The Tragedy of the Cognitive Commons
10 August 2026
I read a paper this weekend called The Tragedy of the Cognitive Commons, by researcher Nolan Lovett, published in Human Resource Development Review. It names something I think most people who work with AI daily have felt without having language for it.
The paper names two kinds of expertise:
- Internalized Mastery: The deep knowledge you only get from years of doing the unglamorous version of a job yourself.
- Distributed Mastery: The newer skill of directing AI well: prompting it, steering it and knowing which tool to use.
Its central claim is simple: you can only catch an AI answer that sounds right but is actually wrong if you still have real, hands-on domain knowledge, not just skill at prompting.
The paper calls this link the Validation Tether. Your ability to use AI well stays tied to, and limited by, how much real expertise you still have.
One study the paper cites found that 80.7% of people who saw a flawed AI recommendation noticed something was wrong with it, and followed the recommendation anyway.
That tells you the problem isn’t necessarily attention. Noticing something is off and having the confidence and expertise to override it are two different skills. Only one of them gets built by using AI a lot.
The incentive problem
What makes this land for me isn’t the AI part. It’s the incentive underneath it.
Every company that skips hiring juniors because AI does that work now saves money right away, and gets to keep all of that saving for itself.
The cost shows up later, and lands on everyone: fewer people around in fifteen years who can actually catch AI’s mistakes.
Even the company that made the cut ends up paying that cost, along with everyone else.
In the moment, though, the decision looks like nothing but an easy win.
The paper is honest that this is a prediction, not something that’s already happened.
But it’s hard not to believe the truth in this conclusion:
Your ability to use AI well only holds up as long as someone nearby still has the slow, unglamorous kind of expertise it depends on.
Reference
