Comment by kshri24
9 hours ago
Determinism is not strictly anti-randomness (though I can see why one can confuse it to be polar opposites). Rather we do not even have true randomness (at least not proven) and should actually be called pseudorandom. Determinism just means that if you have the same input parameters (considering all parameters have been accounted for), you will get the same result. In other words, you can start with a particular random seed (pseudorandom seed to be precise) and always end up with the same end result and that would be considered deterministic.
> You haven't done a lot of gardening if you don't know plants
I grow "herbs".
> there's a biological explanation
Exactly. There is always an explanation for every phenomena that occurs in this observable, physical World. There is a defined cause and effect. Even if it "feels random". That's not how it is with LLMs. Because in between your deterministic input parameters and the output that is generated, there is a black box: the model itself. You have no access to the billions of parameters within the models which means you are not sure you can always reproduce the output. That black box is what causes non-determinism.
EDIT: just wanted to add - "attacked by parasites all the time", is why I said if you have control over the environment. Controlling environment encompasses dealing with parasites as well. Think of well-controlled environment like a lab.
Do you think LLMs sidestep cause and effect somehow ? There's an explanation there too, we just don't know it, But that's the case for many natural phenomena.
I am not saying LLM sidesteps cause-effect. I am saying it is a black box. So yes "we just don't know it" is basically describing a black box.