Why did top value investors (Li Lu, Pabrai) miss the MU AI boom? Was it a mistake, and how do we avoid it?

Look at Dataroma. Value investors like Li Lu and Mohnish Pabrai used to have massive Micron (MU) positions, but then they sold out before this whole run.

When the AI race started, anybody listening to MU earnings calls could tell that MU was going to be profitable. Very profitable: at least for the next few years due to HBM demand.

Why did all these legendary value investors miss this opportunity?

* Was it because memory is so cyclical that they figured they couldn't time the end of the cycle?
* Did they think the projected earnings weren't sustainable or real?
* Or was it just strict adherence to staying within their "circle of competence" on cutting-edge tech?

Would you call this a mistake on their part, or just the necessary trade-off of value investing? And more importantly, what can we as retail investors do in the future to avoid missing structural shifts like this when they're hiding in plain sight on earnings calls?

Curious to hear everyone's thoughts.