Intuition or brute force?

[inocente]

What it tells us that what we call “intuition” is actually a “brute force” method, hidden deep down in the sub-conscious.

Unlikely. A brute force strategy enumerates through every possibility before selecting a solution. For example, in chess a brute force program evaluates every possible move by scoring every possible reply, then scoring every reply to every reply, and so on for many moves ahead. The human brain is too slow and error prone for that.

Intuition is the opposite strategy. Rather than methodically trying for the most correct decision, it uses past experience to make quick and dirty shortcuts based on probabilities.

We can see why intuition would evolve to aid survival while brute force didn’t. When hunting a deer, intuition gives us a hunch of how best to get close enough. Sometimes it works, sometimes not. But anyone who instead relied on brute force algorithms to catch their dinner would never survive to pass on their genes, as the deer would always be long gone before they finished methodically thinking through all the possibilities.

 

[Bagheera]

inocente:

A brute force strategy enumerates through every possibility before selecting a solution.

Strictly speaking, yes. But having a limited number of possibilities, one still must go through every one of them, before one can choose the “proper” one. Two examples.

Many years ago in a monthly chess magazine there was a problem, and the readers were encouraged to solve it, and write down the time it took them to find the solution. In the next monthly issue they presented the time needed by a chess program. We could make a comparison. Obviously the computer solved the problem much faster. But the funny thing is, that I immediately recognized the “theme” of the problem. It is called the “knight’s circle”, where only one of the eight possible moves was the solution. Nevertheless, I had to try each possible move, and it was the least likely move, which was the solution. So there was a “limited” brute force involved.

Another example was when I created a full blown chess problem when asleep. As I woke up I had both the problem and its solution in my head. It was all the result of my subconscious.

inocente:

The human brain is too slow and error prone for that.

The conscious part, yes. The sub-conscious, no. When conducting a pistol duel (for example), the one who uses the instinctive response has the advantage over the one who starts to draw first. This is the reason of the innumerable repetitions that fighters go through, to eliminate the need for the conscious response and let the sub-conscious take charge.

 

[polytropos]

Many years ago in a monthly chess magazine there was a problem, and the readers were encouraged to solve it, and write down the time it took them to find the solution. In the next monthly issue they presented the time needed by a chess program. We could make a comparison. Obviously the computer solved the problem much faster. But the funny thing is, that I immediately recognized the “theme” of the problem. It is called the “knight’s circle”, where only one of the eight possible moves was the solution. Nevertheless, I had to try each possible move, and it was the least likely move, which was the solution. So there was a “limited” brute force involved.

Another example was when I created a full blown chess problem when asleep. As I woke up I had both the problem and its solution in my head. It was all the result of my subconscious.

I don’t think anyone is denying that a human can take a brute force approach to chess. Obviously I can check each possible move for a piece. What everyone seems to be denying is that that is our primary approach when we do something that seems “intuitive.”

It is somewhat like the traveling salesman problem in computer science. Computers that take the brute force route take quite a long time to solve it, even for, say, n = 15, while I believe human adults can solve an n = 15 traveling salesman in a reasonable amount of time with to about 97% accuracy. But humans aren’t doing that because they are trying each of the options that a computer tries - the human brain does not have that much “processing power.” Humans can look at it intuitively and rule out the majority of possibilities immediately, without calculating them all subconsciously. Then when they have a “rough” solution, they can further use intuition to optimize it.

There are, of course, more efficient algorithms for traveling salesman that have been developed since the problem was first formulated. But those were the result of computer scientists collaborating to optimize a difficult problem - and, I think, we can hardly hold that the average adult “subconsciously” knows these more effective algorithms.

 

[MPat]

Strictly speaking, yes. But having a limited number of possibilities, one still must go through every one of them, before one can choose the “proper” one. Two examples.

Er, no, it doesn’t work that way. You were arguing that human brain uses the “brute force” in the sense in which the existing chess programs use “brute force” - that’s why you can conclude that human brain is in no way superior to the current computer.

It doesn’t work if the algorithm the brain uses is somehow “smarter”.

Many years ago in a monthly chess magazine there was a problem, and the readers were encouraged to solve it, and write down the time it took them to find the solution. In the next monthly issue they presented the time needed by a chess program. We could make a comparison. Obviously the computer solved the problem much faster. But the funny thing is, that I immediately recognized the “theme” of the problem. It is called the “knight’s circle”, where only one of the eight possible moves was the solution. Nevertheless, I had to try each possible move, and it was the least likely move, which was the solution. So there was a “limited” brute force involved.

