I’ve noticed something interesting while working on AI features in mobile apps…
Most teams don’t fail because of weak AI. They fail because they put AI in the wrong place.
They start with: “What can AI do?”
Instead of asking: “Where does the user actually struggle?”
And that changes everything.
From what I’ve seen, the real value of AI shows up in very specific moments:
When users are overwhelmed with choices.
When they repeat the same task again and again.
When they need to decide fast but lack clarity.
When too much information slows them down.
Not in chatbots. Not in flashy assistants. Not everywhere.
Just in the right place.
I recently broke down how I personally decide where AI belongs in a mobile app before building anything: dev.to/varsha_ojha_5b45cb023937b/this-is-how-i-de…
Curious how others here approach this:
Do you start with AI capability or user friction?
What’s one AI feature you’ve seen that actually reduced effort instead of adding complexity?
Feels like most products are still optimizing for “visible AI” instead of “useful AI.”
**
Want to see if others are thinking the same way.**
No responses yet.