Here's my AI ROI tracking data from the last quarter:

[Paste your tracker — task, time saved, quality notes, patterns]

Also, here's what I tried that didn't work well:
[List tasks where AI didn't help much or made things worse]

Analyze this data and give me:
1. **ROI Summary** — Total hours saved, estimated dollar value (use a loaded cost of $[your hourly rate x 1.5] per hour), and the top 3 highest-ROI tasks
2. **Pattern Analysis** — What types of tasks benefit most from AI? What types don't? Why?
3. **Quality Impact** — Where did AI improve quality beyond just speed? Where did quality suffer?
4. **Blind Spots** — Based on my task list, what recurring tasks am I NOT using AI for that I probably should be?
5. **Next Quarter Plan** — Three specific experiments to run next quarter based on these patterns
6. **The "Stop Doing" List** — Any AI-assisted tasks where the time investment isn't worth the return

Be honest. If AI isn't helping in an area, say so. I'd rather reallocate my effort than keep using a tool out of habit.