The taxonomy
Patterns
Thirty patterns across the arc of an AI interaction — from the first prompt to the hundredth session. Each one is dissected the same way: what it is, when to reach for it, how it fails.
01 / How people find their way in
Invitation
The first moments decide whether AI reads as capable or as a gimmick. Invitation patterns set expectations, seed the first prompt, and make an empty text box feel like a place to start rather than a test.
PTN—01
Prompt starters
Curated example prompts on the empty state that teach range and get people to a first success fast.
Live specimen
PTN—02
Contextual entry points
AI affordances placed at the moment of need inside existing workflows, instead of a separate destination.
PTN—03
Capability disclosure
Honestly communicating what the system can and cannot do, before the person finds out the hard way.
PTN—04
Suggested actions
System-proposed next steps offered mid-flow, turning each response into a fork the person can take with one tap.
PTN—05
Guided first run
A structured first session that produces a real result from the person’s own context instead of a tour.
PTN—06
AI identity
The AI’s name, avatar, and visual treatment as an expectation-setting system — recognizably synthetic, consistent everywhere, never a fake human.
Live specimen
02 / How people say what they want
Expression
Free text is the most flexible input ever shipped and the least discoverable. Expression patterns give intent structure — context attachment, constraints, modes — so people can steer without learning to prompt.
PTN—07
Context Attachment
Let people ground a request in specific material — files, selections, screens — instead of describing it.
Live specimen
PTN—08
Constraints & Tuners
Expose the request’s key parameters as explicit controls instead of burying them in prose.
PTN—09
Modes
Named operating states that trade speed, depth, or autonomy — chosen once, applied to everything after.
PTN—10
Prompt Refinement
The system improves the request itself — clarifying, expanding, or rewriting — before spending effort on the answer.
PTN—11
Structured Input
Replace part of the free-text prompt with fields, templates, or forms that guarantee the request arrives complete.
PTN—12
Follow-up questions
When a request underdetermines the task, ask one specific clarifying question — with tappable answers — instead of guessing.
Live specimen
PTN—13
Model management
Frame model choice as a tradeoff people can reason about — fast versus thorough — not a benchmark menu.
Live specimen
PTN—14
Style presets
Named, reusable voice and tone presets — including user-saved ones — instead of re-prompting “make it friendlier” every time.
Live specimen
03 / How the system shows its work
Response
AI output is not a page load; it unfolds. Response patterns shape the unfolding — streaming, progress, working transcripts, and the artifacts that outlive the conversation.
PTN—15
Streaming Response
Render output token by token as it is generated, so the wait becomes part of the answer.
Live specimen
PTN—16
Working Transcript
A live, inspectable log of the steps an AI takes — searches, tool calls, files touched — while it works.
Live specimen
PTN—17
Artifacts
Promote substantial outputs — documents, code, designs — out of the chat stream into standalone, editable objects.
PTN—18
Progressive Disclosure
Lead with the answer at skimmable depth and put reasoning, sources, and detail one deliberate step away.
PTN—19
Stage and Conversation
Split the interface into a persistent workspace for the work and an adjacent chat for directing it.
PTN—20
Variations
For subjective output, generate labeled alternatives instead of forcing serial regenerate-and-pray.
Live specimen
PTN—21
Targeted regeneration
Regenerate the sentence that’s wrong, not the paragraph that’s right — redo scoped to a selection, like inpainting.
Live specimen
PTN—22
Draft mode
A cheap, clearly-labeled draft pass for exploration; the expensive polished pass is an explicit upgrade.
Live specimen
04 / How the system earns belief
Trust
Trust is built structurally, not claimed. These patterns expose sources, communicate uncertainty honestly, and make sure people always know when they are talking to a machine — and when it is guessing.
PTN—23
Citations
Attach retrievable sources to specific claims so people can verify the answer instead of trusting the vibe.
Live specimen
PTN—24
Calibrated confidence
Express uncertainty at the strength the evidence supports, so hedges and firm answers both mean something.
Live specimen
PTN—25
AI self-identification
Make it unambiguous when a person is interacting with AI or reading AI-generated content.
PTN—26
Showing limits
State what the system cannot do — scope, staleness, blind spots — before people discover it by failing.
PTN—27
Verification affordances
Build checking into the interface — diffs, previews, traces, one-tap source access — so verifying is cheaper than trusting blindly.
PTN—28
Context budgets
Show the context window as a visible budget — per-item cost, remaining headroom, and a remedy — before quality silently degrades.
Live specimen
PTN—29
Content provenance
AI-generated content carries a human-readable origin marker that survives export and records the human’s edits.
Live specimen
PTN—30
Data boundaries
Make consent, retention, and incognito first-class interface — itemized, per-use controls instead of a buried all-or-nothing toggle.
Live specimen
05 / How people stay in charge
Control
The more an AI can do, the more the interface owes the person a firm grip: approvals before consequences, reversibility after them, and a way to interrupt in between.
PTN—31
Approval gates
Pause before consequential actions and present a reviewable proposal the person must explicitly approve or reject.
Live specimen
PTN—32
Reversibility
Make AI actions cheap to undo, so people can accept output provisionally instead of auditing it exhaustively.
PTN—33
Interrupt and steer
Let people stop or redirect the AI mid-flight without losing completed work or starting the task over.
Live specimen
PTN—34
Human-in-the-loop escalation
Detect when a task exceeds the AI’s competence or authority, and hand it to a person with full context.
PTN—35
Scoped autonomy
Give the AI explicit, adjustable boundaries — what it may do alone versus what requires a person.
PTN—36
Action plans
Before executing a multi-step task, show the plan — steps, order, consequences — and let the person approve or edit it.
Live specimen
PTN—37
Checkpoints & branches
Give long AI sessions save-points: restore to before a bad turn, or branch to try an alternative without losing the original.
Live specimen
06 / How the relationship compounds
Adaptation
A system that learns changes under people’s feet. Adaptation patterns make memory legible, personalization inspectable, and feedback worth giving.
PTN—38
Legible memory
Show people exactly what the system remembers about them, in plain language they can read, edit, and delete.
PTN—39
Inspectable personalization
When output is shaped by a profile, let people see why, adjust the levers, and preview the unpersonalized version.
PTN—40
Feedback loops
Collect user feedback at the moment of reaction, then visibly change behavior because of it.
PTN—41
Graceful degradation
When the AI can’t perform — offline, rate-limited, low confidence, out of scope — fall back to something still useful.
PTN—42
Relationship continuity
Carry context, preferences, and unfinished work across sessions so every return doesn’t start from zero.