Can AI Become a Scientist? How Machines are Starting to do Research

About this episode

The episode explores whether AI can transition from being a research assistant to becoming a full-fledged scientist.

Can AI move beyond being a research assistant and become an actual scientist? In this episode of Today in Tech, Keith Shaw speaks with Mayank Kejriwal, research associate professor at the University of Southern California and CEO of Grail, about the rapid evolution of AI in scientific research. From generating hypotheses to designing experiments and even writing research papers, AI is beginning to take on tasks that were once the exclusive domain of human scientists. But what does that really mean for the future of discovery?

People in this episode

Host: Keith Shaw

Guest: Mayank Kejriwal

Topics covered

• AI in research
• scientific discovery
• machine learning
• hypothesis generation
• experiment design

Keywords

• AI
• scientist
• research
• hypotheses
• experiments
• discovery
• machine learning

Mentioned in this episode

Organizations: University of Southern California, Grail

More episodes of Today in Tech

• AI is Creating More Code, But is it Breaking Software? · June 9, 2026 · 35 min
• AI Can Write the Sales Pitch, But Can it Close the Deal? · June 2, 2026 · 38 min
• AI is Making Business Scams More Convincing Than Ever · May 26, 2026 · 36 min
• Are We Losing Control of AI Agents? · May 19, 2026 · 35 min
• Will AI Start Moving Your Money? The Rise of Agentic AI in Finance · May 12, 2026 · 35 min
• AI Agents are Here, But Your Data Still Isn’t Ready · May 5, 2026 · 31 min