Home / Sprint 3 / Claims in Practice

~1 min · Why I built this

Looking back, I hope this course taught you these five things. Now I want to know honestly whether they landed — for you, not for me. Your reflection on this is more critical to your learning than how good a job I have done.

— Prof. Sathya

Claims in Practice — AI Discussion SDL IS

Sprint 3 Transition · Due · 5 points

📓 Prerequisite: Handwritten Claims Reflection

Complete the Bridge Reflection before starting this activity. The discussion starts from what you wrote by hand — if you skip that step, the AI has nothing real to push on.

What This Activity Does

  1. Work through five claims one at a time. Each claim from the course framework gets its own focused discussion.
  2. Describe where you see each claim in your Sprint 1-2 work. Be specific — name the assignment, the moment, the decision.
  3. The AI generates 3 follow-up questions pushing for specificity. Vague acknowledgments ("I definitely grew in this area") get challenged.
  4. Write a refined response that incorporates the pushback. This is where the real thinking happens.
  5. Download JSON and upload to Canvas. Your responses are saved as you go.

What the AI Will Push On

The AI is looking for evidence, not agreement. Here's an example:

VAGUE (will get pushed back on):

"I did good problem framing in Sprint 2 — I asked my partner a lot of questions before I started building."

SPECIFIC (this is what we're looking for):

"In Sprint 2, my partner needed a way to track medication interactions. I'd never built anything medical before. I used Claude to learn about drug interaction databases, but the key decision — to focus on the top 5 interactions her patients actually encounter instead of trying to cover everything — that was mine. The AI would have built something comprehensive and useless."

Start Claims Discussion →

What to Submit

Complete the interactive activity, download your responses as JSON, and upload to Canvas.

Format: JSON file download from the activity
Due:
Points: 5
Submit on Canvas →