10 Comments
User's avatar
Christina Muller's avatar

So fun collaborating with you, Joel!

Zara's avatar

"Judgment at work is built by doing the job, getting feedback and making mistakes" is the part AI can't shortcut. It's also why experienced professionals underprice themselves: twenty years of reps is the one thing a new hire and a chatbot can't buy.

Angel Radcliffe's avatar

The 19 percent gap only looks like a hiring story. Underneath it is a leadership story the same company will face in five years. Every judgment-building rep the tool absorbs is a rep nobody on the team got to work through. The managers running those rollouts are quietly writing their own future leadership bench.

Matt Doan's avatar

Handing my best work to the firm to own. I got that one wrong about 15 years in a row before I noticed.

Melanie Goodman's avatar

Hi Joel! Christina's closing argument is the part I had not expected, that the people entering work now could come out ahead precisely by choosing the slower route and getting things wrong early. It turns a worry about lost opportunity into something an individual can act on without waiting for an employer to protect the right work for them.

Sherry Heyl's avatar

Some of the most valuable judgment I've developed came from working through difficult situations, getting things wrong, and having to figure out what to do differently.

As someone focused on change leadership and AI readiness, I keep coming back to this question: Are we building a more capable workforce, or simply a workforce that can produce results faster with AI?

I think we need to be much more intentional about protecting the experiences that develop judgment, especially if we want to continue to innovate and find new creative solutions.

Anthony Tsigourakos's avatar

19% for 22 to 25 year olds is rough. where do the reps come from now if AI does the first draft of everything??

John Brewton's avatar

Small mistakes early save you from big ones later.

Breeze Dong's avatar

Christina’s third question, whether someone is becoming more efficient or quietly losing a skill, is one I wish every AI rollout plan carried.

Judgment also has a quieter half that is easy to skip: the patience to accept or reject what comes back, and to notice when a polished answer is wrong. It looks less impressive than producing the draft, so it rarely gets practised on purpose. As I wrote recently, leaving someone alone with a powerful model is not a development plan.

From coaching I’d add that people learn this half fastest when someone more experienced reviews the review with them: what did you accept, and why?

Christina, in the working groups you describe, do managers find time to review the review, or does it get squeezed out by the pressure to show adoption numbers?

Christina Muller's avatar

Breeze, thank you for reading so closely. "Review the review" is such a helpful way to put it. I worry it's exactly the kind of time that gets squeezed when adoption numbers are what leaders are watching. I'd want a working group to put it right on the agenda, so a mgr can sit w/ someone and talk through what they accepted and why.