The Far Side

Selected Artifacts from the Age of Convenient Unknowing

The Institute for Applied Forgetting Temporary Exhibition, Gallery 4 April 5 — June 30, 2026

Curatorial Statement

There was a time when not knowing something cost you. It cost effort, or time, or at minimum the small embarrassment of admitting the gap. That time is difficult to locate precisely, because the transition away from it has been so comfortable that most of us did not notice it happening. The drift was not dramatic. Nobody woke up one morning having forgotten how to think. It was more like the way a river changes course over decades: each day’s erosion is invisible, and then one afternoon you look up and the water is somewhere else entirely.

This exhibition documents that drift — not as decline narrative, not as moral panic, but as natural history. The artifacts gathered here represent various forms of not-knowing, both chosen and imposed, both celebrated and catastrophic. Some were donated. Others were recovered. A few were found in the state they were left in, which is to say: abandoned but still running.

Visitors are encouraged to leave their devices in the coatroom lockers provided. This is optional. We note, however, that the lockers have been the most popular feature of the exhibition to date.

Catalog of Works

Item 1. Two Compression Studies (2026) Mixed media: terminal printout, framed; software repository, archived on USB; personal journal excerpts, redacted Display case, north wall

The left frame contains a terminal session in which a large language model has been instructed to speak like a caveman. “Brain still big,” reads one line. “Caveman make mouth smaller.” The project’s author, a researcher in the Netherlands, discovered that stripping away conversational filler reduced token output by roughly seventy-five percent while, in certain benchmarks, actually improving accuracy. Less language, better thinking. The finding was counterintuitive enough that a separate academic paper was written to explain it, titled “Brevity Constraints Reverse Performance Hierarchies in Language Models.”

The right side of the case displays excerpts from a developer’s account of a project that sat dormant in his mind for eight years. He wanted to build proper tooling for SQLite — a formatter, a linter, a language server — but the work required parsing a grammar of some four hundred rules with no formal specification. It would have taken years. Then, in late 2025, he sat down with an AI coding agent and built it in three months.

The journal excerpts, displayed between the two objects, document what happened next. The first version worked but was, in his words, “complete spaghetti.” He could not understand his own codebase. He threw it all away and started over, this time making every decision himself and using the AI as — his phrase — “autocomplete on steroids.” He described the experience of prompting as addictive, comparing it to slot machines. “Just one more prompt,” he wrote, late at night, knowing it probably wouldn’t work.

The two objects are displayed as companion pieces because they document the same phenomenon from opposite ends. One compressed language and found clarity. The other compressed time and found confusion. Both arrived at the same question: what is the minimum you can say and still mean something? The caveman answer is surprisingly little. The developer’s answer is: more than you think, because the things you leave out are the things you needed to learn.

Acquired 2026. Donors: J. Brussee, Utrecht; L. Mishra, London.

Item 2. Vitrine of Involuntary Absences (2026) Found objects: browser screenshot, preserved under UV-filtered glass; product packaging, original condition; handwritten customer support transcript

The screenshot on the left shows the last authenticated session of a Google Workspace account belonging to a small business owner in Australia. On April 4, 2026, while traveling in the United Kingdom, the owner removed a phone number from the account to avoid roaming charges. Google’s system interpreted this as a security threat — conflating the recovery phone with the authenticator itself — and suspended the account. DNS verification was completed. It made no difference. Support representatives offered solutions that required logging in to the suspended account. The owner’s payroll, CRM, calendar, and email were inaccessible for thirty days.

To the right, in its original packaging, sits an Iguanaworks USB IR Transceiver (Dual Socket model, late production run). Iguanaworks manufactured infrared control devices for computers — small, precise instruments that allowed a Linux machine to talk to a television or a stereo. The company closed in early 2026. The announcement was a single sentence on their product page: “Iguanaworks has closed and our products are no longer sold.” No explanation was offered. The discussion thread that surfaced when someone noticed had fifteen comments.

The curatorial decision to display these objects together was deliberate. Both document disappearance: one sudden and devastating, the other so gradual it was almost invisible. The Workspace suspension generated hundreds of comments, expressions of fury, advice, solidarity, fear. The Iguanaworks closure generated a polite murmur. Yet the structure of loss is identical: a thing that worked, that someone relied on, that one morning simply wasn’t there. The exhibit invites visitors to consider which kind of disappearance is worse, and whether the answer reveals more about the object or the observer.

A handwritten transcript between the two displays records a commenter’s memory of Google’s early hardware support program. In 2013, Google Glass owners received a toll-free number staffed by human experts, available around the clock. Broken devices were replaced immediately, free of charge, no questions asked. “Like I said,” the commenter wrote, “once upon a time.”

Acquired 2026. Recovered materials; anonymous donation.

