[wtbdqrs]

I appear to be reasoning at times but I have mostly no idea what I am talking about. I hit a bunch of words and concepts in the given context and thus kind of hallucinate sense.

Given a few months of peace of mind and enough money for good enough food, I could actually learn to reason without sounding like a confused babelarian.

Reasoning is mostly a human convention supported by human context that would have been a different one if the Fascists had won the war or the Soviet Union wouldn't have gotten corrupted.

But none of that has anything to do with pulling up a whiteboard to draw some flowcharts and run some numbers, all of which is why I am certain there is nothing the devs have "to fix". It took most reasonable humans many generations to learn stuff. Very few of us did the actual work.

It's all just a matter of time.

[voxic11]

Yeah, I think these chatbots are just too sure of themselves. They only really do "system 1 thinking" and only do "system 2 thinking" if you prompt them to. If I ask gpt-4o the riddle in this paper and tell it to assume its reasoning contains possible logical inconsistencies and to come up with reasons why that might be then it does correctly identify the problems with its initial answer and arrives at the correct one.

Here is my prompt:

I have a riddle for you. Please reason about possible assumptions you can make, and paths to find the answer to the question first. Remember this is a riddle so explore lateral thinking possibilities. Then run through some examples using concrete values. And only after doing that attempt to answer the question by reasoning step by step.

The riddle is "Alice has N brothers and she also has M sisters. How many sisters does Alice’s brother have?"

After you answer the riddle please review your answer assuming that you have made a logical inconsistency in each step and explain what that inconsistency is. Even if you think there is none do your best to confabulate a reason why it could be logically inconsistent.

Finally after you have done this re-examine your answer in light of these possible inconsistencies and give what you could consider a second best answer.

[cpleppert]

There isn't any evidence that models are doing any kind of "system 2 thinking" here. The model's response is guided by both the prompt and its current output so when you tell it to reason step by step the final answer is guided by its current output text. The second best answer is just something it came up with because you asked, the model has no second best answer to give. The second best answers always seem strange because the model doesn't know what it means to come up with a second best answer; it 'believes' the output it gave is the correct answer and helpfully tries to fulfill your request. Sometimes the second best answer is right but most of the time its completely nonsensical and there is no way to distinguish between the two. If you ask to choose it will be strongly influenced by the framing of its prior response and won't be able to spot logical errors.

Asking it to do lateral thinking and provide examples isn't really helpful because its final output is mostly driven by the step by step reasoning text, not by examples it has generated. At best, the examples are all wrong but it ignores that and spits out the right answer. At worst, it can become confused and give the wrong answer.

I've seen gpt-4 make all kinds of errors with prompts like this. Sometimes, all the reasoning is wrong but the answer is right and vice versa.

[wtbdqrs]

well, the tuning of training data results in at least some predictions that resemble varying models of systems 1 and 2 thinking. there is no reasoning at all. it's all models of reasoning, tokenized by opinionated taxonomical algorithms and degrees of systemic, academic/conventional human interpretation (tags) that are far from capturing the general human experience.

[daveguy]

> After you answer the riddle please review your answer assuming that you have made a logical inconsistency in each step and explain what that inconsistency is. Even if you think there is none do your best to confabulate a reason why it could be logically inconsistent.

LLMs are fundamentally incapable of following this instruction. It is still model inference, no matter how you prompt it.

[wtbdqrs]

isn't any instruction a subclass of inference? and doesn't any phrasing (lexicology) simply translate "down" to the heaviest values which, varying with the fine tuning, are the words that are, consensually and conventionally, the simplest ones that convey the meaning of the original word in the prompt, which should be the least ambivalent/least interpretable (again, fine tuning can broaden the scope) oneS. thus the LLM fulfills the "translated" instructions step by step and comes up with both or either the correct reasoning and answer.

details and technicalities, especially liminal ones, aren't as conventional and consensual as the name of the current set it is to be interpreted in.

so almost all mistakes of LLMs can be blamed on the lack of variety of human translations. multiple translations are only common for subtitles, mangas and manhwa as far as i know, or when some dude or dudette is proficient and passionate in two languages and reads a bad/weak translation of a (usually classic) novel. why the fuck would a human properly retranslate automated documentations or googles dev blog? or books on logic, in any science, books on art and aesthetics and whatnot. technical people don't need to care because, practically, there are no interpretations in algorithms and the rest of the code, except when a programming language does something weird on the (or someones) machine, which isn't that common by design.

[voxic11]

For me it did in fact follow the instruction and found the right answer. I have found this strategy of asking the LLM to analyze why its answer might be wrong results in better final answers. I think this allows the LLM to "use tokens to think" which allows the attention layers to gather the right cognitive resources to answer more subtitle questions effectively.

[daveguy]

You are confusing "providing the right answer" with "following the instruction". Cause vs correlation. From an algorithmic perspective it is literally doing the exact same thing regardless of prompt.

[zeknife]

If you had a prompt that reliably made the model perform better at all tasks, that would be useful. But if you have to manually tweak your prompts for every problem, and then manually verify that the answer is correct, that's not so useful.

[wtbdqrs]

the fact that you can manually tweak your prompts for any problem and agent, is still super useful (joke: *and the only reason our civilization still exists)

[wtbdqrs]

I'm not gonna read that book. I started and stopped after few chapters because it is based on and aims at manufacturing minds that follow game theory logic. Science (studies, reviews and application) got damaged quite a bit when too many people started following game theory logic.

We are, from our aware POV, a very young civilization.

And you only ever need game theory logic when you have to survive, got no thing and no skill to trade and you are too pathetic to move back in with your parents to work on your mind and or fuckability. Making money by ways of game theory logic compensates for all that but also diminishes the survival chance of the users' offspring to zero once super-unalligned AGIs start to assess the entire supply chain of wealth and how it impacts the evolution of human organisms and the ones inside them.