Every responsible conversation about AI ends with the same advice: check the output. But being 'the human in the loop' is often hard, draining work, and it gets harder when everything AI produces looks like an end product.

In this week's episode of The Mindtools L&D Podcast, guest host Julie Dirksen and Claire are joined by Dr Markus Bernhardt of Endeavor Intelligence to explore the human behaviors behind validating AI outputs, and what organizations need to do to support them.

We discuss:

• why being the human in the loop requires psychological stamina
• how the agentic 'hamster wheel' erode our ability to validate outputs
• triaging tasks by how close they sit to your core expertise and values

Julie's new book on skills development is Design for How People Build Skills.

For more from Markus, visit endeavorintel.com. For more from Julie, visit usablelearning.com.

In 'What I Learned this Week', Claire shared a new term from a scuba-diving friend: the 'incident pit'. It describes how one small problem, like a cracked mask, escalates into a bigger one.

Julie recommended The Score: How to Stop Playing Somebody Else's Game by C. Thi Nguyen. It looks at how scoring systems, from climbing grades to university rankings, can quietly change what we value.

For more from Mindtools Kineo, visit mindtools-kineo.com. There, you'll find out how we can help your organization build AI skills, and how our Skills Practice scenarios help build human skills.

Connect with our speakers

If you'd like to share your thoughts on this episode, connect with us on LinkedIn:

• Julie Dirksen
• Claire Gibson
• Dr Markus Bernhardt