Assume this is talking about majoring in Computer Science.  Do you think getting high GPA (in CS major) is highly, moderately or only slightly aligned with improving AI risks research(both conceptual and emprirical) abilities? Some said they think a lot of college courses are not so useful[1], and getting good grades on tests requires practcing irrelevant skills and cramming stuffs(such as Brain Tomasik writing:"Drawbacks to Formal Education"[2]). 

However, some also think most things taught in CS courses are valuable, rigorous training is beneficial to thinking skills(even if the knowledge in course itslef is not related to AI risks). I don't know what theory is more correct. Do you think chasing for great GPA(define it as getting top 20% GPA　in a top 100 university) is itself benefial, or is it just mostly for credentials and signaling?

[1]https://forum.effectivealtruism.org/posts/hgiLaE3eL76ovcfdH/many-undergrads-should-take-light-courseloads

[2]https://briantomasik.com/drawbacks-formal-education/