6 ms·

The word "reason" came into this thread with the original comment: 3. LLMs are bad at solving reasoning problems. 4. Prolog is good at solving reasoni

The word "reason" came into this thread with the original comment: 3. LLMs are bad at solving reasoning problems. 4. Prolog is good at solving reasoning problems. I agree with you. In Prolog "?- 1=1." is reasoning by definition. Then 4. becomes "LLMs should emit Prolog because Prolog is good at executing Prolog code". I think that's not a useful place to be, so I was trying to head off going there. But now I'll go with you - I agree it IS reasoning - can you please support your case that "executing Prolog code is reasoning" makes Prolog more useful for LLMs to emit than Python?

YeGoblynQueenne 11mo ago
This is not my claim: >> "executing Prolog code is reasoning" makes Prolog more useful for LLMs to emit than Python? I said what I think about LLMs generating Prolog here: https://news.ycombinator.com/item?id=45914587 https://news.ycombinator.com/item?id=45914587 But I was mainly asking why you say that Prolog's execution is "not reasoning". I don't understand what you mean that '"?- 1=1." is reasoning by definition' and how that ties-in with our discussion about Prolog reasoning or not.
- jodrellblank 11mo ago
  "?- 1=1." is Prolog code. Executing Prolog code is reasoning. Therefore that is reasoning. Q.E.D. This is the point you refused to move on from until I agreed. So I agreed. So we could get back to the interesting topic. A topic you had no interest in, only interest dragging onto a trangent and grinding it down to make ... what point, exactly? If "executing Prolog code" is reasoning, then what? I say it isn't useful to call it reasoning (in the context of this thread) because it's too broad to be a helpful definition, basically everything is reasoning, and almost nothing is not. When I tried to say in advance that this wouldn't be a useful direction and I didn't want to go here, you said it was " not a great way to have a discussion". And now having dragged me off onto this academic tangent, you dismiss it as "I wasn't interested in that other topic anyway". Annoying.
  - YeGoblynQueenne 11mo ago
    I'm sorry you find my contribution to the discussion annoying, but how should I feel if you just "agree" with me as a way to get me to stop arguing? But I think your annoyance may be caused by misunderstanding my argument. For example: >> If "executing Prolog code" is reasoning, then what? I say it isn't useful to call it reasoning (in the context of this thread) because it's too broad to be a helpful definition, basically everything is reasoning, and almost nothing is not. Everything is not reasoning, nor is executing any code reasoning, but "executing Prolog code" is, because executing Prolog code is a special case of executing code. The reason for that is that Prolog's interpreter is an automated theorem prover, therefore executing Prolog code is carrying out a proof; in an entirely literal and practical sense, and not in any theoretical or abstract sense. And it is very hard to see how carrying out a proof automatically is "not reasoning". I made this point in my first comment under yours, here: https://news.ycombinator.com/item?id=45909159 https://news.ycombinator.com/item?id=45909159 The same clearly does not apply to Python, because its interpreter is not an automated theorem prover; it doesn't apply to javascript because its interpreter is not an automated theorem prover; it doesn't apply to C because its compiler is not an automated theorem prover; and so on, and so forth. Executing code in any of those languages is not reasoning, except in the most abstract and, well, academic, sense, e.g. in the context of the Curry-Howard correspondence. But not in the practical, down-to-brass-tacks way it is in Prolog. Calling what Prolog does reasoning is not a definition of reasoning that's too broad to be useful, as you say. On the contrary, it's a very precise definition of reasoning that applies to Prolog but not to most other programming languages. I think you misunderstand this argument and as a consequence fail to engage with it and then dismiss it as irrelevant because you misunderstand it. I think you should really try to understand it, because it's obvious you have some strong views on Prolog which are not correct, and you might have the chance to correct them. I absolutely have an interest in any claim that generating Prolog code with LLMs will fix LLMs' inability to reason. Prolog is a major part of my programming work and research.
    - jodrellblank 11mo ago
      > "how should I feel if you just "agree" with me as a way to get me to stop arguing?" Triumphant? Victorious? magnificent, successful, proud, powerful, insert any adjective which applies to a situation where someone wanted something, and then got it. > "And it is very hard to see how carrying out a proof automatically is "not reasoning. The same clearly does not apply to Python, because its interpreter is not an automated theorem prover; it doesn't apply to javascript because its interpreter is not an automated theorem prover" And that does not stop Python or Javascript from being used to find solutions to e.g. an Einstein Puzzle, something a human might call "a reasoning problem". This means Prolog 'doing reasoning' must not be the thing which solves the 'reasoning problem', something else must be doing that because non-reasoning systems can do it too. If Prolog 'doing reasoning' meant it could solve 'reasoning problems' that no other programming language could, that would be a strong reason to use Prolog, but that is not something you or the other 'reasoning' commenters have claimed or offered examples of. Clearly the word 'reasoning' has different definitions in the different sentences and that is important here because I am responding to one and youall on the other. If 'doing reasoning' is not the thing which makes it useful for 'solving reasoning problems' - if that neither compels one to use Prolog when working to 'solve a reasoning problem', nor convinces one to avoid other languages - if the definition does not influence one's decision in any way - it's very hard to see how it is the relevant version of 'reasoning' to focus on, and what point is trying to be made by this insistence on focusing on it, except academic one-upping.

