As AI reshapes how software is built, what does it actually mean to be a good engineer?
Chelsea Troy, who leads Machine Learning Operations at Mozilla and teaches computer science at the University of Chicago, argues that the future of engineering isn’t about typing faster but thinking better.
In this episode of Distributed, Chelsea joins host Jack Hannah to unpack the three skillsets that matter most in the age of AI: investigative, evaluative, and innovative thinking. She explains why AI tools still fall short in areas that require human reasoning and judgment, how engineers can uplevel the skills that set them apart, and why collaboration and a solid grounding in computer science remain essential.
Website: https://chelseatroy.com
LinkedIn: https://www.linkedin.com/in/chelseatroy
LinkedIn: https://www.linkedin.com/in/jack-hannah/
Website: https://tuple.app/
(00:00) Intro
(01:08) The 3 most underrated skillsets engineers rely on every day
(08:45) How an interpreter workshop and a Rust compiler project revealed AI’s limits
(14:59) How to level up investigative, evaluative, and innovative skills
(21:25) Why collaboration still matters but isn’t a new challenge created by AI
(25:38) The value of early career pair programming and why it’s hard but worth it
(29:09) Why pairing with an LLM has value but can’t replace human collaboration
(34:33) Formal CS degrees vs. nontraditional paths and why studying CS still matters
(43:49) Rapid fire round
What can we expect of LLMs as Software Engineers?: https://chelseatroy.com/2025/07/14/what-can-we-expect-of-llms-as-software-engineers
Bob Nystrom’s blog: https://journal.stuffwithstuff.com/
Tim Peters: https://github.com/tim-one
Crafting Interpreters: https://craftinginterpreters.com
Using the Python Interpreter: https://docs.python.org/3/tutorial/interpreter.html
Alan Turing: https://en.wikipedia.org/wiki/Alan_Turing
Ada Lovelace: https://en.wikipedia.org/wiki/Ada_Lovelace