Your question needs reframing.
As it stands you’re stuck.
If there’s no product roadmap showing how ai will help your product, being “ai first” Is meaningless bullshit.
But maybe if you reframe your question you’ll get a different answer.
What does your company do? (You left out all the relevant information that might lead to a positive answer so you’re not even making room for it to be possible. If this is going to work out, you need to make room for the possibility, so start there. “What would it take for this to work? And what can I do to help”)
As a technical person you might be best suited to explore how AI could make your product better. A less technical person can dream things up but without knowing what is technically possible it might not be useful.
The absolutely essential next step in the conversation with your CEO is how do you envision AI being useful for our product. you might need to assist in this ideation (if not you, the most senior technical person at the company, even if you’re not most senior sometimes fresh eyes see something key)
Imagine you make a smart thermostat like a product built around the recent post about the bloke who reverse engineered his landlord’s thermostat to enable smart features and mobile control.
Is there room in that scheme for AI? Absolutely. If he rebuilt his project to be AI first he’d probably build it around voice control or integration so it can predict the user’s needs. He might grant the user the ability to ask conversationally for added features (I’ll be away for 3 weeks let’s turn the heat down and save some money. Turn it back on when you get an alert that my phone is 2 subway stops away.)
Or he could build it around silently predicting what the user needs.
But honestly it might not be worth recreating that project to be ai first. If it works and does the 5 things he wants what more is there to do?
Saying “this is not a place that benefits from AI” is perfectly reasonable.
One place I’ve found AI to be useful is when interacting with the infinite variability of the real world. AI can look at data and transform it into the proper shape in a way that would take structured programming an infinite number of corner case loops.
Another area is offering conversational interface with large data sources or complex systems.
Another way is in rapidly building one-off integrations. (Staying on the smart thermostat page maybe there’s a new site where the system needs a new integration, you might use cursor to rapid prototype the integration)
In the same way there are places that are not a good fit for ai. Enumerate those too. Show up and run the good places for ai and bad places for ai against the list of features and technology you currently employ or to your known or imagined customer needs.
Ai is hugely powerful but just shoehorning it in everywhere is stupid and wrong. You should only use it if there’s a thing you can do better by using it. If you can’t see what that is you won’t ever be able to do it.