Breaking out of the mundane Part 3

What if average AI output is exactly what we need?

Here’s the position I’ve been testing that initially felt radical, perhaps we should stop fighting AI sameness and instead use it strategically. Let the AI handle the conventional baseline whilst we focus human attention on genuine exceptions and edge cases.

Consider customer support responses, where we can implement AI assisted drafting for our support teams. The responses became more uniform, they also became faster and more consistent. But here’s what we discovered through careful measurement, customer satisfaction didn’t decline. For 87 percent of support interactions, the “samey” AI response was exactly what customers wanted. Clear, professional, consistently formatted information that solved their problem without personality quirks that might be charming or might be annoying depending on the reader’s mood.

The 13 percent of interactions that needed something different, truly exceptional service for complex situations or frustrated customers, those got escalated to humans. But because humans weren’t spending time on the routine 87 percent, they could invest much more attention in the exceptional 13 percent. The average got more average, the exceptional got more exceptional and overall customer satisfaction increased.

This pattern holds across knowledge work, most market research actually should use standard frameworks. Porter’s Five Forces exists because