A PhD is an apprenticeship in research – we can’t let AI take that away

2026-03-12 · Source: Artificial intelligence (AI) – The Conversation · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Cybersecurity & Data Privacy · Depth: Intermediate, quick

Summary

Academics initially doubted OpenAI's ChatGPT-5 "PhD-level" intelligence claims, yet many, including mathematician Terence Tao, now use generative AI tools like ChatGPT and Anthropic's Claude Code for research assistance. The author, a computer security researcher, details a project where they used AI to develop mathematical ideas, refine definitions and theorems, translate mathematics into a computerised format for proof checking, and implement Python programs. This AI-assisted workflow allowed the author to complete research equivalent to a year of PhD student work in just six weeks. While acknowledging the productivity gains, the author highlights a significant hidden danger: the sacrifice of student learning and the apprenticeship model crucial for developing future research leaders.

Key takeaway

For AI Scientists considering integrating generative AI into their research workflows, you should proceed with caution regarding its impact on mentorship. While AI can dramatically accelerate research output, ensure you are not inadvertently sacrificing the crucial learning and development opportunities for junior researchers and PhD students, which are vital for cultivating future scientific leaders.

Key insights

Generative AI significantly boosts research productivity but risks undermining the critical apprenticeship model for PhD students.

Principles

• AI tools require careful use.
• Authors retain ultimate research responsibility.

Method

Use ChatGPT for mathematical idea development and definition refinement, Claude Code for translating math to computer format and implementing programs, and rigorous manual checking at each step.

In practice

• Use AI for proof-writing and programming.
• Verify AI-generated content meticulously.

Topics

• Generative AI
• Academic Research
• Apprenticeship Model
• AI Productivity
• Mathematical Collaboration

Best for: AI Scientist, AI Researcher, Research Scientist, AI Student

Related on AIssential

• AI and the PhD student: friend or foe? — AI tools offer efficiency gains for researchers but necessitate rigorous human verification to prevent skill erosion and factual errors.
• Will AI spark a scientific renaissance — or a diffuse monoculture? — AI boosts scientific output but risks narrowing inquiry and critical thinking through automation.
• How AI is changing the nature of mathematical research — AI tools are revolutionizing mathematical research by automating proof generation and contributing novel ideas, accelerating discovery.
• [D] Supervisor support — PhD supervisor support varies widely, impacting student independence, research direction, and overall experience.
• May 27, 2026Economic ResearchCoding agents in the social sciences — AI coding agent adoption among social scientists is low but shows early productivity gains and significant demographic disparities.
• Are modern ML PhDs becoming too incremental, or is this just what research looks like now? [D] — ML PhDs often prioritize incremental benchmark improvements due to academic and industry incentives.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial intelligence (AI) – The Conversation.