitarianism and generating similar results.

Assigning a value of 1 to every goal including possible goals and summing linearly across goals achieved and goals rendered impossible creates ethical deadlock with every action having an ethics value of 0 or negative infinity, depending on syntax.

(The above is a summary, not an argument. Nothing I just wrote hasn't been said several times in this thread before.)

The problem isn't that people aren't understanding you. You presented an idea, people pointed out the problems with the idea, and after your initial responses met with further and more severe criticism, you assumed that we must have misunderstood because your idea is just so good. Then you went looking for ways to "clarify" it rather than assessing the criticism with your eyes open. In the process, you've ended up ignoring inconvenient arguments and just restating your original position over and over in slightly different ways.

Stop, go back, and attack your own idea along the above lines. Move forward with whatever remains.

If your idea holds up and you find good specific reasons why every criticism doesn't pan out, share them. If parts of it crumble and others hold, see if you can assemble new knowledge out of what's left. Maybe you don't have an viable ethical system, but you might have a fragment of a bigger one. And if you come to the conclusion that ignorance is all you have left, file it carefully and with respect: specific, well-considered ignorance is a kind of knowledge, too.

[Elaro]

Alright, fine, forget goals, replace it with "life projects". And forget that equation! It was never meant to be definitive, only indicative of the principle of the thing.
I suppose any equation used to calculate the ethics of a goal should take into account cost, likelihood of success, the number of other real goals that are dependent on achieving that goal, and the number of other real goals that are dependent on failing that goal.

[Purple]

How are you going to run that math without essentially being omniscient?

[Starglider]

I have an amusing project specced that might demonstrate just how hard this is to get right. I had to put it on hold late last year due to our son arriving very premature, but it's still on the list to do after I've completed some more urgent projects.

[Simon_Jester]

Is there any way you can outline why it might demonstrate that which you say it might?

[madd0ct0r]

Im assuming this is starglider's ethical symbol manipulation program, wgere tou set up a problem, convert it into formal statements and itll spit out a result.

[Starglider]

Yes, I need to make this a bit more user friendly so that people not experienced in propositional logic can have a go, but it should be amusing.

[Elaro]

But we are in agreement that the preferred state is the one in which there are more beings who are in a state of satisfaction or, equivalently, not in a state of dissatisfaction, with the caveat that a "state of satisfaction" is based on goals achieved, not emotional state?

So the thing with the cookie? If nobody eats the cookie, then everybody's goal of eating the cookie is "unachieved", whereas if one person eats the cookie, then everybody minus one's goal of eating the cookie is "unachieved", and since everybody-1<everybody, then there is less dissatisfaction in the world for somebody (me) having eaten the cookie. Right?

[Terralthra]

Sure, but would the world be better if you ate the cookie, or I ate the cookie? The world is equally +1 goal satisfied in either case. Neither case is distinguished in your logic, and that's only the basest level, which doesn't take into account things like the fact that I paid for the cookie, or you could easily buy ten cookies, and so on confounding factors. That's what people are trying to express to you, and it's not getting through.

[Esquire]

I think you'll find that overall dissatisfaction is increased when somebody gets something that might (ought?) to have been shared out. One person gets a cookie's worth of satisfaction, eight get a cookie's worth of dissatisfaction and one cookie-theft's worth of dissatisfaction.

[Terralthra]

I think you'll find that overall dissatisfaction is increased when somebody gets something that might (ought?) to have been shared out. One person gets a cookie's worth of satisfaction, eight get a cookie's worth of dissatisfaction and one cookie-theft's worth of dissatisfaction.

All of these things are true, but none of them are adequately represented by goal-satisfaction-based logic.

[Esquire]

That was rather the point.

[Elaro]

Sure, but would the world be better if you ate the cookie, or I ate the cookie? The world is equally +1 goal satisfied in either case. Neither case is distinguished in your logic, and that's only the basest level, which doesn't take into account things like the fact that I paid for the cookie, or you could easily buy ten cookies, and so on confounding factors. That's what people are trying to express to you, and it's not getting through.

Yes it is, because the goal of "me eating a cookie" and "you eating a cookie" are different, and must be evaluated individually. How many goals are helped along by you eating the cookie versus me? And this theory encompasses everything that is real, so it does count emotional state, especially if you need to not be distressed to achieve goals, and if me stealing a cookie would distress you, then this theory gives a clear answer as to why you shouldn't steal the cookie.

Elaro:
Assigning a value of 1 to every goal is exactly as arbitrary as assigning different values to different goals.

Only a priori. The crux of this theory consists that individual goals have different values, where one goal has as worth how many goals are facilitated by the accomplishment of that goal.

Assigning a value of 1 to every goal doesn't comport with reality, in which people assign their own goals different values without ever considering knock-on effects to other goals.

Yeah, and people are beacons of moral rectitude and never make any ethical mistake ever, right? I don't care what people do, I care what people should do, and what they should do depends on what they want.

Assigning a value of 1 to every extant goal and summing linearly across immediate extant goals achieved and goals rendered impossible assigns high ethics values to outcomes that are conventionally evil, following similar reasoning as the superficial application of hedonist utilitarianism and generating similar results.

Does it?

Also, maybe the convention is wrong?

Assigning a value of 1 to every goal including possible goals and summing linearly across goals achieved and goals rendered impossible creates ethical deadlock with every action having an ethics value of 0 or negative infinity, depending on syntax.

Yes, which is why possible goals should not be given the value of 1 but the value of the probability of a person wanting that goal.

Like I said, I explained poorly. I'm actually currently working on a formal definition of this theory, but I'm tripping over the actual definition of a course of action.

[Feil]

I'll be honest, I haven't seen 'wait four months and THEN repeat myself in more strident terms' as an argumentative tactic before. Congratulations on the novelty, I guess.

Unfortunately, I have thirty trillion things I want to do more than try to coax critical thinking out of you, so it would be profoundly evil by your own metric for me to type another wo