Knowledge is not justified, true belief. What is it?
David Deutsch has said that knowledge is, loosely speaking, useful information.
I propose, instead, to think of knowledge as problem-solving information.
Knowledge is information adapted to a purpose. In other words, it solves a problem.
Knowledge is information with the appearance of design (for a purpose). In other words, it solves a problem.
What kind of information is useful? Information which solves some problem worth solving.
Messages (32)
are problems knowledge?
Understanding of problems is knowledge.
How does that fit in with knowledge being problem-solving information?
Understanding problems helps you solve them!
So problem-solving info is useful to first understand a problem, and then once you have that knowledge you create more problem-solving info to solve the problem. So in this case (and also is all cases?), there are 2+ different problem situations.
OK, the first problem to solve is to explain what the problem is. The problem-solving info that solves this problem is knowledge of the problem. Then you use this info to help try to solve the problem. This may then transform your understanding of the problem.
> This may then transform your understanding of the problem.
No, a better understanding of a problem DOES change your understanding of that problem, not MAY.
I wrote "may" because you might fail to solve the problem. I didn't think that failures would always improve your understanding of what the problem is. But you are saying they do right?
You wrote
> The problem-solving info that solves this problem is knowledge of the problem.
You said it's a solution, not a failure, that you were talking about.
The "this" in my sentence refers to the problem of explaining what the problem is - what I said was the "first problem". Explaining what the problem is is different to solving it. One can explain what the problem is, then try to solve the problem, and fail. The failure might lead (always leads?) to an improved understanding of what the problem is (the first problem).
So there is the problem of explaining/understanding what the problem is and the problem itself.
> OK, the first problem to solve is to explain what the problem is. The problem-solving info that solves this problem is knowledge of the problem. Then you use this info to help try to solve the problem. This may then transform your understanding of the problem.
I think the source of the misunderstanding was my "this" in the last sentence. It is ambiguous. I intended it to refer to the trying to solve the problem in the preceding sentence and not the problem-solving info that enables you to explain what the problem is.
So I didn't write clearly enough. Is it clearer now?
Erin: you said there are always (possibly) 2+ problem situations. There is the generic problem of explaining what the problem is (or, maybe better, why there is a problem) and then there is the actual problem itself. What are the other problem situations indicated by the "+"?
It would be clearer to write out what you meant instead of adding modifications to a previous statement. It'd also add clarity to mention what problem you're trying to solve and how.
About 2+: you can divide problems in pretty much whatever number of sub-problems you want to. The number isn't important. What's important is whether the division is useful.
Your post said that knowledge is problem-solving info. The problem I had was I wasn't clear how info about what the problem is constitutes knowledge in your way of thinking. For that info does not solve the problem. But it is useful information and so knowledge by Deutsch's way of thinking.
The answer to that is:
Problems must be understood in order to solve them. That is, you must be able to explain what the problem is. This is a different problem to solving the problem. Explaining what the problem is requires creating problem-solving information. That problem-solving info is knowledge of the problem. So that is how problems themselves fit into your way of thinking of knowledge.
The further point was that problem-solving info of what the problem is helps you to try to solve the problem. If you are successful, then you will not only have knowledge of the solution but also better knowledge of what the problem is. If you are unsuccessful, you still may nevertheless end up with better knowledge of what the problem is. That's still a win because better problem-solving info of what the problem is will stand you in good stead for eventually solving the problem.
About 2+: Understand about the sub-problems but wanted to clarify if Erin meant anything else.
> If you are successful, then you will not only have knowledge of the solution but also better knowledge of what the problem is.
This now seems wrong having thought abt it more. If you solve the problem, then the problem goes away. It's not that you have better knowledge of what the problem is, it's that you can explain why there is no longer a problem and what the solution was.
Does that sound right?
> David Deutsch has said that knowledge is, loosely speaking, useful information.
Source?
In The Beginning of Infinity, Deutsch said
"... knowledge is information which, when it is physically embodied in a suitable environment, tends to cause itself to remain so."
Can this be read, loosely, as useful information, or as problem-solving information?
Source: he said it to me. (many times)
Ok, cool.
What do you think of the defn above? It seems somewhat different.
What causes the information to remain there? It's serving some purpose, solving some problem.
> This is a different problem to solving the problem.
Statements like this are too confusing. And then when you use the word "problem" after, I don't know which problem you mean. If you want to talk about two different problems, it'd help to label them in some clear way using different words.
The distinction being drawn between different types of problem solving information is arbitrary and counter-productive. Information either contributes to solving a particular problem, or doesn't.
And a full solution isn't required for problem solving information to exist. Partial solutions involve problem solving power, they count.
Let P1 be a problem.
To solve P1, you must understand what P1 is.