YeGoblynQueenne 11mo ago

This is not my claim: >> "executing Prolog code is reasoning" makes Prolog more useful for LLMs to emit than Python? I said what I think about LLMs generating Prolog here: https://news.ycombinator.com/item?id=45914587 https://news.ycombinator.com/item?id=45914587 But I was mainly asking why you say that Prolog's execution is "not reasoning". I don't understand what you mean that '"?- 1=1." is reasoning by definition' and how that ties-in with our discussion about Prolog reasoning or not.

jodrellblank 11mo ago
"?- 1=1." is Prolog code. Executing Prolog code is reasoning. Therefore that is reasoning. Q.E.D. This is the point you refused to move on from until I agreed. So I agreed. So we could get back to the interesting topic. A topic you had no interest in, only interest dragging onto a trangent and grinding it down to make ... what point, exactly? If "executing Prolog code" is reasoning, then what? I say it isn't useful to call it reasoning (in the context of this thread) because it's too broad to be a helpful definition, basically everything is reasoning, and almost nothing is not. When I tried to say in advance that this wouldn't be a useful direction and I didn't want to go here, you said it was " not a great way to have a discussion". And now having dragged me off onto this academic tangent, you dismiss it as "I wasn't interested in that other topic anyway". Annoying.
- YeGoblynQueenne 11mo ago
  I'm sorry you find my contribution to the discussion annoying, but how should I feel if you just "agree" with me as a way to get me to stop arguing? But I think your annoyance may be caused by misunderstanding my argument. For example: >> If "executing Prolog code" is reasoning, then what? I say it isn't useful to call it reasoning (in the context of this thread) because it's too broad to be a helpful definition, basically everything is reasoning, and almost nothing is not. Everything is not reasoning, nor is executing any code reasoning, but "executing Prolog code" is, because executing Prolog code is a special case of executing code. The reason for that is that Prolog's interpreter is an automated theorem prover, therefore executing Prolog code is carrying out a proof; in an entirely literal and practical sense, and not in any theoretical or abstract sense. And it is very hard to see how carrying out a proof automatically is "not reasoning". I made this point in my first comment under yours, here: https://news.ycombinator.com/item?id=45909159 https://news.ycombinator.com/item?id=45909159 The same clearly does not apply to Python, because its interpreter is not an automated theorem prover; it doesn't apply to javascript because its interpreter is not an automated theorem prover; it doesn't apply to C because its compiler is not an automated theorem prover; and so on, and so forth. Executing code in any of those languages is not reasoning, except in the most abstract and, well, academic, sense, e.g. in the context of the Curry-Howard correspondence. But not in the practical, down-to-brass-tacks way it is in Prolog. Calling what Prolog does reasoning is not a definition of reasoning that's too broad to be useful, as you say. On the contrary, it's a very precise definition of reasoning that applies to Prolog but not to most other programming languages. I think you misunderstand this argument and as a consequence fail to engage with it and then dismiss it as irrelevant because you misunderstand it. I think you should really try to understand it, because it's obvious you have some strong views on Prolog which are not correct, and you might have the chance to correct them. I absolutely have an interest in any claim that generating Prolog code with LLMs will fix LLMs' inability to reason. Prolog is a major part of my programming work and research.
  - jodrellblank 11mo ago
    > "how should I feel if you just "agree" with me as a way to get me to stop arguing?" Triumphant? Victorious? magnificent, successful, proud, powerful, insert any adjective which applies to a situation where someone wanted something, and then got it. > "And it is very hard to see how carrying out a proof automatically is "not reasoning. The same clearly does not apply to Python, because its interpreter is not an automated theorem prover; it doesn't apply to javascript because its interpreter is not an automated theorem prover" And that does not stop Python or Javascript from being used to find solutions to e.g. an Einstein Puzzle, something a human might call "a reasoning problem". This means Prolog 'doing reasoning' must not be the thing which solves the 'reasoning problem', something else must be doing that because non-reasoning systems can do it too. If Prolog 'doing reasoning' meant it could solve 'reasoning problems' that no other programming language could, that would be a strong reason to use Prolog, but that is not something you or the other 'reasoning' commenters have claimed or offered examples of. Clearly the word 'reasoning' has different definitions in the different sentences and that is important here because I am responding to one and youall on the other. If 'doing reasoning' is not the thing which makes it useful for 'solving reasoning problems' - if that neither compels one to use Prolog when working to 'solve a reasoning problem', nor convinces one to avoid other languages - if the definition does not influence one's decision in any way - it's very hard to see how it is the relevant version of 'reasoning' to focus on, and what point is trying to be made by this insistence on focusing on it, except academic one-upping.