Dear Community, 

Anthropic CEO Dario Amodei has warned of substantial losses in entry-level white-collar employment. 

I don’t see a graduate jobs apocalypse as a foregone conclusion. But as a university dean, I find the question behind it impossible to ignore: how should students and universities prepare? 

If AI can generate answers, write code and analyze data, what exactly should we now teach? 

 AI is changing what universities need to deliver. Answers are becoming easier to produce. The scarce bit is knowing what to ask and judging whether the answer makes sense. 

In two Princeton-led research tests, AI agents completed the engineering work but failed to make substantial progress on the research questions. That’s a useful distinction: the ability to execute does not guarantee the judgment to decide which work is worth doing. 

Students and parents ask me all the time which discipline is future-proof. My advice is to get really, deeply good at something. Engineering. Computer science. Finance. Law. Biotech. 

 But being brilliant at one thing is not enough. The knack is applying that expertise to real problems, connecting it to other disciplines, and even creating something new with AI. 

For universities, this means three things. 

First, keep advancing knowledge. Original research and discovery remain central to our purpose, with more emphasis on doing with industry as a collaborator. 

Second, rethink assessment. If AI can produce a polished answer, testing students mainly on reproducing one makes little sense. How did they frame the problem, work with AI and test the result? Can they explain their thinking? 

Third, put students into situations where they have to think on their feet, adapt and exercise judgment. Knowledge matters. So do resilience, curiosity, drive and intellectual energy. 

The jobs question adds another challenge. Routine work gets things done while teaching junior people how to do them. If AI takes over more of that work, where does the next generation gain experience? 

Universities cannot answer this from inside campus walls. Students, faculty and industry partners need to tackle problems whose answers are still unknown. Research informs practice, practical challenges generate research questions, and students learn by contributing. 

Bring these together, the best universities will evolve into living labs. The degree has to come with real, demonstrated capability. 

Better preparation alone will not create jobs. We also need new products, services, and ventures. As AI cuts the time and cost of familiar tasks, we must aim higher and solve the problems that have previously been considered unsolvable: from healthcare at a global scale to the energy transition to new scientific applications. 

AI should ultimately empower humanity. The purpose of the university is to make sure it helps us thrive.


  • A diverse group of people posing together in a professional setting.
  • A motivational phrase encouraging individuals to empower their entrepreneurial spirit.
  • A motivational quote encouraging innovation with a proactive mindset.
  • A message emphasizing the importance of cherishing diversity in people, mind, and experience.
  • This image features a motivational phrase emphasizing a humanistic approach in a stylish font.