Discussion about this post

User's avatar
Jacob Felson's avatar

Your analysis of how AGI might reshape academic research raises important points, though I find several of your assertions problematic.

First, your claim that "AGI systems launching now can reason, learn, and solve problems across all domains, at or above human level" appears to contradict the consensus among AI researchers themselves, who broadly agree that true AGI has not yet been achieved.

Second, while you suggest that professors who cannot surpass AGI in generating new knowledge may become obsolete, this overlooks the fundamental reality of higher education: most institutions prioritize teaching over research. Consider that despite the widespread availability of high-quality recorded lectures online, students continue to enroll in traditional colleges and attend live classes in significant numbers. This persistence of in-person education—even when potentially superior alternatives are freely available—raises intriguing questions about the nature of effective teaching that predate any AGI considerations.

Additionally, your argument would benefit from clearly distinguishing between research universities and other types of higher education institutions. There's a common tendency among research faculty to universalize their experience, assuming all professors focus primarily on research and all students are traditional full-time undergraduates living on campus. This oversimplification undermines what could otherwise be a more nuanced analysis of AGI's potential impact on different educational contexts.

Expand full comment
Kash's avatar

I agree that some adjustments are needed, particularly with traditional assignments, but lecture courses have been replaceable for a long time. Schools - especially elite schools - are providing signaling and networking

Expand full comment
76 more comments...

No posts