Working memory (WM) is often associated with intelligence and can supposedly be measured and trained by games like dual n-back. My question is if WM is task specific or has at least a task specific component. Or rather, my intuition is that it is: Me training dual-n back 20 minutes are day will improve my ability to play dual n-back far more than it will improve my ability to juggle different ways of asking this question in my head while writing it.
I think this is testable, so did anyone test if WM is task dependent and what did they find?
Edit to add:
This question with answers on the cross-discipline benefits was marked as a possible duplicate. My question is not addressed: The accepted answer to the question explores different ways cognitive abilities might 'transfer' from one discipline to another. One important aspect here is specific sub skills - psychologists may have a strong math background that may be useful when they learn. Another important aspect is the use of metaphors.
The supposed duplicate is relevant when discussing skill transfer between tasks. WM training is often hyped as general intelligence training. There are more straightforward ways of addressing this hype, my (maybe a bit roundabout) way is questioning wether WM trained on one task is as useful for another task.