PhD: What AI Accelerates and What It Doesn't Replace

A PhD student in machine learning and ecology argues that AI agents are now at the heart of thesis routines, yet they do not solve data problems rooted in their context. He highlights tangible gains in coding, experiments, and writing, but emphasizes the irreplaceable role of human interpretation and critical thinking. He also anticipates more collaborative laboratories with AI, refined methods, and a shift in skills.

What AI Does Not Replace in a Thesis

AI agents do not resolve data quality issues that require an understanding of context. Human interpretation remains essential to make sense of the results obtained. Critical thinking and creativity are indispensable in the research process. Even if the code runs faster, experiments are optimized, or drafts are generated, the challenges related to data with a history do not disappear with automation.

Where AI Agents Save Time

AI agents accelerate code development by automating certain programming tasks. They facilitate the optimization of experiments by producing faster data analyses. They can also generate drafts of research documents, simplifying the writing process. These contributions pertain to the operational aspects of doctoral work.

Laboratories: Collaboration, Methods, and Evolving Skills

According to this PhD student in machine learning and ecology, collaboration between researchers and AI agents could intensify to improve outcomes. Research methodologies could be adapted to more effectively integrate AI tools. The skills expected of researchers would evolve, with increased importance placed on mastering AI tools and data analysis.