Pretty soon, we may see hierarchical decision fatigue with regards to AI.
- Level 1: which GPT should I choose? This could come at the vendor and/or specific model choice.
- Level 2: What if I don’t like my answer? Should I tune my prompt and try again until I get the answer I want?
At what point will an end user get frustrated and take the first option they’re given, or opt-out of the process altogether?
Iterating on a design is critical for refining ideas and making a better product. But users won’t want to iterate when what they are trying to do is effectively “smart search.” The cognitive costs to them are too high.