One of the most important aspects of AI is its ability to analyze patterns at a scale that’s hard for brains. 

That means it can help us notice what we had not. Those who use GenAI often will have experienced this underappreciated aspect.

My dissertation in the early 90s used neural nets to analyze the EEGs of people with and without Multiple Sclerosis (MS). But I had no delusions it would be an operationalized diagnostic. I trained it in an unsupervised mode, where I didn’t tell the AI whether the EEG was from an MS patient or not but rather let it find some pattern in the data where the differences between patients was maximized. The analysis pointed out an important feature of the brain waves the doctors (who already used this diagnostic method) hadn’t included in their heuristic interpretations.

We can apply this same unsupervised lens to our own creative and professional output to find the patterns we miss. We are often the worst judges of our own work because we are too close to our own intent.

I see this now when I ask AI for title suggestions; it often exposes an underlying theme I didn't realize was interfering with my main argument. Similarly, an analysis of past emails might illuminate a communication pattern—like unintended terseness or excessive hedging—that others find annoying but remains invisible to the sender.

Such insights are also possible at team and organizational levels, but that’s of course more complex in its implications. 

By providing a mirror that doesn't share our ego or our context, AI helps us spot the gap between the signal we think we are sending and the one actually being received.