PUBLIC BUILD EXHIBITION
2026—
I NOTICE
WHAT'Smissing.
Then I build it.
I’m building 100 useful things from questions I keep running into across technology, law, neuroscience, place, research, work, and everyday life.You do not need to be a technologist, lawyer, scientist, or expert to understand or use what is here.
Every numbered build has two parts:
A / THE THING — a working tool, system, experiment, or experience you can use
B / THE STORY — a visual build that makes the problem, idea, or hidden system easier to see
I publish the process too: what I tried, what failed, what the evidence changed, what you can reuse in your own work, and what each build leaves behind for the next one.
001-A
Can a human really stop or change an AI-assisted decision?
Come for whatever is useful.
Try working tools and experiments. The first path is written for people who do not already know the field.
See motion, maps, diagrams, data stories, film, and creative code that explain why each build exists.
See what changed, what failed, the evidence behind decisions, reusable methods, and how later projects inherit earlier work.
Built across disciplines, with boundaries visible.
RN Collins is an interdisciplinary systems builder and law student with graduate training in anatomy, neurobiology, developmental psychology, and medical education.
Released work, lab builds, synthetic rehearsals, evidence limits, and incomplete archive layers are labeled separately. A working interface is not presented as legal, medical, scientific, or operational authority.
For research, systems architecture, public-interest technology, or a closer look at the method, contact RN.
What this exhibition actually is.
Each build starts from a question that turned out to be too messy for a single discipline to answer on its own. The question comes first and the artifact second, which is why builds sit next to each other that would never share a portfolio category: a review-authority framework, a legal-workflow decomposition, a place-feasibility model. What they have in common is the shape of the problem rather than the subject matter.
A working tool proves an idea survives contact with real input. A visual build makes the argument legible to someone who has no intention of operating anything. Most work of this kind ships one and gestures at the other, and the gap is where the reader loses the thread. Building both is the point rather than a presentation choice.
The builds are numbered because each one is meant to inherit from the ones before it. A build that produces a validated entity model hands that model to the build that has to decide when two records describe the same thing. The lineage view traces those inheritances as concrete capabilities rather than themes, so the sequence can be checked instead of asserted.
How to use it without reading all of it.
Open the exhibition and pick any build whose question you recognise from your own work. Use the A side. Nothing assumes you have read anything earlier, and every build is written so a first-time visitor can reach a judgement without background.
Go to a public build record instead. It states what changed, what another person can reuse, what evidence supports the system and what the build is not entitled to claim, in that order, on one page. It is the fastest way to see how the claims here are bounded.
Status labels are load-bearing. A build marked as a lab prototype or with its evidence review open is saying something specific about what has been checked, and those labels are kept accurate in preference to being kept flattering. Where a record is incomplete, the page says so rather than omitting the layer.