In today's @AcademicRenewal article, Thomas K. Lindsay of @TPPF shows how "UT Austin's new core curriculum puts history, literature, civic education, and the enduring questions of liberal education back at the center." jamesgmartin.center/2026/09/…

It's the first week of classes here at NYU! Some scattered lukewarm takes on AI and higher ed right now (some philosophy-specific): UG Assessment: all assessment that is not intended to test ability to use AI should now be in person, whether that's a written test or an oral exam. If you can't staff an oral exam, and you don't want to test AI use, use a written exam. Anything else is a disservice to the students who don't cheat. I know, it's annoying! We had a different practice before. But that ship has sailed. UG Applications: this seems like a disaster, and I don't know what to say. High schools haven't generally caught up to the need for all in-class assessment and assessment was anyway a mess there. We now also can't trust application materials. It's not a popular view, but in person standardized tests look like they're one of the few things that can carry signal (without relying on a more biased proxy like "we know this school and these teachers"). Journal reviewing: I expect journal submissions to grow *a lot* in the near future. We are already over capacity for referees. We need to develop new ideas here, probably it involves AI desk-rejections. This is harder in our field than in say math where you can at least check correctness as a proxy. But we need something; the system won't be able to tolerate a lot more submissions. Open access/working papers. I'm in an unusual and obviously somewhat biased position, but I would like AI companies to train on my scholarly work. It's how my work will reach other scholars and it's how it will impact the ideas of the future. Others may feel differently. Your thinking on this should heavily impact what journals you submit to. Citation / influence. Related to the above, many fields use google scholar as a proxy for impact in the field. I expect cite counts to go up a lot (even more than they did already) as AI models make it easier to "cite everyone" by automating bibliographic entry. It wasn't clear before that cite count was a great proxy for influence; this will become even less clear going forward. I don't have a positive solution but I think we need new ideas. Recommendation letters/promotion letters. Summaries of work no longer carry the signal of effort they once did. We should not expect that everyone will "just be honest" with their letters. Everyone is too busy. This system already needed rethinking, even more so now. Importance of scholarly community and virtues. Now more than ever we should value the virtues of scholarship and scholarly community. Rigor and seriousness about intellectual work are going to be threatened even more than they were before as it becomes easier to produce a large volume of reasonable-looking work. In my opinion, one of the most important things we can share with our students is modeling for them what it looks like to be conscientious about details even when no one else will check, and what it looks like to care about the truth and open inquiry rather than reputation. The university is distinctively well-positioned to foster these ideals, and they will still be of great value to our student in many of the wildly different futures they may live through.