A thesis isn't enough for a PhD (lemire.me)

Recent discussions at Harvard University have highlighted a growing concern in academia: the traditional PhD assessment model, heavily reliant on the thesis, is becoming inadequate due to advancements in AI-generated writing. Historically, a well-crafted thesis has often been seen as a sufficient qualifier for earning a PhD, even if the candidate struggles during the oral defense. However, the emergence of sophisticated AI tools capable of producing thesis-like documents raises critical questions about the validity of this assessment approach. As a response, Harvard faculty are advocating for a shift towards more comprehensive evaluation methods. They emphasize the need for rigorous, multi-faceted assessments that gauge a student's mathematical knowledge and independent achievement beyond just their written work. This recommendation could fundamentally alter the landscape of doctoral education, ensuring that candidates are assessed through practical and interactive evaluations that reflect their true understanding and capabilities, rather than allowing AI-generated texts to stand in for genuine scholarly effort. This shift not only addresses concerns over academic integrity but also enhances the evaluative framework to better prepare students for their future careers.