Software Without Understanding: Leading When Coding Outpaces Comprehension

Not too long ago, each new feature in our products came with a reasonable assumption: there was a human programmer in the loop. Writing the code was how engineers learned their systems, and how organizations built knowledge of how those systems worked.

Agentic coding breaks that feedback loop. Without understanding, we risk generating tomorrow's legacy systems at agentic speed.

Drawing on cognitive psychology and hands-on experience leading AI-first teams, Adam Tornhill explores what this shift means for technical leaders.

You’ll learn:

When pushing agent throughput backfires, and how to set a sustainable pace.
Where to invest in automated verification versus what to leave for humans.
Practices that keep teams learning their systems when they no longer write the code.

About Adam Tornhill

Adam Tornhill is a programmer who combines degrees in engineering and psychology. He’s the founder of CodeScene where he designs tools for code analysis. Adam is also the author of multiple technical books, including the best selling Your Code as a Crime Scene and Software Design X-Rays. Adam’s other interests include modern history, music, retro computing, and martial arts.

More About Adam »