Podcast: The AI Revolution Fails Without Psychological Safety For Developers: A Conversation with Erin Doyle

The AI Revolution Fails Without Psychological Safety For Developers: A Conversation with Erin Doyle - InfoQ

InfoQ Homepage Podcasts The AI Revolution Fails Without Psychological Safety For Developers: A Conversation with Erin Doyle

Architecture & Design

InfoQ Certified High-Performing Teams Program (online, Nov 17): Build a clearer operating model for your team.

The AI Revolution Fails Without Psychological Safety For Developers: A Conversation with Erin Doyle

Sep 28, 2026

Podcast with

• Erin Doyle

by

• Michael Stiefel

Like

new dropdown rading list

• Reading list

In this podcast, Michael Stiefel spoke to Erin Doyle about how artificial intelligence tools are transforming the software development process, and how they are affecting software engineers. Architecture skills are now critical because engineers need to understand how to build systems and handle the constraints and associated ambiguity in the requirements. At the same time, given the rapid change both in the nature of the AI tools, and the amount of code they can produce which can overwhelm developers, psychological safety becomes essential for successful AI adoption. Teams need a secure, non-judgmental environment to share their successes, admit failures, ask for help, and express skepticism.

Industry lacks a plan to train junior engineers because AI is taking over the tasks they were traditionally trained on, and university education is not adequate preparation for this new reality. The conversation concluded by discussing how software developers and organizations must retain complete ownership and assume full liability for the output generated by AI tools. To manage this, there is an urgent need to develop advanced verification and telemetry tools that can support the review process.

Key Takeaways

• Artificial Intelligence tools shift the focus of software engineers from writing code to understanding system architecture. Grappling with ambiguity becomes an increasingly critical problem..

• Psychological safety is essential for AI adoption, as team members need a secure environment to share successes and failures, ask for help without judgement as well as to be able to express skepticism. Success at using AI requires team alignment on goals and processes.

• The industry has no plan to train junior engineers as artificial intelligence takes over the tasks that these engineers were traditionally trained on. Current university education does not seem adequate.

• The rapidity at which AI tools can produce code has led to a situation where engineers can feel overwhelmed at the need to verify what has been produced. There is a need to create advanced verification and telemetry tools to support this process.

• Developers and organizations must be prepared to maintain complete ownership and liability for the output of AI tools. They cannot be used as scapegoats for errors.