AI can write, plan and act. The interface still asks us to squeeze intent, context and control into one text box. I built a pattern library to explore what happens when that work gets its own surface.
Plan the report. Make it concise. Use these sources. Ask before publishing.
THE WORK, MADE OPERABLE
01IntentReport outline
02ControlConcise · editable
03Evidence3 sources attached
04PermissionApproval required
Research → Framework → 80 prototypesScroll to unpack the work ↓
RoleResearch, framework, design and code
Timeframe2024—ongoing
Scope80 interaction patterns across five families
StatusPublic library Research ongoing
01 / STARTING POINT
A single field is doing six jobs.
For a quick question, chat is useful. For longer work, one field has to carry intent, constraints, context, feedback, control and revision. People must remember what the system knows, what it will do next and where they can intervene.
Six tools. One pattern.A blinking cursor in a box.
Claude
App builder
Perplexity
ChatGPT
Manus
DeepSeek
Examples shown in the original talk. Intent, context, constraints, control, feedback and revision all meet the same small field.
The design question became practical: which parts of AI work should be visible, steerable and checkable before they disappear into a prompt?
02 / DIRECT CONTROL
Change the work without rewriting the request.
The Property Panel turns part of a prompt into direct controls. This small illustration lets you change a sample response; the full prototype ↗ explores the pattern in more depth.
PROPERTY PANEL / ILLUSTRATION01—03
Draft a project update
“Summarise the status for the team. Keep the facts and dates intact.”
PREVIEW Changes with your controls
Example source: fictional status note
You can inspect the result before using it. ↗
The content stays explicit. Form, detail and evidence become decisions you can see and revise.
03 / CONTROL-AWARE UX
Control starts before the interface.
The research separates a principle from the process and the controls that carry it. That distinction matters when AI moves from answering questions to acting on someone’s behalf.
WHY Normative
Keep judgment with people.
AI can open possibilities. People should retain the authority to select, intervene and take responsibility for what is done.
HOW Process
Scaffold intent and delegation.
Structured choices and adjustable autonomy make the scope of a task clear before and during the work.
WHAT Interface
Make control operable.
Visible state, scoped actions and evidence give that principle a concrete place in the product.
PROCESS / AN EXAMPLE
Automation is a decision at every stage.
A trip planner can do substantial work before anything is booked. Explore where the interface can prepare, reveal and hand a decision back to a person.
Easy to reviseApproval before commitment
AI prepares · person can inspect
Gather routes and charging stops.
The system can collect options, while showing where they came from and when the information was checked.
INTERFACE MOVEVisible sources and freshness
Illustrative trip-planning scenario from the BMW UXCOMM talk. The degree of autonomy depends on the risk and reversibility of each action.
The 2025 essay asked whether an AI system could be observed, anticipated and directed. As the pattern library developed, explainability became a fourth question. DOPE is a design lens for the interface layer.
Select a circle to explore it. Trustworthy AI autonomy is the ambition at the center, not a measured trust score.
DDirectability01 / 04
Can I steer the work?
Give people a way to edit the plan, set boundaries and change course while the system is working.
Beyond Chat collects interaction patterns already appearing in products alongside emerging and speculative directions. Every entry pairs a concrete idea with a prototype, its use case, risks and examples.
80patterns in the catalogue 5families across the AI workflow
INLINE & CONTEXTUAL
Property Panel
Steer a result with visible parameters instead of writing the same instruction again.
I began with writing about autonomy in design, then studied seven designers’ AI-assisted work. The public library turned those questions into prototypes. Subsequent research and product-example audits exposed claims that needed to become more precise.
2024—25
Frame the question
Essays and a qualitative study with seven professional designers explored curation, autonomy and responsibility in AI-supported design.
2026
Build the vocabulary
Five pattern families, a shared schema and interactive demos made the ideas comparable across writing, visual work, agents, trust and memory.
2026 →
Audit the claims
Research reviews and a targeted audit of 70 example rows across 18 patterns checked whether references actually demonstrated the interaction described.
28 mature
40 emerging
12 aspirational
Evidence status of the 80 patterns — distinct from whether their catalogue pages are live.
WHAT YOU CAN SEEDisplayed trace
An account of a result that a person can inspect.
≠not a window into
WHAT YOU CANNOT VERIFY FROM ITInternal process
The route the model actually took.
Design response Treat traces as checkable claims. Pair them with sources and limits.
AN EXAMPLE OF A CORRECTION
“Reasoning Trace” needed a narrower claim.
A displayed chain of reasoning can be a useful account of a result, but it is not necessarily a faithful window into how a model arrived there. The pattern’s risk language was revised after the research review.
That correction matters: a pattern library should help teams ask better questions, including where its own examples are uncertain.
Beyond Chat is live as a public design resource: 80 catalogue entries across five families, interactive prototypes and research notes. The work also informed publications and talks on control in human–AI interaction.
The next question is empirical. A separate study platform has been built to compare chat with structured controls; participant testing and outcome data are still ahead. The library makes possibilities tangible. It does not, by itself, prove better performance or trust.
Beyond Chat.
Claude
App builder
Perplexity
ChatGPT
Manus
DeepSeek
SELECTED PATTERN PREVIEWS