Tony Feng, a mathematics professor at UC Berkeley, warns that «the very foundation of modern academia has been blown to bits.» In the past, a math Ph.D. cultivated resilience and critical thinking, but «suddenly you can produce a passable Ph.D. thesis with the push of a button»—not just in mathematics, but across sociology, computer science, and engineering.

Feng traces the erosion: from physical libraries where scholars memorized papers, to the web and Google Scholar compressing literature review from years to hours. Ph.D. output surged—in the United States from 36,000 in 1990 to over 58,000 in 2024, while China exploded from a few thousand to roughly 88,000 annually. Now agentic AI can generate a 2015-quality thesis on demand. Since most theses were «painful to read» anyway, the bar was never high.

Oral defenses rarely fail candidates. Feng recalls a colleague praising a student's «extensive review of the literature» produced in record time—while proudly opposing ChatGPT. Feng stayed silent. The system, he concludes, is not merely shaken but «wiped out.»