Field Note · 003 iamhonee ↗

Amsterdam AI · #AI020 · The Netherlands

Where design met its next hard problem

Event — AI020 Amsterdam Researchers, ML engineers Industry seats: few
Scroll to enter
01 / 06
The AI020 venue entrance in Amsterdam — ornate brick building with AI020 banners flanking the doors.
The venue — a grand Amsterdam building, not a tech campus
The empty auditorium before the event — red velvet seats, Welcome to AI020 on the projection screen.
Before the crowd arrived — Welcome to AI020
Selfie in the auditorium — Honeylyn in front of the red seated hall with AI020 on screen.
One of the few product designers in the room
Audience from behind during a keynote talk — packed room, ornate ceiling, speaker at the front.
Researchers, engineers, and policy people — every talk grounded in real work
Hand holding an AI020 badge reading Honeylyn Tutor, Senior Product Designer, Harver.
Senior Product Designer, Harver — the identity I brought into the room

01 — Arrival

A different kind of room.

Amsterdam AI's AI020 was held in a grand venue in the city — ornate ceiling, red velvet seats, the kind of space that signals something is being taken seriously.

Not a tech campus. Not a design conference. Something else.

02 — Before the crowd

The quiet before a room fills with intent.

I arrived early. The seats were still empty, the screen read "Welcome to AI020," and I had a moment to register where I was — a room built around the question of what AI is actually for.

Keynotes from Prof. Max Welling and Prof. Henk Marquering. Not product launches. Research with weight.

03 — In the minority

Researchers, engineers — and me.

The crowd was mostly AI researchers, ML engineers, and policy people. Product designers were rare. I was probably one of very few.

That asymmetry was the point. I didn't come to present my perspective. I came to update it.

04 — The talks

Every talk was grounded in real work.

Welling's framing: not what to build, but what problems are worth solving. Marquering walked through AI's history — the rises, the winters, the hype cycles — and what it means that this wave feels different.

⚖️ Understanding where AI has been changes how you read where it's going.

05 — The identity I brought

Senior Product Designer. That badge matters.

The badge said Harver — an HR assessment platform where the people I design for grapple daily with accuracy scores, probabilistic outputs, and model confidence.

Tools like Claude or Figma handle the surface. The harder work is building mental models that help decision-makers trust — and interrogate — what an AI is actually telling them.

The gap isn't technical. It's about what humans do with AI outputs.

An accurate model that nobody trusts, understands, or can act on responsibly is not a solved problem. It's a design problem wearing a research coat.

That's where I want to work.

The challenge, stated plainly

What good AI product design actually requires.

good design = trust(model, decision-maker)
    + fairness(output, context)
    − hype(feature, timeline)

Trust requires legibility — not just accuracy. Fairness is a constraint, not a feature. And hype is the thing that gets subtracted when real decisions are on the line.