Design Session 2 Focus
You diagnosed your learning experience across the four lenses in the last design session. Now synthesize your findings, define what your redesign must accomplish across those same lenses, and write a clear design challenge statement.
Design Time
~60 minutes
This design session is about pinpointing the AI vulnerability problem and what you want to fix in your redesign. Here’s the design path we’ll follow:
Design Decisions Document
Work through the Define section of your copy of the document to:
- Frame the problem: name the most significant sources of vulnerability across the four lenses and the core design problem to solve
- Set your design requirements: what learners must be able to do, what thinking to keep central, what to make visible, and what evidence would make learning credible
- Set your design criteria: the role AI should play, and what the redesign should preserve, protect, or avoid
- Write your design challenge statement, and name what success would look like
Your Design Reflections
When you finish this design session, reflect on your design challenge statement. What does success look like in terms of a redesigned learning experience? Jot your thoughts in your Design Decisions Document.
Optional | Podcast
Want more context about this design topic? Episode 2 of the AI for Educators Design Lab podcast explores learning goals in the age of AI.
Jennifer’s Design Session 2: Define
My Design Reflections:
Like everyone who worked through the live Design Sprint, I’m working through an AI-vulnerable design challenge of my own. What follows are my design session notes and reflections that align with the same design decisions you’re working through.
Yours in learning,
Jennifer
Problem Framing
- Overview: Looking across the four lenses, the most significant sources of AI vulnerability are related to Tasks, Visibility, and Measurement, but Goals need tweaking too (e.g., adding learners’ evaluation of AI output).
- The pattern: it is a combination, but the through-line is that the evaluation and synthesis I value are never made visible or required as evidence. Even sound goals fail here, because the task elicits a product rather than the reasoning behind it.
- The core design problem, in two sentences: The annotated bibliography asks for a product that AI can now produce on its own, while the reading, judging, and connecting it is meant to remain invisible and unverified. How might I redesign it so that evaluation and synthesis are required, visible, and credible to the student, even where AI is available?
Design Requirements
- Learning Goals: What should learners be able to know or do, even where AI is available? Judge whether a source is credible and methodologically sound, and explain how it bears on their own research question, including recognizing when a source is not real.
- Learning Tasks: What learning tasks best support those goals, even where AI is available? Tasks that require learners to work with their own questions and the actual sources: comparing sources against one another, justifying why they selected each, explaining relevance in their own words, and accounting for where AI helped or misled them.
- Learning Process Visibility: What needs to become more visible during the learning process so I can better guide and scaffold learners in those tasks? The selection reasoning and the reading itself, so I can see why these sources, what the student actually took from each, and how the thinking is developing, and intervene before the final product.
- Learning Evidence: What kinds of evidence would credibly measure learning of the intended knowledge and skills? Evidence that shows synthesis across sources and judgment about their fit to the student’s question, rather than a single polished artifact.
Design Criteria
- What role should AI play going forward? Guided. These students will and arguably should use AI in literature work; the redesign should make that use visible and require verification and judgment on top of it.
- Why is this the right role for AI? A ban is unrealistic and forfeits a real AI-literacy goal, while guided use mirrors authentic scholarly practice.
- What would count as credible evidence in this AI context? Here the emphasis is attribution: evidence I can credit to this specific student, such as a live account of their own choices or a synthesis tied so closely to their evolving question that a generic AI output could not reproduce it. Note how close this sits to the evidence requirement above. In Requirements, I named the kind of evidence but here I am naming what makes it credible given AI, which is that it can be attributed to the learner.
- What should a redesign preserve, protect, or avoid? Preserve the critical evaluation and synthesis, protect the connection to the student’s own research question, and avoid turning this into a detection exercise or busywork.
Core Design Challenge
My core design challenge is to redesign the annotated bibliography so that learners can demonstrate genuine evaluation and synthesis of sources for their own research question, in a context where AI is available and may be used as a verified aid.
- What success would look like: I can see, from what each student produces and how they account for it, that they read and judged real sources and connected them to their own question in ways a generic AI output could not.
- Uncertainties before Develop: how far to formalize AI use, whether an oral or in-class component is feasible given class size, and how much process to require without overloading students.
Next: Design Session 3 | Develop
Previous: Design Session 1 | Back to Design Sprint On-Demand