Never let a good crisis go to waste. That is the adage that universities should take to heart as they confront the threats of the Trump administration. Because the reality is that universities have been lost for decades and require dramatic changes if they want to survive at all in the future.

Here are a series of major issues that universities urgently need to address:

1. Frontal teaching, high stakes testing, even traditional essays are all outmoded. The rapid progress of AI is making many of these obsolete. Universities urgently need to figure out what learning in an AI world looks like. I suspect that one-on-one tutor meetings as in the Oxford model will have a big renaissance (where some of the tutors may be AIs).

2. Students as customers is fundamentally a bad model. So is donors as customers. Universities need to have the pursuit and sharing of knowledge as their central value and everything else must follow from there.

3. Administrations have become absurdly bloated and self-serving. Universities need to cut back to the minimal support required to enable professors and students to pursue their missions.

4. Research is still often locked up in expensive academic journals and much of it doesn’t reproduce with a growing number of cases of outright fraud. Universities need to pare back dramatically on what research is done in the first place, publish in open access, require the publication of data, code, experimental setup and foster reproduction work.

5. Way too much research is incremental at best or worse outdated at inception while work on foundational questions is underfunded. Universities must set priorities for what work is actually worthwhile.

6. With lifespans increasing, bad ideas and outdated paradigms stick around ever longer. As Planck put it “science change funeral by funeral.” Exhibit A here is probably physics which has for far too long been dominated by string theory. Universities need to figure out how to make room for emerging ideas and heterodox approaches. 

7. Universities are long on buildings and short on digital infrastructure. Donors like putting their names on buildings but this is the past. Universities must emphasize the digital infrastructure for sharing teaching and research, producing and storing data, and applying AI across all efforts. Investing in lab automation also needs to be a high priority.

Universities had a great run. In some cases they have lasted for hundreds of years. But now they are facing an existential crisis. Can universities survive at all? I sure hope so because I believe short of the singularity, we absolutely need places that are dedicated to creating and sharing knowledge. But incremental changes won’t get the job done. Universities need to fundamentally reinvent themselves.

What could this look like? Almost certainly for a lot of places it will mean shrinking possibly by unbundling into separate efforts. But not in the outdated schema of disciplines, such as biology or economics, but around problems and projects. Programs need to become much more flexible in how they identify promising students and what the ideal student profile looks like. Same for professors. This is an incredibly exciting time for any institution that’s willing to do the hard work of going back to first principles of knowledge creation and dissemination.