> still don't want to accept the reality that their research has become obsolete.

I see it that most of legacy research is not obsolete but can be significantly amplified by AI.