It really depends on the type of research. Writing, in artificial intelligence, can be especially useful when it’s still unclear whether your research is reason

It really depends on the type of research. Writing, in artificial intelligence, can be especially useful when it’s still unclear whether your research is reasonably implementable. Machine learning, for example, was mostly theoretical until the engineering issues were resolved.