From Lab to AI Lab: A Physicist's Story

Dr. Sarah M. is a postdoctoral researcher in condensed matter physics at a major research university. She spends her days studying quantum phase transitions, work that is deeply technical, rewarding, and, like most academic research, not especially well-compensated.

When Pasiflora AI launched this April, she was among our first experts to join. We sat down with her to understand what the experience has been like.

What Made You Join?

"Honestly, the pay was part of it," she says. "Postdoc salaries are what they are. But what actually drew me in was the idea that my domain knowledge could contribute to something beyond my own research. I spend years building expertise that gets used in a fairly narrow slice of academia. The idea that it could also help make AI systems more accurate in physics, that was interesting to me."

What Does the Work Look Like?

"It varies. Sometimes I'm reviewing AI-generated explanations of quantum mechanics and flagging where they're subtly wrong, the kind of wrong that a non-physicist wouldn't catch. Sometimes I'm solving problems and explaining my reasoning step by step, which apparently teaches the model something about how expert physicists approach problems. Sometimes I'm evaluating two responses and deciding which is more accurate or more useful."

"The Bloom Program made the transition easy. I understood the domain, obviously, but I had to learn how to apply that expertise to training data tasks. The calibration sessions were genuinely useful."

Has It Changed How You Think About AI?

"It has, actually. I used to think of AI as something that happened at technology companies, somewhat removed from my world. Now I see that the accuracy of these systems in technical domains is directly downstream of whether people like me are involved in building them. That's motivating. And a little sobering, because I've seen how much bad physics content is out there that a non-expert would never catch."

"The thing I didn't expect was how intellectually engaging the work is. It's not annotation drudgery. It's actually using your brain in a different way than research does.", Dr. Sarah M.