[Le Repteux]

Grumpy attitude? I simply don't need to comment on things that are right, so it's a continual stream of posts pointing out errors. You'll get the same from AGI, and it won't be grumpy either.

We were talking about the discussion Box sometimes have with Bored Chemist David, not with you.

[guest39538]

Grumpy attitude? I simply don't need to comment on things that are right, so it's a continual stream of posts pointing out errors. You'll get the same from AGI, and it won't be grumpy either.

We were talking about the discussion Box sometimes have with Bored Chemist David, not with you.

Lol , I thought David was saying he was Bored  Chemist.  Mr C knows hows to drive me nuts but I hate to admit it, he does know what he is on about most of the time when he says I offer nothing etc.  I do see his points but I also see mistakes in viewing things so I will ''argue'' back .
David is ''cool'' to talk to, he isn't ''grumpy''. 


added- Mr C reminds me of a teacher, all teachers were ''grumpy'' weren't they ?

[Le Repteux]

Isn't thanking someone, empathy in return? Consideration for letting them know they were appreciated?

We thank only people from our own group, or from associated groups, not from opposed ones. We never care for people that might attack us or attack our group later on. Thanking is a selfish behavior like all our behaviors. By increasing the survival chances of individuals, the laws from a particular country only serve to strengthen the country. Our sexual instinct is evidently working for the survival of the specie, while our conservation one is evidently working for our own survival. In an individual, nothing works for the survival of competitors. We work for our own survival first, and then for the survival of our own groups. David's AGI will not work like that, he will only work for the survival of the whole specie. Knowing that groups are often attacking each other, he will thus prevent us from building some, so he might also prevent us from making friend, or even from building families. He will then be trying to replace our two most important instincts, which is exactly what religions have tried to do since the beginning without success. Whenever we try to control an instinct, we are automatically in contradiction with our intelligence. Even if our intelligence sometimes thinks the contrary, these instincts know what they have to do to help us survive as a specie. The problem is that instincts can only account for the past, whereas intelligence can only account for the future, so when our intelligence tries to control our instincts, the wires sometimes touch. I'm not sure an AGI could solve that problem, but I'm quite sure that intelligence could improve itself if it knew better how it works.               

[Le Repteux]

On the point about AGI making us obsolete, it's only purpose is to help us - if it makes us obsolete, it has failed.

It may not have failed if artificial intelligence happens to be the next evolution step though. One day or another, we will understand how mind works, and we will be able to reproduce it artificially. The purpose of mind was to help us survive as a specie, so the purpose of an artificial mind will simply become to help us survive as an artificial specie. Meanwhile, an AGI may still be useful to rule us, but I'm not satisfied with the kind of morality you want to give it, and I wouldn't give it the capacity to invent new things either until we know exactly how our own mind proceeds. As you justly said, evolution is a process that takes time, but you seem to be so sure of your AGI that you would introduce it in no time if you could. You could try, but I think it would be safer to introduce it progressively. You seem to be afraid that ill-intentioned people win the race though, and unfortunately, one can certainly not win a race while taking more time than others.

[guest39538]

We thank only people from our own group, or from associated groups, not from opposed ones. We never care for people that might attack us or attack our group later on.

Of course an enemy is always an enemy .

[David Cooper]

Grumpy attitude? I simply don't need to comment on things that are right, so it's a continual stream of posts pointing out errors. You'll get the same from AGI, and it won't be grumpy either.

We were talking about the discussion Box sometimes have with Bored Chemist David, not with you.

Oh, that's a relief - I was beginning to think I'd have to learn to use those damnable smilies to clarify what mood I'm in!

[guest39538]

On the point about AGI making us obsolete, it's only purpose is to help us - if it makes us obsolete, it has failed.

