Only completion ticks are saved in this browser. No account, analytics or learner submissions.
LESSON 01
Inventory what you actually pay for
List every AI service used or billed in the last 30 days. Record monthly cost, billing cycle, next renewal, typical task, and whether your data goes to the provider. Check statements rather than guessing.
Do this
Fill subscriptions.csv for at least two tools or one tool plus a no-AI baseline. For annual plans, divide the remaining bill by the remaining months only for comparison, not as a refund entitlement.
Completion check: Your inventory has real plan names, prices, renewal dates and one actual use case for each tool.
LESSON 02
Run an identical task comparison
Select one real, non-sensitive task, such as summarizing a public five-paragraph article. Give the *same* input and same scoring rules to every candidate tool, plus a manual baseline. Test again once if an answer looks unusual. Never paste confidential records into consumer tools without approval.
Do this
Use comparison.csv to grade correctness (0–3), time saved (0–3), effort to verify (0–3), privacy fit (0–3) and actual monthly cost. Record a concrete example, not a vague impression.
Completion check: You have at least two comparable results, evidence notes and a human quality check.
LESSON 03
Decide without a sunk-cost trap
Mark each subscription KEEP, DOWNGRADE, PAUSE or REVIEW. Use the least expensive tool that passes your minimum quality and privacy checks. A cheap but risky answer is not a saving. Make the decision reversible whenever possible.
Do this
Complete decisions.md. Note whether cancellation can be completed now, the renewal date, a manual alternative, and the reason. Do not assume a trial will automatically cancel.
Completion check: You can explain each decision in one sentence with an observed test, not a marketing claim.
LESSON 04
Run the 30-day review
Set a 30-day reminder in your own calendar. Record whether you missed a cancelled tool, whether a downgrade blocked a task, and whether observed outcomes matched the estimate. Recheck provider terms, plans and costs before any new decision.
Do this
Complete review.md with baseline monthly amount, current monthly amount, changes, failures, and next review date. Keep receipts for any changes.
Completion check: The final log shows a decision, who approved it, and a calendar date to review it.
Reusable prompt (human review required)
Compare these two draft outputs only using the attached *public or synthetic* task input. For each output, score factual completeness, unsupported claims, clarity, verification effort and privacy implications. List uncertainties and citations to the supplied input. Do not make a purchasing decision for me.
Fictional worked example
Example only: A fictional freelancer pays €24 and €12 per month for two writing assistants. Using a public 400-word press release, Tool B is accurate enough for the same task; Tool A adds no verified benefit. Decision: pause Tool A before renewal, keep Tool B for 30 days, monitor one live task. No actual savings claimed.
Take the working toolkit
Includes a complete field guide PDF, four editable templates, a prompt, fictional example and completion checklist. All files work offline after download.