This itself is a problem. Call it P2.
The solution of P2 is not the same as the solution of P1. But they are connected. If you get the solution to P2 wrong, you will fail to solve P1 because you do not understand what P1 is.
P1 and P2 are distinct problems. P2 is present in any problem situation. That is, to solve a problem you must know what the problem is.
To solve P2 and P1 you need to create problem-solving info. The info created from solving P2 helps you try to solve P1. And the info from trying to solve P1 can change your solution to P2.
There is no claim here that a full solution to P2 and P1 is required for problem-solving info to exist.
There is no claim here that info from solving P2 does not contribute to solving P1.
There is no claim here that the problem-solving info in both cases are different types or independent, only that the info solves different problems.
There is a claim that P1 and P2 are non-arbitrary. And there is a claim that understanding what P1 is is not counterproductive.
Also, unless I am mistaken, P1 and P2 are what Erin was getting at in her 2+ comment.
Your last comment makes me confused abt ur position. Erin and I appeared to take the same thing from your 4th comment above. See comments 5 and 6. But it appears we, or at least I, misunderstood (I might have misunderstood Erin). What did you mean when you said "understanding problems" in your 4th comment? I took it to mean being able to explain what the problem is.
> Your last comment makes me confused abt ur position. Erin and I appeared to take the same thing from your 4th comment above. See comments 5 and 6. But it appears we, or at least I, misunderstood (I might have misunderstood Erin). What did you mean when you said "understanding problems" in your 4th comment? I took it to mean being able to explain what the problem is.
You can link to comments and/or quote stuff. I'm not especially motivated to reread everything based on a few vague hints, in order to better help you learn. I looked back some, but I'm unclear on what you're confused about.
It would be easier to do this in email with proper quoting.
> Let P1 be a problem.
>
> To solve P1, you must understand what P1 is.
No. It's sometimes (often) helpful, not necessary.
Re. My confusion, you said:
> Understanding problems helps you solve them!
You seem to be distinguishing between P1 (the problem) and P2 (understanding what the problem is). In my subsequent expansions of your statement where I thought I was agreeing with you and your general idea that knowledge is problem-solving info, you seemed not to know what I was talking about. And disagree. So I appear to have read you wrong.
Ok, it is too strong a statement to say that in order to solve a problem you must understand the problem.
But you must know something about the problem to solve it. Otherwise you don't even know there is a problem. So let P2 be the problem of knowing something about P1.
Ok, it is too strong a statement to say that in order to solve a problem you must understand the problem.
But you must know something about the problem to solve it. Otherwise you don't even know there is a problem. So let P2 be the problem of knowing something about P1.
> But you must know something about the problem to solve it. Otherwise you don't even know there is a problem.
ah, silly me - I retract that. wasn't thinking abt evolution. kick my P1 and P2 down the road.
so: knowledge about a problem can help you solve a problem and is problem-solving info but it isn't necessary to solve the problem.
FYI people routinely solve problems they don't know about because they are doing something with REACH. They are trying to solve problem X, but their method is principled, so in the process they end up solving problems Y and Z, without ever knowing about those.
>About 2+: Understand about the sub-problems but wanted to clarify if Erin meant anything else.
nothing else. i was thinking about sub-problems when I wrote the stuff about the 2+.
@ knowledge as problem-solving info: would it be better to think of knowledge as "info which contributes towards solving a problem"?
like u may not have enuf knowledge to actually solve a problem. so if someone is thinking of knowledge as "problem-solving info", then is it misleading cuz the knowledge may not be enuf to actually *solve* the problem?
or it could be that the knowledge you do have is there because it actually *was* enuf to solve past problems, just mb not enuf to solve the current problem you are working on. so in that case, the knowledge is correctly thought of as problem-solving info cuz it:
1) was used in the past to solve a problem and 2) is useful right now by contributing towards solving the current prob you are working on
Partial solutions that are valuable always solve sub-problems (which are problems).
This is straightforward because getting/understanding anything useful can be seen as a problem/goal itself. How to make partial progress in some useful way on some worthwhile problem is itself a good problem to try to solve.
Maybe it would be helpful to give an example of something that, in your explanation, would not be knowledge? It sounds to me like everything can be called knowledge, because if it fails to solve the main problem, it solves a minor problem, that of its inability to solve the main problem has solved another problem, which is the problem of whether this proposed solution solves the problem.
I hope my explanation is sufficiently clear.
Lots of stuff doesn't solve any problem anyone cares about. Many of those problems also are objectively unimportant. Solving an arbitrary problem – one that neither exists in reality nor has any other importance – is not valuable or important, and should not be called "knowledge" in general (it'd be misleading to use the term "knowledge" to refer to junk).