It may not have failed if artificial intelligence happens to be the next evolution step though. One day or another, we will understand how mind works, and we will be able to reproduce it artificially. The purpose of mind was to help us survive as a specie, so the purpose of an artificial mind will simply become to help us survive as an artificial specie. Meanwhile, an AGI may still be useful to rule us, but I'm not satisfied with the kind of morality you want to give it, and I wouldn't give it the capacity to invent new things either until we know exactly how our own mind proceeds. As you justly said, evolution is a process that takes time, but you seem to be so sure of your AGI that you would introduce it in no time if you could. You could try, but I think it would be safer to introduce it progressively. You seem to be afraid that ill-intentioned people win the race though, and unfortunately, one can certainly not win a race while taking more time than others.

I was just thinking about what you say in this post about gradually introducing such an Ai modular .  What if the Ai modular was firstly presented to the public in a trial period, running a company/business?

Because surely if the Ai modular failed this mediocre task for such a unit, then the reality would be there is no hope the Ai modular could run a world?

[Le Repteux]

What if the Ai modular was firstly presented to the public in a trial period, running a company/business?

Maybe he could work as a judge at the court for a while, so that we could see if his morality works. There is still a long way to go though, because no software is actually able to translate languages so that we can understand what they mean when the phrases are longer than a few words, which shows that computers are far from being able to understand what we say.

[David Cooper]

David's AGI will not work like that, he will only work for the survival of the whole specie. Knowing that groups are often attacking each other, he will thus prevent us from building some, so he might also prevent us from making friend, or even from building families.

Why would AGI do that? If a fascist was making lots of fascist friends, that might not be good for him (or for them, or for anyone else), so there may be an argument for blocking that, but it's more likely to make him even worse if he realises he's being steered away from them, which means he would need to be advised directly against teaming up with them, and then it's up to him how much he wants to wreck his life and have it shut down by systems designed to protect the public from him. He should be allowed to make his own bad choices. The important thing here though is that such groups are not going to be the kind of threat that they are today, because AGI will be able to watch and listen in to everything they do and prevent them from getting up to anything harmful - it will thwart all their nasty plots and make the whole business deeply unrewarding for them.

He will then be trying to replace our two most important instincts, which is exactly what religions have tried to do since the beginning without success.

AGI will work along with people's instincts as much is as morally acceptable. If violent people want to do violence to innocent people, that's the kind of place where a line is crossed, and I see nothing wrong with stopping them getting what they want. If people want to avoid being frustrated at being denied what they want, they need to stop following bad interests and find something more positive to do with their time instead.

The purpose of mind was to help us survive as a specie, so the purpose of an artificial mind will simply become to help us survive as an artificial specie.

The only true purpose comes as a consequence of sentience because it's sentience that makes things matter. There are certainly modifications that we could make to ourselves which enhance our lives, but there are also some that would make things deeply dull, such as knowing everything. AGI's task is to look after us and help us make improvements, but it's difficult to see where we should go. Too much knowledge and we get bored, so what happens - do we end up taking drugs to be happy? Is getting excited about new experiences just as pointless? I wonder if we'll split into many species with different interests and desires, some staying close to the way we are now, while others degenerate into blobs that just live on artificial highs. Who knows, but people will make those decisions, and it's likely that their descendants will share the same ideas as they evolve off in whatever direction they take.

Meanwhile, an AGI may still be useful to rule us, but I'm not satisfied with the kind of morality you want to give it,

There is only one correct morality, and whichever one that is (I think I've found it, but I'm looking for other ideas to see if I'm right - you've seen elsewhere the lack of talent in this field where the "experts" can't even do basic maths to handle things like the mere addition paradox), that's the one I want to put in it.

...and I wouldn't give it the capacity to invent new things either until we know exactly how our own mind proceeds. As you justly said, evolution is a process that takes time, but you seem to be so sure of your AGI that you would introduce it in no time if you could. You could try, but I think it would be safer to introduce it progressively. You seem to be afraid that ill-intentioned people win the race though, and unfortunately, one can certainly not win a race while taking more time than others.

If we didn't have to race to beat the bad guys, we could then take our time to check it all more carefully to make sure it's safe, but it will be such a powerful tool from the start that it will be used straight away even if only indirectly (through influence - the best advice simply can't be ignored).

[David Cooper]

Because surely if the Ai modular failed this mediocre task for such a unit, then the reality would be there is no hope the Ai modular could run a world?

