I don't get what you mean by "unknown reasons", we understand that counting tokens requires a type of introspection transformer models can't do while operating

I don't get what you mean by "unknown reasons", we understand that counting tokens requires a type of introspection transformer models can't do while operating on tokens. What I described is tried, and works, but the models are still not cheap/fast/reliable enough to always do what I described for every query. The difference between what I described and directly asking the model to count is that we know the models can get cheaper, faster, and more reliable at what I described without any earth shattering discoveries Like I don't see any reason why GPT 10 will ever be able to count how many letters there are in the word strawberry without a complete paradigm shift in model building... but going from GPT 3 to GPT 4 we already got a model that can always write the dead simple code required to count it out, and the models that can do so are already getting cheaper and faster every few months without any crazy discoveries.