Guest: Alex Craig, Chief of Staff at PortSwigger

Host: Scott Amenta, Founder, Chief of Staff Network

Most companies say they're adopting AI. PortSwigger actually did it, and the results weren't what anyone expected. Alex Craig breaks down how a bootstrapped cybersecurity company built a real AI culture from the ground up and why, for Chiefs of Staff, the hardest part isn't the technology. It's figuring out where human thinking has to show up in the work.

About Alex Craig

Alex didn't apply for a Chief of Staff role. He kept raising his hand for problems no one else owned, and eventually the title caught up to the job he was already doing. He works directly with Daf, PortSwigger's CEO and founder, at a bootstrapped cybersecurity company of roughly 200 people in northwest England. His core function is reading organizational decisions and correcting course toward how Daf would actually make them.

Key Takeaways

• The fastest AI adopters weren't the engineers. Kitchen and front-of-house staff embraced it first. Developers pushed back hardest, resistant to anything non-deterministic.

• AI eliminated the lower-stakes work that gives you room to breathe. Slide formatting, meeting prep, basic synthesis. What's left is only the complex decisions, all the time.

• The biggest emerging problem isn't bad AI output. It's employees forwarding AI-generated work they haven't read, pushing decisions upward with no personal thinking attached.

• The incognito-context test: if your manager could produce the same output in five minutes with no context about you, there's no "you" in the work.

• A personal AI system is only as useful as the training behind it. Alex spent months building one that reflects his actual views, reading habits, and listening history. The mental model is a new hire you develop over time, not a tool you pick up and put down.

Alex is publishing a full technical breakdown of his personal AI Chief of Staff system alongside this episode. Look for his blog post, "I'm a Chief of Staff. So I Built Myself One."