phd / thesis examination / mental health

A PhD student in machine learning says after 18 months they feel underprepared and lacking foundational theoretical knowledge, and that peers and advisors often dismiss such gaps. They report scrambling to catch up on core theory—statistics, probability, optimization and ML fundamentals—and suggest graduate programs may tolerate low theoretical standards. The post argues this is common in ML academia, raising concerns about training quality, preparedness for research, and consistency across advisors and institutions. It matters because gaps in theoretical grounding can affect research rigor, reproducibility and the long-term development of reliable ML systems, with implications for hiring, curriculum design and mentorship in AI programs.

The author, concluding their PhD in computer science (ML/vision), offers a candid survival guide for prospective and current PhD students, arguing the degree grants freedom, ownership of work, status, and future options compared with industry roles. They weigh pros — autonomy, personal growth, higher hiring value in applied ML, and career flexibility — and acknowledge variability across programs and advisors. The guide targets decisions like whether to pursue a PhD and how to navigate its psychology and productivity, noting the intense learning and identity shifts involved. It’s framed as practical, experience-based advice rather than prescriptive rules, with caveats about domain-specificity.

A first-year PhD student asked on Reddit whether it is a red flag that their dissertation topic keeps changing every few months. The poster says they are not narrowing down to a single research question, instead exploring a broad range of applications within their domain. They describe their “core topic” as niche but note they are likely working on the application side, applying the niche method or idea across multiple areas. While they enjoy the variety, they worry it may signal a lack of focus or depth—“playing around” rather than mastering a specific research direction. The post provides limited details about the field, advisor guidance, or concrete milestones, so the discussion centers on the general concern of topic drift early in a PhD.

A Reddit user posted that waiting for their PhD thesis examination results is harming their mental health. The author says they submitted their thesis a little over two months ago and have experienced persistent anxiety since submission, adding they have been posting frequently about the situation. The post is framed as a request for others’ experiences and support, highlighting the emotional strain that can accompany long academic review timelines. No university, examiners, field of study, or expected decision date is provided, and the excerpt contains no details about the examination process, institutional policies, or any actions taken to seek professional help. The available information is limited to the author’s self-reported anxiety and the two-month wait after thesis submission.