# Should we pick an AI tool and then find use cases for it?

**The prior:** ~~Pick the AI tool, then find the use case~~

## The short answer

No. That's backwards, and it's how a lot of AI projects start: someone sees a demo, someone says "we should use that", and the team spends a quarter looking for a problem shaped like the tool. Start from the outcome instead. Step one: write the result you want in one sentence, with a number in it. Not "use AI". Something like "invoices paid in two days instead of nine" (example). Step two: find the slowest step between here and there. It's usually a person waiting on another person. Step three: fix that step with the cheapest thing that works. Sometimes that's an agent. Sometimes it's a shared spreadsheet. Sometimes it's a phone call. The tool comes last. It's the least interesting decision in the room. If your AI roadmap is a list of tools, it's a shopping list.

## Outcome first
- Write the result in one sentence.
- Fix the slowest step.
- Pick the tool last.

By Eduardo Martínez, updated 2026-10-19. Source: https://rethinkyourpriors.com/priors/tool-first/

---
Rethink your priors by Eduardo Martínez. AI implementation. Outcomes over tools. And everything else.
YouTube: https://www.youtube.com/@rethinkyourpriors · Instagram: https://www.instagram.com/rethinkyourpriors/ · Work with Eduardo: https://edumartinez.ai/