It seems like a flavour of the rubber duck method; by trying to explain it to a third party, you think about it in a different way and find a solution.
Trust me bro(ette): Rubber duck is the SHIT. I don’t even program save for a few rare instances, but any complex issue where you just know something is wrong but can’t quite put your finger on it? It works miracles. A lot better tbf if you are actually explaining it to someone who can ask questions, but any object that you can look at is a good substitute.
I think it’s a bit more than that. I think that the idea is that you simplify the problem so that the rubber duck could understand it. Or at least reformulate it in order to communicate it clearly.
It’s the simplification, reformulation or reorganisation that helps to get the breakthrough.
Just thinking out loud isn’t quite the same thing.
Even though this is true for like 90% of my thinking (that I can see when I try), so far I’m concinced this ist because I am a predominantly language-and-normal-grammar-rules thinker.
There are people that mostly think via associations of words that don’t have to be formulated/ cast into grammar.
And then there supposedly people mainly thinking in pictures or smth, without words.
Anyways for some people rubber duck mode reoresents a change in thinking method, I think
Yes, saying thinks out loud requires a different change in thinking because you are verbalizing the thoughts in addition to approaching it as an explanation instead of just an understanding. I know how a phone works, but describing how it works is a different thing from knowing. The duck is just a stand in for someone else to get the mindset of explaining
I’m one of the latter that doesn’t really think in words, and a LOT of the time, thoughts have to be greatly simplified or at least much more organized to be stated in clear sentences. It’s that pause-and-refine that often gets the breakthrough for me. Sometimes it takes clear until I’m trying to put it in understandable sentences instead of a big ramble, but it still largely boils down to ACTUALLY stopping the task work to loop back over the landscape.
A lot of people do the same thing physically. Like when you’re climbing a big ladder and suddenly realize how high up you are, or how unstable the ladder is. Just a pause and broadening of attention is often enough to cue different thoughts and realizations.
AI in general is only a glorified rubber duck for most cases. The amount of bullshit cobbled together is too high for many uses
I’ve been using it like that. I have been trying to program this macropad thing I bought that uses python without having done much programming and it has yet to give me a solution that works. But in the course of explaining to it why whatever it gave me doesn’t work I’ve made a lot of progress so that’s nice at least.
