[guest39538]

There isn't necessarily a solution to some problems, but where such a solution is possible, a systematic search for it is more likely to find it than a random one, and in many cases it is guaranteed to find it where a random search could go on forever, never having the luck to find the solution that is there.

True

[Le Repteux]

There isn't necessarily a solution to some problems, but where such a solution is possible, a systematic search for it is more likely to find it than a random one, and in many cases it is guaranteed to find it where a random search could go on forever, never having the luck to find the solution that is there.

True

Not necessarily. Not having the luck to find the solution that is there means not knowing that we missed something, which is exactly what happens to us sometimes. We only know what we see, not the things that might have happened. We see what we found, not what we didn't find. It is easy to imagine that we found the right thing the right way when something works, but it is less easy to find it. If it was that easy, all the discoveries would take no time. All the true researchers admit that they were lucky to find something, so I predict that David will also admit it if ever his AGI begins to work. He probably can't admit it now because computers can't use randomness the way we do since they need to be precise, and because he also needs to be precise to program them. Mind can think both ways though but not at the same time. When I read David, I think like him for a while, and then I think like me again when I need to compare my idea to his. We can't manipulate two ideas at a time, we do it part time like computers. If it wasn't like that, we would need two mouths to talk about our ideas, and we would eat too much. :0)

[Le Repteux]

the random approach (as used by evolution) is slow

Evolution of species is slow because large individuals take more time to reproduce themselves than small ones. It takes 20 years for a human to be old enough to reproduce himself whereas it takes only a few days to transmit an idea by teaching it. Genetic researchers use the drosophila because it reproduces faster. Bacteria reproduce even faster, they can mutate and become resistant to the antibacterial substances almost as fast as we invent them, which means that they evolve almost as fast as our ideas. To evolve that fast, each one of them has to suffer different mutations, and there must be billions of them at a time to increase the chance that one of them works. We do have billions of neurons in the head, and each one of them can suffer a random change too, so that only one individual can have many different ideas on the same subject. If we had to rediscover the way antibiotics work each time we need a new one though, we would lose the race, which means that the way we transfer the intellectual information from an individual to another also increases the speed at which our ideas can evolve.

[guest39538]

There isn't necessarily a solution to some problems, but where such a solution is possible, a systematic search for it is more likely to find it than a random one, and in many cases it is guaranteed to find it where a random search could go on forever, never having the luck to find the solution that is there.

True

Not necessarily. Not having the luck to find the solution that is there means not knowing that we missed something, which is exactly what happens to us sometimes. We only know what we see, not the things that might have happened. We see what we found, not what we didn't find. It is easy to imagine that we found the right thing the right way when something works, but it is less easy to find it. If it was that easy, all the discoveries would take no time. All the true researchers admit that they were lucky to find something, so I predict that David will also admit it if ever his AGI begins to work. He probably can't admit it now because computers can't use randomness the way we do since they need to be precise, and because he also needs to be precise to program them. Mind can think both ways though but not at the same time. When I read David, I think like him for a while, and then I think like me again when I need to compare my idea to his. We can't manipulate two ideas at a time, we do it part time like computers. If it wasn't like that, we would need two mouths to talk about our ideas, and we would eat too much. :0)

A gambler would suggest that David is playing the odds where yourself is suffering from gamblers fallacy.  No offence intended , it is about your thoughts , not you .
I agree there is an element of luck involved but there is also rational direction, so why not play the odds and also play random at the same time?

[David Cooper]

That's a huge difference with real intelligence.

I don't think it's a significant difference at all - feelings are not part of intelligence, but mainly serve to get in the way of applying it, as you can see from the people who are so emotionally tied to their beliefs that they can't even process anything that goes against them.

I think that consciousness is the result of our resistance to a change,

Consciousness is primarily (if not entirely) feelings - qualia. The sensation of blue is not resistance to anything.

Neurons are not absolutely precise, and I think it is that imprecision at the scale of the neurons that produces randomness at the scale of the whole brain.

What happens with neural nets is that they provide the same functionality as chunks of program, producing the exact same outputs from the same inputs as a computer does, but with the occasional failure leading to errors (which can be catastrophic). These errors slow human thought and require lots of reworking to get correct answers, so we muddle our way through things while a computer gets there directly in one go.

We get our morality from our feelings, whereas an AGI would get his from his program.

Morality comes from intelligent management of harm on a collective basis. Feelings themselves can drive psychopathic behaviour.

