Requirement for live design/code review and oral assessment in graduate admissions and hiring
Problem
Current graduate admissions and hiring processes rely primarily on written applications or code-only reviews, lacking live design verification or oral defenses, which creates mismatches between candidate abilities and role requirements, especially as AI-generated work becomes common.
Pain points
- Graduate admissions evaluate candidates solely based on written applications, ignoring live presentation skills.
- Hiring decisions often prioritize asynchronous code review comments over in-person design or code reviews, risking poor design coherence.
- AI-generated mathematical proofs are produced but are messy and difficult for humans to review, indicating a need for better validation tools.
Questions to verify
- Would implementing a mandatory live design review improve hiring outcomes?
- Do graduate programs currently require an oral presentation as part of the admissions process?
- How can AI-generated math proofs be validated for clarity and correctness before publication?
Evidence
quote · sourcecomment_1
The author argues for evaluating Ph.D. candidates based more on the oral thesis defense than on the actual thesis. By essentially the same reasoning, I’ve been arguing for prioritizing in‑person design/code reviews over code‑only async PR comments. The important thing is to verify that the human has a coherent design in mind and can demonstrate that it got implemented, regardless of who or what was at the keyboard.
quote · sourcecomment_3
We already interview faculty hires; we must now do the same for graduate admissions... In Germany, the applicant gives a talk (30-40 min) to the research group they want to join, usually presenting their master's thesis, engage in discussion, often share lunch with the group, then do 1 on 1s with individual members of the group and a longer one with the PI.
quote · sourcecomment_5
The central complaint was that it was just too difficult for humans to review. Similarly now we're getting AI doing math. The proofs compile but are a mess. So just make the models better at writing clean proofs and explaining what they're doing to humans.
Disclaimer: this card is distilled by AI from community discussions as “pain point evidence”. It is a starting point for need analysis and validation, not a “validated market need”. AI has analytical limits; please open the original discussion above to manually verify before key decisions.