The traditional academic career path of becoming a replica of one's Principal Investigator (PI) is largely obsolete. A PhD provides a broad skill set in critical thinking and data management applicable across many sectors. Young researchers should focus on the big problems they want to solve, not just replicating a disappearing job.

With industry dominating large-scale compute, academia's function is no longer to train the biggest models. Instead, its value lies in pursuing unconventional, high-risk research in areas like new algorithms, architectures, and theoretical underpinnings that commercial labs, focused on scaling, might overlook.

Instead of choosing a career based on its perceived "safety" from AI, individuals should pursue their passions to quickly become domain experts. AI tools augment this expertise, increasing the value of experienced professionals who can handle complex, nuanced situations that AI cannot.

With industry dominating large-scale model training, academia’s comparative advantage has shifted. Its focus should be on exploring high-risk, unconventional concepts like new algorithms and hardware-aligned architectures that commercial labs, focused on near-term ROI, cannot prioritize.

Because models excel at mechanical theorem-proving within defined boundaries, the mathematician's core differentiator is shifting toward mathematical taste. Deciding which real-world problems to model, asking insightful questions, and framing theories will increasingly outweigh the routine mechanics of proof verification.

When evaluating young builders, the most compelling signal is not breadth of knowledge but evidence of having followed a single thread of curiosity to its depths. This demonstrates rare traits like persistence, intrinsic motivation, and the ability to teach others something new.

With industry dominating large-scale model training, academic labs can no longer compete on compute. Their new strategic advantage lies in pursuing unconventional, high-risk ideas, new algorithms, and theoretical underpinnings that large commercial labs might overlook.

A major frontier for AI in science is developing 'taste'—the human ability to discern not just if a research question is solvable, but if it is genuinely interesting and impactful. Models currently struggle to differentiate an exciting result from a boring one.

The ultimate skill of a great scientist isn't performing calculations but identifying the most fruitful questions to pursue. While AI is becoming superhuman at answering well-posed problems, the human role of taste and strategic direction-setting remains paramount for breakthroughs.