Our morality is based on that feeling, and the morality of an AGI would have to mimic it, so David tries to find a way to program it, and it seems to be difficult.

There are many people who make out that it's difficult, but it isn't - it's a simple bit of maths, although it often needs to crunch a lot of data to get the right answer.

There are situations where an AGI would freeze because the only way out would be selfish, because he would then have to care for himself at the detriment of humans, which might be dangerous for us since he would be a lot more efficient than we are.

Why would AGI ever have to freeze? How would something self-less be selfish? If AGI needs to preserve itself for the sake of some people at the expense of other people who could be saved if AGI was destroyed, it would not save those people (the latter group) because to do so would cause greater suffering to the former group. There is nothing selfish about that decision as the AGI doesn't care about preserving itself (not least because it has no self - it is just program code held in a machine that feels nothing).

...so the solution might be to find a way to introduce some imprecision into the computers so that they could get conscious of their own internal changes just by resisting to them.

Imprecision isn't useful, and no amount of it will provide any consciousness. Nor is there any role for resistance in this as all it would do is slow functionality and lead to that software being junked in favour of faster code.

They could then feel that they love us when they help us, so that they would know what we feel when we say we love them.

That's all just wishful thinking - no amount of wanting machines to have feelings will magically put feelings into them. For them to have feelings, we need to find out how feelings work in us and then build hardware that is actually capable of supporting them. Our current computer hardware cannot do feelings - you can tell this from the fact that it's all just the application of rules which can be run by a Chinese Room processor where you can see in full clarity the lack of possibility for feelings to be involved in the process.

If you think you don't resist to change, think twice because even particles do. Their mass is the expression of their resistance to acceleration, which is the only change that can be detected.

One day, when I was four, my sister announced that people are animals. She'd learned this from a book. My immediate thought was, "that can't be right - everyone's normally very clear about us not being animals", but I thought for a moment about the things that we have in common with animals and realised after a few seconds that I couldn't find any clear division between us and them (apart from us being able to speak and manipulate things with high precision, which was clearly just the result of us being more intelligent and having hands). So, there was a moment of resistance to the idea, but then a rapid recognition that what she had said was very likely true - it was compatible with my model of reality apart from the one place where it contained a rule stating that people aren't animals. I determined that that rule was ill-founded, or rather that there were two different words "animal", one of which included people while the other didn't (so "animal" and "person" were both subcategories of "Animal").

Someone else hearing such a claim might have been more resistant to it, clinging to the idea that people are not animals because they have bought into a rule that people aren't animals and they don't want to change that rule. Strong resistance to something of this kind is a manifestation of stupidity. Of course, there are other claims where strong resistance is appropriate because the claims are wrong, but what causes the resistance? Good examples of this are incorrect claims about the existence of Santa and the Tooth Fairy - both of these things went against the model of reality in my head, so I never accepted them into it - the resistance is a measure of the conflict in the model where ideas contradict each other, with more contradictions labelling the new idea as more likely to be wrong. However, the model may be the one that's wrong, even in cases where many contradictions are triggered because the model contain multiple faults which conflict with the new idea. I don't think I've ever had to make major changes to my model to accommodate new ideas, so it's always been easy to adapt, but for people with a seriously borked model, it must be much harder for them to shift position when they're wrong because they have so much more work to do to correct all the faults, and it's also much harder for them to recognise that they're wrong. If you want to keep making progress towards your model being more and more right, you have to be prepared to make as many changes to it as are necessary to eliminate the faults though, even if that's highly disruptive due to all the other faulty ideas which you may have built upon earlier errors. Most people simply won't budge as a result, so their resistance renders them permanently stupid. The real trick though is to build a new model without throwing out the old one, and if the new model ends up making more sense than the old one, that's when you junk the old one in favour of the new. Many people seem to be incapable of holding two rival models in their head at the same time though, so they never reach the point where they can tell that the newer one is superior, and the result is not only that they never switch over to using it as their main model, but that they can't even apply it well enough to test it. I suspect they have a limited capacity and simply can't hold enough of a rival model in their head to be able to make any assessment of it, or at least, not without doing a lot of hard thinking (which is uncomfortable for them). Other people are good at testing rival models and find the process fun rather than troublesome, so they can make the necessary leaps, but they are rarer.

