AI philosophy · Today

AI philosophy in practice

AI is a superpower, not a replacement for deep thinking. As an ever-evolving tool, AI is able to broaden thinking and speed up the process, but always kept in check with design intuition + data.

Project lifecycle
Hover a phase to see example prompts
Tap a phase to see example prompts
What users or use cases am I not accounting for?
Organize all the research. Extract primary versus secondary user problems.
Using the existing patterns in our product, what are different ways we can impact x goal?
Use ideas from 1 + 3, along side these sketches to create a new concept
Use the following data set and workflow, along with the leading concept to build a prototype
Iterate on the proto with the following feedback in mind. Estimate time saved per task compared to the old flow
Are there any other solution ideas we should consider? What project assumptions were false?
Write documentation for each new concept and pattern and create a retro board in FigJam.
Research + planning

Supplemental opinion and organization for research / problem synthesis

Idea proliferation

Explore wider and deeper before narrowing the solutions, at 3x speed

Prototyping + executing

Build loops and artifacts that users can touch and use for quick validation

Analyzing results

A second pair of eyes on results and analysis. Retro the project and work

AI philosophy · What's next

One designer. A team of agents.

The next leap isn't a smarter tool — it's a better team. Each designer leads a crew of specialized AI agents across research, brainstorming, iteration, development and shipping. You set the direction and make the calls; an orchestration layer routes the work and brings it back for your review.

Agentic design team
Hover over each category to see agent examples
You
The designer
Sets intent · judges quality · makes the call
Orchestration · shared context
Research
Interview synthesizer
Competitive scout
Analytics analyst
Brainstorm
Concept generator
Journey mapper
Devil's advocate
Iterate
Prototype builder
Design critic
Usability tester
Develop
Spec & handoff writer
Component coder
Visual QA
Ship
Release & comms
Experiment monitor
Feedback listener
● Specialist agent◌ Orchestration — routes tasks, keeps context, returns work for your review

Human at the center

Agents expand what one designer can explore and execute. Taste, judgment and accountability stay with the person.

Shared context

Every agent works from the same design system, research repository and product data, so output stays on-brand and on-strategy.

Faster loops, same bar

Parallel agents compress the cycle from insight to shipped product, with a human checkpoint at every gate.