Open almost any app you used happily a year ago and there is a new little sparkle icon in the corner, promising an AI assistant you did not ask for and probably will not use. The bolt-on chatbot is the defining product decision of the year, and I think it is quietly making a lot of software worse. I say this as someone who loves the underlying technology.
The reflex nobody questioned
Somewhere in the last year, "add AI" became a box every product team had to tick, usually by grafting a chat box onto an interface that was working fine. The result is everywhere: assistants that duplicate features you can already reach in one click, that answer questions the app never had, that exist mostly so the company can say the app has AI. It is feature theater, and users can tell.
A chatbot bolted onto a good interface is not an upgrade. It is a confession that nobody knew what the AI was for.
What it costs
- Clutter. Every bolted-on assistant is another thing on the screen, another prompt to ignore, another decision the user did not want to make.
- Worse defaults. Effort that could have improved the actual product went into the AI feature the roadmap demanded instead.
- Eroded trust. When the assistant is bad, and the rushed ones usually are, it makes the whole product feel less reliable, not more modern.
Where AI in products actually works
The good implementations are almost invisible. AI that quietly makes an existing feature smarter, autocomplete that reads your intent, search that finally understands what you meant, a tedious task that now just happens, needs no sparkle icon and no chat box. It disappears into the product and makes it better. That is the opposite of the bolt-on, and it is much harder, which is exactly why so few teams do it.
The takeaway
The question a product team should ask is not "how do we add AI." It is "what is genuinely worse for our user than it needs to be, and could this technology quietly fix it." Sometimes the answer is a chat box. Usually it is not. The best AI in a product is the AI you never notice, doing a job you used to hate. The worst is a chatbot in the corner, blinking for attention, built to satisfy a roadmap rather than a user. We are drowning in the second kind, and I would trade all of it for a little more of the first.