The 'local AI lab' folder exercise: a practical first workflow to try today

A simple starting exercise: create a folder called 'local AI lab' and put 10 files that matter to your work inside it — sales calls, meeting transcripts, old tweets, ideas, anything. Then run a local model like Gemma against that folder and have it produce one useful artifact. Ask it to create a weekly business pulse, find what's changed in customer conversations, group feature requests by the actual pain behind them, or review drafts and surface what your audience keeps responding to. The key insight here is that a chat answer is nice, but a useful artifact — a memo, a checklist, a brief, a report, a review — is what actually changes a workflow. A model reads the folder, the model writes the file, you inspect it, you improve the workflow, then you run it again.