Another example was when I created a full blown chess problem when asleep. As I woke up I had both the problem and its solution in my head. It was all the result of my subconscious.

Yes, we can perform “brute force” as well. But it doesn’t mean that we normally do so.

OK, let’s try this “endgame study”:

FEN string: 8/6k1/5pPp/4pP1P/3pP3/2pP4/2P1K3/3B4

White: Ke2, Bd1, p. c2, d3, e4, f5, g6, h5 (8)
Black: Kg7, p. c3, d4, e5, f6, h6 (6)
White to win (obviously).

How much time do you need to see the solution? (Very little, right?) Now, how much time would you need to perform full “breadth-first-search” and find all variations of required length? (Let me guess: you won’t have patience to do that. No, in this case it does not matter if patience is a virtue.)

So, what can you conclude from this experiment? I’d say you could conclude that you did not do a full “breadth-first-search” (otherwise the speed would have been comparable). So, what algorithm do you use? In this case you cannot just take your article of faith “Human mind works just as computers do.” and conclude something useful or interesting from it - that example was meant to support that “article of faith” in the first place. And if you do not know the actual algorithm, how can you support such claims…?

Now, if you had only tried to claim that anything human mind does could be simulated by a computer - in principle -, that would have been much harder to disprove… But you didn’t make such more modest claim… And now you have to revise the claim you made - or defend it…

 

[MPat]

I don’t think anyone is denying that a human can take a brute force approach to chess. Obviously I can check each possible move for a piece. What everyone seems to be denying is that that is our primary approach when we do something that seems “intuitive.”

It is somewhat like the traveling salesman problem in computer science. Computers that take the brute force route take quite a long time to solve it, even for, say, n = 15, while I believe human adults can solve an n = 15 traveling salesman in a reasonable amount of time with to about 97% accuracy. But humans aren’t doing that because they are trying each of the options that a computer tries - the human brain does not have that much “processing power.” Humans can look at it intuitively and rule out the majority of possibilities immediately, without calculating them all subconsciously. Then when they have a “rough” solution, they can further use intuition to optimize it.

There are, of course, more efficient algorithms for traveling salesman that have been developed since the problem was first formulated. But those were the result of computer scientists collaborating to optimize a difficult problem - and, I think, we can hardly hold that the average adult “subconsciously” knows these more effective algorithms.

It is probably a good place to add a link to a related game about traveling salesman problem: math.uwaterloo.ca/tsp/games/tspOnePlayer.html …

Unfortunately, sometimes it seems to be working badly…

 

[Bagheera]

polytropos:

I don’t think anyone is denying that a human can take a brute force approach to chess. Obviously I can check each possible move for a piece.

Certainly. But this is not what I am talking about. What we call “intuition” is actually our thinking process, considering and evaluating a whole lot of options, and coming to a conclusion – **before **that conclusion emerges from the subconscious into the conscious. That is what was substantiated by the experiment of monitoring the rapid eye-movement. The “eureka” moment is actually the result of a lot of subconscious analysis. That is all.

 

[polytropos]

What we call “intuition” is actually our thinking process, considering and evaluating a whole lot of options, and coming to a conclusion – **before **that conclusion emerges from the subconscious into the conscious.

But in the first example you gave in #22, you seemed to imply that you did manually and consciously consider each of the 8 possible moves:

It is called the “knight’s circle”, where only one of the eight possible moves was the solution. Nevertheless, I had to try each possible move, and it was the least likely move, which was the solution.

How does that support your larger point?

That is what was substantiated by the experiment of monitoring the rapid eye-movement. The “eureka” moment is actually the result of a lot of subconscious analysis. That is all.

Could you respond to MPat’s #9? It seemed pretty decisive but was never responded to. I am just not seeing the evidence that the subconscious is the seat of all these brute calculations.

I should add, regarding rapid-eye movement/saccades, I now recall from my psychology studies that when one looks at a complex painting, for instance, scientists could track the sequence of points where the subject looked. But presumably the subjects were not trying to “brute force” the painting, anymore than their rapid eye movements belie a brute force approach to chess.

 

[Bagheera]

polytropos:

But in the first example you gave in #22, you seemed to imply that you did manually and consciously consider each of the 8 possible moves

Yes, of course. I recognized the theme of the problem, but I had to investigate the possibilities. However, I did not go through the process of analyzing ALL the possible moves (and the problem had about 12-15 pieces). So the process was partly pattern recognition (based upon my previous exposure to many different types of problems) and partly a “brute force” of testing all the possible moves within that arrangement. Now, in a sense I could have “cheated” – but only a little. Based upon my prior knowledge, I was pretty sure that the “least likely” knight move will be the actual solution (which it was ). Quite frequently there are alternate solutions, so they also had to be analyzed.