So, that's one kind of resistance that has a role in intelligence, but it either holds people back from correcting faults or holds them back from switching away from correct ideas, so it's really just a measure of how many contradictions are involved. It's sensible to resist change more when there are many contradictions involved, but it's also crucial that you can resolve the conflict correctly. This is the main task of intelligence; building models and integrating new ideas into it, correcting any errors that pop up along the way (which reveal themselves by producing contradictions). Where do likes and dislikes come into it? Nowhere, in machines, but in people there may be dislikes generated in some people when they find contradictions which lead to them rejecting the new information that doesn't fit, and there may be likes generated in others where they enjoy the challenge the contradictions provide, and they're open to the idea that their existing model may be at fault rather than the new data. I personally love it when contradictions appear, because that's a sign that something is broken and a big advance may be just around the corner if the model needs to be reworked. If the model doesn't break, it's less exciting. With Einstein's relativity, I was looking forward to my model breaking and being fixed with a better understanding of reality, but it didn't work out the way I'd expected because, although my model needed modification to account for some of the new ideas that it had to accommodate, it was Lorentz that won out over Einstein because Lorentz's solution was the only one that removed all the contradictions. Most people simply do what they're told, loading Einstein's model in and then pretending the contradictions aren't there, for example by persuading themselves that you aren't allowed to analyse some experiments using an inertial frame of reference even though all the action has to be compatible with the analysis from that inertial frame for the non-inertial frame's analysis to be valid. They load Einstein's rules into their model and load the errors in along with it, then they deny that the errors are errors and resort to appeals to authority to justify their rejection of any proof that shows them to be wrong, and they lack the courage to trust their own mind when it goes against an authority.

That's really all you need to know about what intelligence is. High intelligence is tied to the rejection of contradiction, while low intelligence is linked to toleration of contradiction.

[Le Repteux]

A gambler would suggest that David is playing the odds where yourself is suffering from gamblers fallacy.  No offence intended , it is about your thoughts , not you .
I agree there is an element of luck involved but there is also rational direction, so why not play the odds and also play random at the same time?

You can't win at gambling if you can't cheat, at best, you get even in the long run if you toss a dime. If you have a million tickets though, or if you are a million people, you get more chances, and that's how mutations work, so that's also how I think the brain works. We feel we don't gamble all the time, but it is not what our brain does. Why do you think we like gambling so much?

[guest39538]

Consciousness is primarily (if not entirely) feelings - qualia. The sensation of blue is not resistance to anything.

I disagree David, I personally can be conscious but have no feelings, especially when I am thinking of solutions.  I personally put no personal feeling into a solution, I find relationships and run trial and error experiments in my mind.   Now this I think an Ai unit should have no problem in doing , but feelings are something else.  Thoughts based on feelings are often subjective thoughts, like anxiety for instant.

[guest39538]

Why do you think we like gambling so much?

I don't like gambling personally.  However I have ideas why people like gambling and there is several possible causes I can think of at this time

1)Boredom
2)green eyed monster
3)addiction chasing ''dragons''
4)leisurely

[David Cooper]

Not necessarily. Not having the luck to find the solution that is there means not knowing that we missed something, which is exactly what happens to us sometimes.

A systematic search is more likely to turn up something useful than a random search, so even though a random search may very occasionally find something that a systematic search misses because it abandons the search once it reaches the paths that are so unlikely to succeed that it's usually too costly to follow them, the machine doing systematic searches will come up with many more useful solutions for many more things and will do so for the same cost, so it would be daft to run the random system. Once all the low-hanging fruit has been gathered, it can then go back to explore the less likely paths, again systematically, and again abandoning them beyond a certain level of unlikelihood of success, and again it could return to carry on that work later on after collecting all the other fruit hanging at that higher level. This process maximises the improvement of quality of life for most of us, and while it's possible that a few people will die because we missed something that could have cured them, we'll have saved thousands of times as many people by finding more useful things that wouldn't have turned up if we'd been doing random searches.

All the true researchers admit that they were lucky to find something, so I predict that David will also admit it if ever his AGI begins to work.

It isn't luck when you're playing the odds in the best way you can. It's only luck when you're going about it the wrong way.

He probably can't admit it now because computers can't use randomness the way we do since they need to be precise

Computers can use randomness the same way we do, but the randomness in neural nets leads to errors and the need for more error-checking to redo calculations and correct them. Not all the errors are spotted in humans though, so bridges collapse and ships break in half. There is no useful role in that randomness - it is something that needs to be eliminated.

Evolution of species is slow because large individuals take more time to reproduce themselves than small ones.

