Skip to content
← All workKirti KumarNext →

UX Research · The Regency Town House × University of Brighton · 2025

Does a visitor trust an AI that makes things up?

Coding and synthesising 40+ visitor interviews from the public evaluation of REX — a retrieval-augmented virtual expert deployed in a real Regency drawing room.

My role
UX Researcher — evaluation
Context
Heritage AI · AHRC-funded
Method
40+ visitors · coded interviews
Output
Trust & design findings

[ 01 · The brief ]

A heritage AI that sounded certain even when it was wrong.

REX was deployed as a virtual expert in a working heritage house, answering visitor questions about the building using retrieval over a curated archive. The team needed to know whether visitors actually trusted it — and what happened to that trust when the model got something wrong.

My scope was the evaluation: designing the interview protocol, running sessions on the floor with real visitors, and coding the transcripts into a set of findings the build team could act on.

REX in the drawing room
Visitor-facing interface
Retrieval architecture

[ 02 · Process ]

What I did

  • Designed a semi-structured protocol covering expectation, verification behaviour and repair
  • Ran sessions with 40+ visitors in situ, immediately after their interaction with REX
  • Coded transcripts for trust signals, hesitation moments and abandonment triggers
  • Mapped findings against the retrieval architecture to locate where confidence was being manufactured

“Confident hallucination, not limited capability, was the real threat to trust.”

Headline finding, REX evaluation

[ 03 · Findings ]

What the evidence said

Visitors forgave a model that said it did not know. They did not forgive a model that answered fluently and turned out to be wrong — and once that happened, they stopped verifying anything else it said.

The interface was doing nothing to signal the difference between a grounded answer and an extrapolated one. Confidence was uniform, so trust collapsed uniformly.

Question framing
Feedback capture
40+
Visitor interviews coded
6 mo.
Field research on trust in AI
1
Headline finding that changed the build

[ 04 · Outcome ]

What changed

The findings fed directly into how the team surfaced provenance and uncertainty in the interface. The recommendation was not to make the model better — it was to make its confidence legible.

The lesson I carry into every AI product now: users do not need a model that is always right. They need one whose certainty they can calibrate against.

Reflection
[ Next case ]

An AI agent that fixes your storefront while you watch

Kirti Kumar © 2026Start a project →