Imagine preparing for a conversation with a clear result in mind. The answer comes back differently. Now there is the event itself, and the expectation it interrupted.
In predictive-processing frameworks, differences between expected and incoming information are often described as prediction errors. How those differences influence learning depends on the model and context, including how reliable the information is considered to be.
That does not mean a fixed percentage of all learning happens through surprise, or that painful experiences automatically rewire someone for the better. Those would be much larger claims.
The practical question is what this event gives you reason to reconsider. Was the expectation well supported? Is the new information clear? What else might explain the response?
An immersive can place an unexpected turn inside a story. AATAM uses that creative possibility to invite attention and curiosity, while leaving any personal learning or change as an open question.