Signals

❝

AI-Native Behavior · 2025

The Prompt Is Now the Interface — and Nobody Designed It

Users who know how to prompt well receive a fundamentally better product than users who don't. This isn't a learning curve. It's a capability gap baked into the interface itself. The people who understand how AI thinks will compound advantages over time. The people who don't will receive a degraded experience while believing they're using the same product.

This is the new accessibility problem. And unlike color contrast or screen reader support, it has no checklist fix. The design challenge has shifted from visual hierarchy to conversational priming — and most teams haven't noticed yet because their power users are also their best prompters.

Video Essay · On Stage

Embed your talk or presentation video here

01

Digital Identity · Teens

Avatars as Aspirational Infrastructure

For teenagers, the avatar is not a representation of who they are — it's a prototype of who they're trying to become. The wardrobe choices, the display name, the accessories: behavioral experiments in social signaling before the physical identity is settled. Platforms that treat this as cosmetic are missing the emotional architecture they've been handed. The avatar is doing the psychological work of trying on a self before committing to one.

Podcast · Short-form audio

Add a podcast clip or audio note here

00:00 — short-form thinking

This card is set up for a podcast embed or SoundCloud clip. Drop in an <iframe> from Anchor, Spotify, or SoundCloud to embed a short audio take on any of the topics on this page.

02

Virtual Economies

Ownership Without Scarcity Produces Anxiety, Not Abundance

Virtual item economies that remove scarcity entirely don't produce satisfied users — they produce users who can't form attachment to anything. The psychological value of owning something is inseparable from the possibility that others might not have it. Designing for belonging means designing for exclusion, at least partially. Most product teams are not ready to have that conversation with their stakeholders.

03

Organizational Behavior

Why Research Gets Weaponized Instead of Used

In most organizations, research is deployed after the decision is made. To confirm it, defend it, or deflect accountability for it. The structural problem: research is positioned as a deliverable rather than a decision-making process. Teams that use research well are the ones where the researcher is present before the question is finalized — not after the solution is already being built.

Video · Research Observations

Drop a short-form video take here

04

Emerging Technology

The Trust Debt Accumulates Before Anyone Notices

Automation features fail not because the algorithm is wrong but because the transition to trusting it was never designed. Users who encounter one unexplained result that violates their mental model don't recalibrate — they disengage. Trust is not built by proving correctness. It is built by narrating uncertainty honestly. This is the design problem most AI product teams are currently losing.

05

Creator Economies

The Line Between Creator and Character Is Dissolving

Audiences no longer expect authenticity — they expect well-executed performance. On platforms where creators become brands and brands hire creators to become characters, the concept of authentic voice is under structural pressure. The behavioral shift: parasocial relationships have evolved to accommodate the performance. The emotional reality hasn't changed. Only the transparency has.

❞

Social Signaling · Consumer Behavior

The New Flex Is Intentional Restraint

Across demographic cohorts showing up in research, the aspirational signal is shifting from maximalism to deliberate reduction. Fewer platforms, curated feeds, physical objects with clear provenance. The conspicuous display of curation is replacing the conspicuous display of volume. This has direct implications for how young consumers will approach virtual ownership in the next five years — and it means platforms designed for accumulation may be optimizing for a value system that's already shifting.

06

Research Practice

Empathy Is Not Enough — Context Is the Variable

Every team says they're user-centered. The distinction is whether the researcher has access to context, not just stated needs. The most consequential insight in the Optym RouteMAX research came not from what planners said they wanted — but from understanding what they were afraid to lose. That's a context question. Empathy gets you to the surface. Context gets you to the structure underneath.

07

AI & Identity

Personalization Is Flattery — and Has the Same Risks

AI systems that personalize at scale are reflecting users' preferences back to them as if those preferences were surprising discoveries. This produces engagement and a narrowing of the behavioral repertoire that users rarely notice and almost never consent to. The experience feels like being understood. The mechanism is closer to being mirrored. These are very different things with very different long-term behavioral effects.

↗

Future of Work · UX Research · 2025

The Research Role That AI Can't Replace — But Will Completely Reinvent

The basic functions of UX research will be automated. Survey design. Participant recruitment. Affinity mapping. Synthesis of large text corpora. These are not ten years away. They're already here. What remains — what cannot be handed to a model — is the judgment work. The ability to recognize when the question itself is wrong. The organizational influence required to make a finding actionable. The capacity to sit in a room with stakeholders who want a specific answer and deliver a different one with enough credibility that it changes the decision.

The researchers who will matter in the next decade are not the ones who run studies. They are the ones who design research systems — infrastructure that makes behavioral intelligence continuous rather than episodic. They are the connective tissue between organizational silos. They are the people who ask "what problem are we actually solving?" before the team commits to the wrong sprint.

AI will make research faster. It will make shallow research much cheaper. The premium will shift entirely to the judgment layer — the layer where behavioral strategy lives. This is not a warning. It is a map.

Published · Optym Blog

Organizational Culture · Nov 19, 2019 · 7 min read

How Empathy Changed the Product Conversations in My Team

Empathy is a very powerful tool that designers use to make effective design decisions. A simple attempt to have meaningful design conversations led to a change in mind-set and affected the product's decision-making culture.

How are we asking different questions? From "Let's do ABC. Does that work?" to "XYZ is the problem. How can we provide a feasible solution quickly?" How are we making product decisions? From "Client said Option 1 looks nice. Let's do it." to "The users found it difficult to interpret. Let's test our medium effort scenario."

Sixteen eight-hour shadowing sessions. Eight different locations. Four days. Templates for note-taking. Training in unbiased questioning. And in the end — fewer design review sessions, but more engaging ones. Because we were talking about user problems, not features.

Read full piece ↳ See the Optym case study this grew from

Published · Medium

Conversational Design · Mar 8, 2018 · 5 min read

UX Best Practices for Conversational UI

Two months of intensive research building a chatbot for GameOn Technology. Multiple design iterations and critique sessions. Here's what the usability work revealed about designing better bot experiences.

Without a personality, a chatbot is just an online form with higher usability issues. Trust tax is real: one undesired interaction loses three times the trust that one desired interaction can build. Design for failure — how you handle out-of-scope conversations is where you lose your users.

Read full piece