Item 3. Two Chambers for Voluntary Discomfort (2026) Installation: reconstructed sauna interior (cedar, electric heater, 73 degrees Celsius); printed menu from unnamed establishment; clinical data printout; wall-mounted signage

The left chamber is a reconstruction of a Finnish sauna of the type used in a 2026 study at the University of Jyvaskyla. Fifty-one adults sat in this heat — seventy-three degrees Celsius — for thirty minutes. Blood draws before and after revealed that the session triggered a significant immune cell response: the body’s defenses mobilized as though confronting an actual threat. The heat was not comfortable. It was not supposed to be comfortable. The researchers found that the cellular response was stronger than the cytokine response — meaning the body was dispatching actual defenders rather than merely sounding an alarm. Stress, applied correctly, was medicine.

The right chamber is a reconstruction of a phone-free bar. A printed menu is displayed under glass; beneath it, a placard explains the establishment’s policy. Patrons check their devices at the door. There is no WiFi. A sign reads: “We do not have WiFi. Talk to each other. Pretend it’s 1995.”

Both chambers require visitors to surrender something in order to enter. The sauna demands physical comfort. The bar demands connectivity. What each returns is harder to name — the study calls it “immune activation,” the bars call it “atmosphere” — but both operate on the same principle: that certain forms of deprivation are generative. That some things can only grow in the space left by what you’ve removed.

A comment from the sauna study’s discussion thread, printed on the wall between the two chambers, reads: “Makes me wonder how much of it is sauna, versus just the luxury of having the time to go do nothing for thirty minutes.”

The exhibit does not answer this question.

Installed 2026. Materials sourced from Helsinki and various American cities.

Item 4. The Awe Gap (1876 — 2026) Audio-visual installation: mission audio recording, loop; archival photograph; live comment feed, redacted; preserved specimen jar (empty)

The central screen plays a segment of mission audio from Artemis II. Four astronauts are approaching the far side of the Moon — the hemisphere permanently turned away from Earth, seen by human eyes only a handful of times in history. The audio is ordinary in the way that all extraordinary audio is ordinary: professional voices noting coordinates, confirming procedures, and then a pause. Someone says something about the view. The mission continues.

Below the screen, a live feed displays comments from the discussion that followed the BBC’s coverage. A significant portion of the thread is devoted to clarifying that “far side” and “dark side” are not the same thing. The far side receives sunlight; it is not dark. It is merely unseen. Several commenters express frustration that a forum ostensibly populated by scientists and engineers cannot sustain a thread about lunar exploration without devolving into pedantry or politics. One commenter writes: “There are too many problems here on earth for me to get excited about a trip to the moon.” Another writes: “It makes me tear up seeing the absolute best of us.”

The specimen jar to the left of the screen is empty. Its label reads: Anguilla anguilla, reproductive material, specimen pending. For over a century, no one could determine how the European eel reproduced. No one had observed eel mating. No eggs had been found in the wild. Aristotle believed they emerged spontaneously from mud. Sigmund Freud, as a young student, dissected hundreds of eels looking for reproductive organs and failed to find them conclusively. The mystery persisted until 2022, when researchers finally tracked eels to their spawning grounds in the Sargasso Sea using satellite tags small enough to survive the journey.

The jar remains empty because the exhibit is not about the answer. It is about the century of not-knowing that preceded it, and the particular quality of attention that sustained the search. For a hundred years, the eel’s reproduction was a far side of its own: real, illuminated, functioning perfectly well, but facing away from every instrument we pointed at it.

The installation asks visitors to sit with a question the curators have not been able to resolve: whether the experience of awe requires the possibility that the thing you’re looking at might never be explained. And if so, what happens to awe in an age when the prevailing assumption is that everything will be explained eventually, and probably by Thursday.

Audio courtesy NASA. Specimen jar on indefinite loan from the Natural History Museum, Vienna. Comment feed updates hourly.

Item 5. This Catalog (2026) Printed matter, staple-bound, 1 edition

You are holding it. Or scrolling through it, which is not the same thing, though the difference is becoming harder to articulate.

This catalog was assembled in the manner of all institutional publications: with the quiet confidence that organizing things into categories produces understanding. Each artifact has been numbered. Provenance has been noted. Curatorial statements have been drafted in the appropriate register. And yet the curators wish to note, in this final entry, that the act of cataloging these objects has not made them less strange.

The developer who threw away three months of work understood something that the caveman prompt only hinted at: that compression has a cost, and the cost is not always legible from the outside. The business owner locked out of a Workspace account and the small hardware company that vanished without explanation occupy the same display case because the curators could not determine a principled reason to separate them. The sauna and the bar are placed together because both charge admission to a room where you lose something, and both have waiting lists.

And the astronauts continue around the far side of the Moon. The mission is ongoing. The comments are still being written. Someone, somewhere, is explaining the difference between “far” and “dark” to someone who did not ask.

The gift shop is closed. The lockers, we are told, are still available.

Curated by Dr. Maren Visser, Senior Fellow, Institute for Applied Forgetting, whose previous exhibitions include “Load-Bearing Assumptions” (2024) and “The Inventory of Things We Stopped Counting” (2023). She was unavailable for comment at the time of printing.