Frame the design decision
Frame the opportunity before discussing features: whose problem are you solving, at which moment in their workflow, and what evidence says it is worth solving?
Pathway 2 · Lesson 0190 min
Design Thinking for AI
The Double Diamond Framework applied to AI products
Learning Objective
By the end of this lesson, you will be able to: Apply the Double Diamond framework to an AI design challenge, distinguish between divergent and convergent thinking phases, and identify responsible AI design principles.
95%
of AI products fail because they solve the wrong problem
Engage
"The most common AI design failure isn't bad technology. It's building technically impressive solutions to problems nobody has. How do you make sure you're solving the right problem?"
Explore
IDEO's introduction to design thinking — the mindset and methods behind human-centred design.
Explain
The Double Diamond: Four Phases
Developed by the UK Design Council, the Double Diamond structures design as two diverge-converge cycles:
Diamond 1: The Right Problem
Diamond 2: The Right Solution
⚠️
The Most Common AI Design Mistake
Jumping straight to the second diamond — building solutions before properly defining the problem. This produces technically impressive AI that nobody uses because it doesn't solve a real human need. The Discover and Define phases are not optional.
Explore
The 'How Might We' technique for reframing problems into design opportunities — essential for the Define phase.
Explain
Responsible AI Design Principles
Every AI product design decision should be evaluated against four principles:
Transparency: Can users understand when AI is involved and why it made a decision?
Fairness: Does the AI perform equally well for all user groups? Where might bias emerge?
Controllability: Can users override, correct, or opt out of AI decisions?
Accountability: Is there a clear human responsible for the AI's outputs and impacts?
Elaborate · Worked Example
Case Study: Designing an AI Hiring Tool
Context
A company wants to use AI to screen CVs and shortlist candidates.
The Situation
Discover: Research reveals that hiring managers spend 70% of their time on initial screening — a genuine pain point. But research also reveals that automated screening has historically disadvantaged women and minorities. Define: HMW use AI to reduce screening time while actively improving diversity outcomes? Develop: Prototype 3 approaches — AI ranks candidates, AI flags potential bias, AI assists human screeners. Deliver: Test with real hiring managers. Measure time saved AND diversity outcomes.
The Outcome
The responsible design process produces a tool that saves time AND improves fairness — because the problem was properly defined before building began.
Try It · Reflect & Apply
Apply the Double Diamond: Choose an AI opportunity in your organisation. Write a Discover question (what would you research?), a Define statement (your 'How Might We'), a Develop idea (one prototype concept), and a Deliver metric (how would you measure success)?
Your notes are saved in your browser for this session.
Evaluate · Knowledge Check
1 / 3In the Double Diamond, what happens in the 'Discover' phase?
Design Practice Studio
AI Discovery Studio
Create the artefact
Create a one-page Double Diamond brief with a user group, observed pain point, ‘How might we’ question, harmful failure to avoid, and measurable success condition.
Test with evidence
Interview or observe three people before committing to a solution. Listen for a consistent pain pattern, not enthusiastic opinions about AI itself.
Spaced revisit
Return after seven days and rewrite the problem statement using the language users actually used. Note what assumption changed.
External video references
Keep Learning on YouTube
Open in a new tab
Design Thinking ExplainedEmbedded lesson videoIDEO's introduction to design thinking — the mindset and methods behind human-centred design.Watch · Watch on YouTube How Might We — Design Thinking MethodEmbedded lesson videoThe 'How Might We' technique for reframing problems into design opportunities — essential for the Define phase.Watch · Watch on YouTube
Ready to move on?
Mark this lesson complete and continue to "Context Engineering for Designers".