Yet, what is a “pattern recognition”? Obviously, it is the result of subconsciously analyzing the position of the pieces. There were other possible patterns (it could have been an “excelsior” theme, too). Still, just by looking at the problem (without consciously analyzing it) I already “saw” the theme. There is no magical power of suggesting the theme out of nowhere.

Regarding the source, I read that article quite a long time ago, in a chess magazine. But the process was described in great detail, especially how the extremely rapid eye movement just preceded the moment of “enlightment”, the moment of finding the special, winning move. And it was not just an isolated experiment. Several players in several situations were involved.

polytropos:

I am just not seeing the evidence that the subconscious is the seat of all these brute calculations.

Where else? The “little grey cells” (to quote Poirot) are miniscule compared to the “white cells”. Are you aware the incredible amount of your memories? Are you aware of the process of how you recognize a face even from a hazy photograph? Of course not. As we grow older, we experience the very frustrating events of not being able to remember certain things… and then the memories emerge (if we are lucky). Our brain monitors and oversees an incredible amount of our bodily functions, and we are never aware it. The conscious part of our mind is like a ship on the waves of the ocean… it is on the top, but it is not in control (sorry for the poetic picture).

 

[MPat]

Yes, of course. I recognized the theme of the problem, but I had to investigate the possibilities. However, I did not go through the process of analyzing ALL the possible moves (and the problem had about 12-15 pieces). So the process was partly pattern recognition (based upon my previous exposure to many different types of problems) and partly a “brute force” of testing all the possible moves within that arrangement. Now, in a sense I could have “cheated” – but only a little. Based upon my prior knowledge, I was pretty sure that the “least likely” knight move will be the actual solution (which it was ). Quite frequently there are alternate solutions, so they also had to be analyzed.

Yet, what is a “pattern recognition”? Obviously, it is the result of subconsciously analyzing the position of the pieces. There were other possible patterns (it could have been an “excelsior” theme, too). Still, just by looking at the problem (without consciously analyzing it) I already “saw” the theme. There is no magical power of suggesting the theme out of nowhere.

Regarding the source, I read that article quite a long time ago, in a chess magazine. But the process was described in great detail, especially how the extremely rapid eye movement just preceded the moment of “enlightment”, the moment of finding the special, winning move. And it was not just an isolated experiment. Several players in several situations were involved.

Try to look at your original statement:

Of course we don’t know how the sub-conscious “works”, but these experiments start to shed some light on this phenomenon. What it tells us that what we call “intuition” is actually a “brute force” method, hidden deep down in the sub-conscious. So there is even less reason to feel “superior”. True, we are happy when we solve a difficult problem, or if we can beat a good opponent (and the computer is not), but that is just emotion, and not intelligence.

That specifically speaks about “brute force”, which is understood to exclude more advanced techniques like pattern recognition.

Thus maybe you shouldn’t be so harsh with “ad hominems” like this:

Further nonsense. Patience is simply waiting with making a fuss about a delay. But, hey! If everything else fails, do the usual tactics of the incompetent – redefine the issue. Goodbye.

The question if “patience” is to be defined as a type of virtue is, at the very least, less clear than question if “brute force” is to be defined to include pattern recognition…

Where else? The “little grey cells” (to quote Poirot) are miniscule compared to the “white cells”. Are you aware the incredible amount of your memories? Are you aware of the process of how you recognize a face even from a hazy photograph? Of course not. As we grow older, we experience the very frustrating events of not being able to remember certain things… and then the memories emerge (if we are lucky). Our brain monitors and oversees an incredible amount of our bodily functions, and we are never aware it. The conscious part of our mind is like a ship on the waves of the ocean… it is on the top, but it is not in control (sorry for the poetic picture).

As I have said, “unconscious of the gaps”…

You know, sometimes some atheists claim that science must ignore all supernatural explanations, because “God did it.” does not explain much. Well, “Unconscious did it.” explains even less. You say that “intuition” does not happen consciously, that we are not aware how it happens…? Well, sure. If it was done consciously, we wouldn’t call it “intuition”. But your claim that those “white cells” do “brute force” calculations of variations and nothing else (for that is what “what we call “intuition” is actually a “brute force” method, hidden deep down in the sub-conscious” means - if it means anything) is very suspicious - and unsupported by any actual evidence. You haven’t even given us a way to find the study that supposedly shows something like that…

 

[Jaaanosik]

Certainly. But this is not what I am talking about. What we call “intuition” is actually our thinking process, considering and evaluating a whole lot of options, and coming to a conclusion – **before **that conclusion emerges from the subconscious into the conscious. That is what was substantiated by the experiment of monitoring the rapid eye-movement. The “eureka” moment is actually the result of a lot of subconscious analysis. That is all.

