I once worked as a mobile developer for a company where my job was to work on a new mobile app for a widely successful web product. The founder wanted me to implement a super obscure feature that I learned only about 1% of users actually use. The founder was the 1% of the users for this feature and wanted it in the app. I found this super frustrating and too much “feature creep” that went against the minimalist and straight-forward UI of mobile apps. Needless to say, I didn’t work long at this company.

Cue to now, and I’m realizing that AI reframes product development to building for ALL the 1% use-cases. That is what products like Instinct, GrokBot, and Muse are doing! This also applies to open-source libraries. Whenever I needed only one small feature from the library, I pretty much always downloaded the whole library anyway as it was usually too hard to pull out only one feature. Now, AI can easily help pull out or build just the one feature I need.