# Rethink your priors

> A deadpan stick-figure video series by Eduardo Martínez about implementing AI (outcomes over tools) and everything else in tech. About a minute each. The core idea: start from the outcome, work backward, and pick the tool last.

Rethink your priors is made by Eduardo Martínez. Episodes are short (about a minute) and posted on YouTube (@rethinkyourpriors) and Instagram (@rethinkyourpriors). Each episode challenges one common belief ("prior") about AI at work. Example numbers in episodes are labelled as examples, not data.

## Priors, rethought (short answers)
- [Do we need an AI strategy?](https://rethinkyourpriors.com/priors/ai-strategy/): A list of outcomes. Each one has an owner. That's the strategy.
- [Should we pick an AI tool and then find use cases for it?](https://rethinkyourpriors.com/priors/tool-first/): Write the result in one sentence. Fix the slowest step. Pick the tool last.
- [Does every company need an AI chatbot?](https://rethinkyourpriors.com/priors/chatbot/): Let people finish a task. Not chat about it.
- [Do more AI agents mean more output?](https://rethinkyourpriors.com/priors/more-agents/): More agents means more review. Pick one outcome first.
- [Should AI implementation work be billed by the hour?](https://rethinkyourpriors.com/priors/billing-hourly/): Price the outcome. Agree on the baseline first.
- [Do AI adoption and usage numbers prove ROI?](https://rethinkyourpriors.com/priors/adoption-roi/): Measure one business number. Before and after.
- [Is prompt engineering the most important AI skill?](https://rethinkyourpriors.com/priors/prompt-engineering/): Know the workflow. Know the outcome. The prompt follows.
- [What is MCP, and is it complicated?](https://rethinkyourpriors.com/priors/mcp/): Decide what the AI may touch. The plug is the easy part.
- [If an AI demo works, is it ready for production?](https://rethinkyourpriors.com/priors/demo-ready/): Ship the last 20%. That's where the product lives.
- [Should we use AI to summarize every meeting?](https://rethinkyourpriors.com/priors/meeting-summaries/): Ask what decision the meeting is for. Then maybe delete it.
- [Do AI benchmarks tell you which model is best for your work?](https://rethinkyourpriors.com/priors/benchmarks/): Test it on your own work. Ten real examples beat a vibe check.
- [Will AI agents replace junior staff?](https://rethinkyourpriors.com/priors/replace-junior-staff/): Agents need supervision. Budget for the review.
- [Should a human approve every step an AI agent takes?](https://rethinkyourpriors.com/priors/human-in-the-loop/): Put a human at the steps you can't undo. Only those.

## Episodes
- [ep. 1: Start from the outcome](https://rethinkyourpriors.com/episodes/start-from-the-outcome/): Most AI projects start with a demo. That's the problem. Transcript included.
- [ep. 0: Episode zero](https://rethinkyourpriors.com/episodes/episode-zero/): Hi. I'm Eduardo. This is episode zero. The one before the first one. Engineers love that. Transcript included.

## About
- [About the series and the narrator](https://rethinkyourpriors.com/about/)
- [Full text of every page in one file](https://rethinkyourpriors.com/llms-full.txt)

## Optional
- [YouTube channel](https://www.youtube.com/@rethinkyourpriors)
- [Instagram](https://www.instagram.com/rethinkyourpriors/)
- [The Compounding Founder, the long-form writing](https://www.thecompoundingfounder.com/)
- [edumartinez.ai, work with Eduardo](https://edumartinez.ai/)