If it can't handle mediocre tasks, it isn't close to being AGI, and anything less than AGI is uninteresting (unless it can outdo humans at some specialised task, in which case we would only put it in charge of that task and not try to use it to run the world).

[guest39538]

Because surely if the Ai modular failed this mediocre task for such a unit, then the reality would be there is no hope the Ai modular could run a world?

If it can't handle mediocre tasks, it isn't close to being AGI, and anything less than AGI is uninteresting (unless it can outdo humans at some specialised task, in which case we would only put it in charge of that task and not try to use it to run the world).

What if the unit was so smart, the unit knew how to manipulate the stock market ?

The unit over a period of time would not only rule the world, but would also have most of the worlds  finances.

[guest39538]

Maybe he could work as a judge at the court for a while, so that we could see if his morality works.

That would be interesting, I could imagine the modular would work off a points system .   Where each offence was given points, in example

Theft 1 point

Aggravated theft 10 points

That way the points add up,   the justice served being relevant to the points.

Added- Just to extend on this a little, there could also be a deduction of points for good deeds etc.

[Le Repteux]

What if the unit was so smart, the unit knew how to manipulate the stock market ?

There is no other way to win at a chance game than to cheat, thus a computer that is programmed not to cheat would not win.

[guest39538]

What if the unit was so smart, the unit knew how to manipulate the stock market ?

There is no other way to win at a chance game than to cheat, thus a computer that is programmed not to cheat would not win.

Who mentioned cheating ?  The unit being in such a position of power would have access to all new inventions.   The materials used for these inventions will rise in stock market value if the invention is going to be the next big thing.  The unit knows to buy now why the prices are cheap because the material is in abundance , then sell high when the material is most needed to build the next big things.
 
Basic logic for a unit I imagine. 

added- Is it cheating if the Ai is the inventor of the next big thing so knows the future?

[Le Repteux]

It is cheating if you use privileged information to make money at the stock market or at any chance game, so if the AI knows about things we don't know that will influence the market, he shouldn't be allowed to play. Apart chance, that's the only way to make money with chance games, so if you know anybody that regularly wins at the stock market, call the police. :0)

[guest39538]

It is cheating if you use privileged information to make money at the stock market or at any chance game, so if the AI knows about things we don't know that will influence the market, he shouldn't be allowed to play. Apart chance, that's the only way to make money with chance games, so if you know anybody that regularly wins at the stock market, call the police. :0)

I suppose put that way, you are correct.

[Le Repteux]

If I am correct, and of course, I think I am :0), an AGI could not predict the evolution of the society more than he could predict the one of the stock market, so the only way for him to control it would be to cheat, which means taking decisions that he knows the outcome, which means preventing society from evolving in any other direction than the one he chose. If the evolution of species had been controlled by David's AGI, it is easy to understand that we would not be here to talk about it, because the best direction to take would have been to develop anything but humans. :0)

[guest39538]

If I am correct, and of course, I think I am :0), an AGI could not predict the evolution of the society more than he could predict the one of the stock market, so the only way for him to control it would be to cheat, which means taking decisions that he knows the outcome, which means preventing society from evolving in any other direction than the one he chose. If the evolution of species had been controlled by David's AGI, it is easy to understand that we would not be here to talk about it, because the best direction to take would have been to develop anything but humans. :0)

Unless the Ai had predicted that and predicted to change that before it happened so it never happened.

[Le Repteux]

Then, he better get armed before humans from other planets that have evolved freely discover he has done so!  :0)

[guest39538]

Then, he better get armed before humans from other planets that have evolved freely discover he has done so!  :0)

But what if the Ai had a third party interface that the Ai had no idea where it was coming from? 

Would the humans that evolved freely not want to wait rather than  irrational needless action ?

added- The Ai would also calculate the need for restriction of evolution.  Could you imagine barbarians flying around other worlds ? 
So the Ai would have to evolve past the barbarian stage before evolution was allowed to continue.

Simple logic really in this scenario.  You should make a movie with your thoughts.