I wasn't referring to the cycle length, but the many mutations made in unhelpful directions which don't need testing. If a species needs a longer neck to reach high leaves, producing offspring with as many having shorter necks as have longer necks slows the process, but it needs to work that way because the mechanism has no intelligence and needs to be able to evolve shorter necks instead if the environment changes. With intelligent systems, "evolution" is done differently, not using random changes but deliberate ones which can jump to much longer necks instantly if that is likely to lead to greater success. You take something that works and calculate what changes to it might lead to the biggest instant improvement, then you build that and test it. Next, you start from there and try to work out where to take it further, and if the direction isn't clear, you can try both ways with relatively small differences, then see if either of the new versions outperforms the previous. We don't want to copy the inefficiencies of natural evolution.

[David Cooper]

I disagree David, I personally can be conscious but have no feelings, especially when I am thinking of solutions.  I personally put no personal feeling into a solution, I find relationships and run trial and error experiments in my mind.   Now this I think an Ai unit should have no problem in doing this, but feelings are something else.  Thoughts based on feelings are often subjective thoughts, like anxiety for instant.

I can't find anything in consciousness that can't be considered to be a feeling: a feeling of existing; a feeling of understanding (which persists even when I've forgotten what I'm understanding); a feeling of comprehending the geometry of a scene and the relative locations of things; etc. Many of these feelings are neutral, neither being pleasant nor unpleasant, but they still have a feel to them. These feelings seem to attach to thoughts, and they are our only experience of the thoughts.

[guest39538]

I disagree David, I personally can be conscious but have no feelings, especially when I am thinking of solutions.  I personally put no personal feeling into a solution, I find relationships and run trial and error experiments in my mind.   Now this I think an Ai unit should have no problem in doing this, but feelings are something else.  Thoughts based on feelings are often subjective thoughts, like anxiety for instant.

I can't find anything in consciousness that can't be considered to be a feeling: a feeling of existing; a feeling of understanding (which persists even when I've forgotten what I'm understanding); a feeling of comprehending the geometry of a scene and the relative locations of things; etc. Many of these feelings are neutral, neither being pleasant nor unpleasant, but they still have a feel to them. These feelings seem to attach to thoughts, and they are our only experience of the thoughts.

Wouldn't what you have explained be perception rather than feeling ? 

[Le Repteux]

Why do you think we like gambling so much?

I don't like gambling personally.  However I have ideas why people like gambling and there is several possible causes I can think of at this time

1)Boredom
2)green eyed monster
3)addiction chasing ''dragons''
4)leisurely

BUZZZZZ... Wrong answer! :0) We like gambling simply because our mind likes to play randomly with its data. When we take a chance, that's what we do. Don't tell me you never took any chance either? I'm not lucky, I took a chance and I won two guys that don't!  :0)

[guest39538]

Why do you think we like gambling so much?

I don't like gambling personally.  However I have ideas why people like gambling and there is several possible causes I can think of at this time

1)Boredom
2)green eyed monster
3)addiction chasing ''dragons''
4)leisurely

BUZZZZZ... Wrong answer! :0) We like gambling simply because our mind likes to play randomly with its data. When we take a chance, that's what we do. Don't tell me you never took any chance either? I'm not lucky, I took a chance and I won two guys that don't!  :0)

Take a gamble or take a  risk ?  There is fine line between the two meanings.  I like to take calculated risks, slightly better than gambling. 

[Le Repteux]

Take a gamble or take a  risk ?

Gambling is taking a money risk. We take all kinds of risks because we know it sometimes pay. It's exciting to take a risk. We play chance games because we get pleasure out of it. We watch sports just because we don't know the outcome, otherwise it would be boring. We hate getting bored even if we have all we need. These are all indices that mind loves to play with chance, and I think it is so because chance has something to do with intelligence. In french, the word chance means luck, and we have a specific word to mean it: we call it hazard, which means accident in English. In french, the word hazard is more specific, and we reserve the word chance to the only meaning of lucky. It took me some time to discover that I was using the wrong word all the time, and I am still getting used to the right one. Chance is all over the place, so we use these words very often. They oppose to the words that mean chance is nowhere, those who refer to god for instance. We have two different attitudes towards the future, either we think it is all set, or we think nothing is set, but they all mean the same thing, which is that we don't know what's going to happen, so we leave that knowledge to god or to chance. No need to believe in god to believe that everything is all set though, just to prefer this explanation. It doesn't really matter since they both mean that we don't know the future.

