Check your lab before the recap.
You're good when
- Tasks 1–3 run on at least two of ChatGPT, Claude, Gemini
- Winner log has Winner + one-line why for at least two tasks (1, 2, and/or 3)
- Task 4 attempted on an image-capable app or logged
skipped — no image tier/Claude N/A
Common fixes
- Rate limit — wait 30–60 minutes; log
cap hit; don't paste real data to "force" a reply - Different answers — normal; log preference, not "wrong AI"
- Thinking only on one paid app — note it in columns; don't compare silently
- Prompt drift — re-copy from the lesson page; don't paraphrase from memory
- One platform only — add a second free account, or finish the Lesson 9 tours, before you call the comparison done
Continue once the boxes match your honest effort.