You don’t know what you are talking about.
A chess player’s subconsciousness does not play chess.
What experiment you are talking about?

 

[AndrewW94]

1. Just as a side note, computers are not unequivocally better than humans at chess. Humans are better at correspondence chess.
2. Humans and computers are very different. Humans understand concepts, computers are just complex arrangements of rocks. Nothing more. Rocks cannot be arranged in such a way as to have an understanding.
3. The eyemovement thing is something I’ve noticed while I play chess (I noticed it, so at least some of it isn’t subconscious). It doesn’t show that humans really take a brute force approach to chess. They recognize patterns. (Which is why I can remember chess games I’ve played almost perfectly for some period after they’re done).

I’m not entirely sure what you mean by intuition, but I am sure that computers don’t think. Computers don’t grasp concepts. That is a very major difference.

 

[polytropos]

Regarding the source, I read that article quite a long time ago, in a chess magazine. But the process was described in great detail, especially how the extremely rapid eye movement just preceded the moment of “enlightment”, the moment of finding the special, winning move. And it was not just an isolated experiment. Several players in several situations were involved.

This is all well and good, but I remember specifically from my psychology textbook that such eye movements present when one examines a nice painting. A chessboard is complex and has a bunch of pieces all over the place. What you are citing does not even rule out the possibility that the eye movements are just a result of visual perception.

But even if it did…

Are you aware the incredible amount of your memories? Are you aware of the process of how you recognize a face even from a hazy photograph? Of course not. As we grow older, we experience the very frustrating events of not being able to remember certain things… and then the memories emerge (if we are lucky). Our brain monitors and oversees an incredible amount of our bodily functions, and we are never aware it. The conscious part of our mind is like a ship on the waves of the ocean… it is on the top, but it is not in control (sorry for the poetic picture).

Yes, I am aware that some mental activities are subconscious. That doesn’t justify the inference that intuition consists in subconscious brute force calculations. It doesn’t help that you haven’t answered the most substantial critiques. I don’t think you’ve responded to any of MPat’s posts.

Even inocente, who is (I believe) a materialist, doesn’t think so. There’s just no evidence for it.

 

[inocente]

Strictly speaking, yes. But having a limited number of possibilities, one still must go through every one of them, before one can choose the “proper” one. Two examples.

Many years ago in a monthly chess magazine there was a problem, and the readers were encouraged to solve it, and write down the time it took them to find the solution. In the next monthly issue they presented the time needed by a chess program. We could make a comparison. Obviously the computer solved the problem much faster. But the funny thing is, that I immediately recognized the “theme” of the problem. It is called the “knight’s circle”, where only one of the eight possible moves was the solution. Nevertheless, I had to try each possible move, and it was the least likely move, which was the solution. So there was a “limited” brute force involved.

Another example was when I created a full blown chess problem when asleep. As I woke up I had both the problem and its solution in my head. It was all the result of my subconscious.

Be careful of getting carried away by introspection . Objective evidence that intuition uses brute force would include a correlation in the types and quantity of decision errors, but I predict that intuition gives a different spread from any kind of analytic thought process. Also according to you MRI scans should light up like Christmas trees as the heart races to cool all those neurons, while I predict you’ll be sitting in the dark.

The conscious part, yes. The sub-conscious, no. When conducting a pistol duel (for example), the one who uses the instinctive response has the advantage over the one who starts to draw first. This is the reason of the innumerable repetitions that fighters go through, to eliminate the need for the conscious response and let the sub-conscious take charge.

Instinct isn’t intuition. To commit the heinous sin of making comparisons with computers: instinct uses built-in logic which is very fast, but the logic only develops over evolutionary timescales. Intuition is, as it were, a learned instinct, slower but custom made (learned) and so very useful. Conscious thought is glacially slow, it’s the programmer needed to learn intuitions (amongst other stuff).

(Btw is the subconscious special? To quote Wikipedia “In psychology, the subconscious is the part of consciousness that is not currently in focal awareness”).

I think also your interest in chess is misleading you as it’s artificial. Try writing a brute force algorithm for how to get close to a deer without spooking it and you’ll see there are far too many fuzzy things going on. Then consider how such a methodical, analytical process could possibly evolve. My intuition is you’ll not manage that with or without brute force.