Ross Horlock is a PhD researcher at University College London working on immunology for cancer treatment. He is an ex AstraZeneca and Syngenta research scientist.
Fede Gambedotti is a PhD researcher at University College London modelling energy transition and energy equity. He is a former power trading analyst at Drax Group.

Will AI Replace Researchers or Transform Them?

This episode cuts through the hype, exploring whether AI is a threat to academic experts or a tool that will redefine the role of research itself. Ross and Fede, both PhD researchers, debate whether future scientists will be skilled problem solvers or just AI operators, and how academia must balance harnessing AI’s productivity with preserving essential human skills.

In this episode:

• The current impact of AI on PhD research, from literature reviews to coding (timestamps 0:41, 1:42)
• The risk of cognitive offloading and losing foundational skills such as methodology and troubleshooting (timestamps 2:53, 4:47)
• How AI’s biases and training data could influence scientific objectivity and integrity (timestamps 7:06, 22:17)
• The divergence between academia and industry: hands-on problem solving vs. AI management (timestamps 8:43, 9:35)
• The future of scientific expertise: Will AI make researchers obsolete or merely shift their focus? (timestamps 14:22, 22:03, 33:40)
• The challenge of maintaining objectivity and combating biases in AI-generated research outputs (timestamps 26:36, 29:19)
• The societal implications: replicating global knowledge equity and the role of academia as a verifier of truth (timestamps 35:40, 36:15)
• Practical insights into how AI is boosting productivity—freeing researchers from mundane tasks to focus on meaningful innovation (timestamps 36:50, 37:35)
  

Key moments of disagreement:

• The balance between utilising AI for maximum impact versus developing foundational skills (discussed at 5:14 and 11:07)
• Whether AI’s inability to provide objective, universal truths makes it unsuitable for certain research fields (around 31:36)

• The potential obsolescence of expert researchers versus AI’s role as an enabler for democratised knowledge and productivity (around 23:44 and 33:40)