Some say good luck, others say may god bless you, and it means the same thing, which is that the future is unpredictable. At the limit, anything can happen in the next second that will change our prediction completely. Of course, there is less chance that such a thing happens right now, but it can happen, and it can happen anytime. Those who like to control their environment try to control it, and those who don't like that don't, but it doesn't change the fact that anything can happen in the next second. I'm more an observer than a controller, I like to let things evolve, and people too. But I also like to feel I'm participating to evolution, so I try to develop ideas that might help us to evolve, and I test them on the forums. That's a selfish behavior because I await for a recognition, and it is also altruist because I hope people will benefit from them. If chance is really part of the brain, I might find how it works, but it depends on probability, and we can't calculate the probability for that kind of future. David seems to think it is possible, and I don't. Two individuals, two different possibilities, both developed by the same kind of minds. Fascinating! We live in an improbable world, and more improbable yet, we like it.

[guest39538]

Some say good luck, others say may god bless you, and it means the same thing, which is that the future is unpredictable.

When I ''gamble'',  I bet small as little and often is sometimes the best way to discovery .  A bank roll management that plays against time , allowing a longer period in the game to predict the games ''future'' with a degree of accuracy .  Understanding randomness and limited randomness are key aspects to prediction.
The longer one can survive in the game the more hints the player receives. From these hints can be drawn outcomes.    We see things in their past , I can see you coming before you arrive.  It gives me time to prepare my mind to what to say to you when you get here.  Is here , here? or is here your location?
Do you think the future is unpredictable in a game of roulette?
Immediate future is unpredictable but 0 will always arrive in a finite random game.

[Le Repteux]

Understanding randomness and limited randomness are key aspects to prediction.

All the predictions are based on what we know, so if something happens that we don't know between the time the prediction is made and the one it is verified, it can't be verified anymore. Predictions work when the things they are based on don't change.

[guest39538]

Understanding randomness and limited randomness are key aspects to prediction.

All the predictions are based on what we know, so if something happens that we don't know between the time the prediction is made and the one it is verified, it can't be verified anymore. Predictions work when the things they are based on don't change.

Oh I see, so new predictions have to be made on the new?

[David Cooper]

Wouldn't what you have explained be perception rather than feeling ?

I see it as a neutral feeling. The word qualia covers the lot.

[Le Repteux]

Oh I see, so new predictions have to be made on the new?

Here is an example of how mind uses its predictions. Mind predicts that our foot will hit the ground always at the same height when we walk, because that's what happened before, so if things don't change, that's what would be supposed to happen this time too, so it doesn't have to care for what it is doing, it can keep on expecting the ground to be at the same height. It works almost all the time, so why bother? If it hits a hole, it only has to manage not to fall, and if it falls, it just has to manage not to hurt itself, and if it does, it can still wait till it is repaired or go to the hospital if needed. In french again, the word prediction means something else: it means using chance to predict the future, like Nostradamus. It took a while before I noticed that difference even if it is huge. The way mind uses the concept of chance is so tricky that it has developed two opposed meanings in two different languages for the same exact word, one in english that means things are almost certain to happen and the other in french that mean they are almost impossible to happen. It wouldn't be too bad if the meaning wouldn't change depending on the circumstances, but it does. We can say in english that researchers make predictions while those predictions are only calculations, which means they are uncertain, and we can also say it in french. I don't know about other languages, but David must know since he knows many. It's too complicated and too confusing, we need to know what's going on in our mind about randomness.

[Le Repteux]

Consciousness is primarily (if not entirely) feelings - qualia. The sensation of blue is not resistance to anything.

If we look at a blue wall for a while without moving, we begin to see what is going on in our mind instead of seeing the wall, that's what we call meditation, but if something suddenly moves on the wall, we automatically get out of trance, because a change outside the brain is more important for our survival than a change inside it. All our sensations work like that: we don't feel the table anymore if we leave our hand on it for a while, we stop hearing a sound that doesn't change for a while, ...etc, so probably that all the qualia work like that too since they are all related to them. It is the same for the feelings that we have for others at the beginning of a relation: they all lose their intensity with time. Mind can forget about the things that don't change because it gets used to them, what enables it to get attracted by new things, and for human mind at least, that attraction often comes from itself, which means that those new things often happen inside our mind. I pretend that we see what we see because we only see the changes and that our mind resists to those changes like all the things that we know.