The word remediation comes from the Latin root for healing.
Elian knows this the way he knows most things that are not directly useful to him, peripherally, the way you absorb information that passes through your environment often enough to leave a residue. He knows it because he has been typing it into documents and status updates and client-facing briefs approximately forty times a day for the past week and at some point the word stopped looking like a word and started looking like a shape, and when words stop looking like words he tends to look them up, not out of curiosity exactly but out of the particular discomfort of using something he does not fully understand.
Healing. That is what they are calling it. Remediation. As though the interfaces are patients and the null-percept is a wound and the work he has been doing for seven days without a break longer than four hours is medicine.
He is not sure it is medicine. He is not sure what it is. He does his work anyway because the work is there and because not doing it would mean sitting with the alternative, which is thinking about what Vera Cho showed them in that conference room and what he found afterward in the full log and what it means that the seam was already open when he noticed it, and he has filed those questions in the part of his mind where the load-bearing questions wait, and he is not ready to approach them yet, and work is the most reliable way he knows to keep a door closed until he is ready to open it.
He opens the logistics component.
The orbital freight platform client sent a message at six this morning, his time, which means someone at the platform sent it at what would have been the middle of their operational cycle, which means someone was awake and at their desk and troubled enough by what they were seeing to write a formal message rather than waiting for a scheduled check-in. The message is three paragraphs. The first paragraph describes the problem in client language, which is the language of impact and concern rather than the language of mechanism. The third paragraph asks for a timeline. The second paragraph, the one Elian reads twice, describes what the operators are actually experiencing, which is that the depth shadow indicators in the spatial navigation interface are rendering in the null-percept and the operators are reporting that the null-percept does not read as depth to them the way black did. They are not misreading the indicators. They are reading them correctly and the reading is not producing the spatial understanding it used to produce. The connection between the visual signal and the cognitive model of three-dimensional space that the operators depend on for safe cargo routing is not functioning as designed.
He sits with this for a moment. Then he opens a new component file and begins to think.
The problem is not the null-percept rendering. He implemented that correctly last week, the systematic replacement of every instance of the zero-value color with the null-percept variable, and the rendering is consistent and technically accurate. The problem is what the null-percept does in a depth-shadow context that black did not do, which is the thing he wrote in his investigation file and has been sitting with since, the quality of threshold rather than terminus. Black said this is the furthest point. The null-percept says this is where known things end. For a person looking at a painting or a user interface on a personal device, the distinction between those two communications is philosophical. For an operator routing cargo trajectories through a three-dimensional spatial interface at a platform where a misread depth indicator has physical consequences, the distinction is operational.
He needs to redesign the depth shadow system entirely. Not patch it. Redesign it from a different starting point, one that does not depend on the dark end of the color spectrum to communicate terminus because the dark end of the color spectrum no longer communicates terminus.
He opens a call to Riku.
It takes four rings, which for Riku means he was not at his desk, which is unusual enough that Elian notes it. When Riku's face appears in the projection above his secondary monitor he is not in his usual workspace. The background is unfamiliar, a narrower room with visible cable management infrastructure along the left wall, and he is eating something, which he stops doing when he sees Elian's expression.
Riku says what happened.
Elian says nothing happened. He says he needs to talk through a depth system problem and wants Riku's read on the back-end architecture implications before he commits to a redesign approach.
Riku looks at him for a moment with the expression he wears when he is deciding whether someone is telling him the complete truth. Then he sets down whatever he was eating and says go ahead.
Elian explains the freight platform situation. He explains the threshold versus terminus distinction that he has been developing in his notes. He explains that his current thinking is to move away from shadow-based depth communication entirely for the spatial navigation component and redesign around a dual-axis system using luminosity gradient on one axis and geometric scaling on the other, both of which communicate depth through mechanisms that do not rely on the dark end of the spectrum at all.
Riku is quiet for about ten seconds, which is a long time for Riku.
Then Riku says that the luminosity gradient approach will work for depth but it will create a conflict with the existing alert overlay system, which uses luminosity gradient to communicate urgency, and if the same visual mechanism is carrying two different kinds of information in the same interface at the same time the operators will spend cognitive resources resolving the ambiguity and that cognitive load is exactly what you do not want in a high-stakes operational environment.
Elian says he knows. He says he was thinking the alert system could be migrated to a motion-based signal, a subtle pulse on the alert elements rather than a luminosity change.
Riku says that motion-based alerts are problematic in spatial navigation interfaces because motion in peripheral vision triggers a specific attentional response in human cognition that pulls focus away from the primary task. He says Sable has data on this from the neural interface research they have been running since Event Zero. He says he can send it.
Elian says please.
Riku sends it. Elian opens it in a secondary window while Riku watches him read, which Elian is aware of and does not comment on because Riku watching him read is Riku's version of interest and is not uncomfortable once you understand that it is not surveillance but engagement.
The data is useful. It is more than useful. It is exactly the thing he needed to read because it closes off two approaches he was considering and opens a third one he had not thought of, which is to use the null-percept itself as the depth communicator not through shadow but through saturation gradient, running the null-percept at varying saturation levels across the depth field so that the furthest elements are most saturated in the null-percept and the nearest are least saturated, which means the null-percept's threshold quality becomes directional rather than terminal, it communicates the far end of the field as the edge of the known rather than the edge of the possible, and an operator whose nervous system reads threshold rather than terminus in the null-percept will read a saturated null-percept as the place where the interface's knowledge runs out which is precisely what the far end of a cargo trajectory field is.
He says this to Riku, quickly, before the idea loses its shape.
Riku listens without interrupting, which is his highest form of engagement. When Elian finishes Riku says that the saturation gradient approach is technically implementable on the back end without significant architecture changes because saturation is a property of the color variable rather than the color system itself, which means it can be added as a parameter to the null-percept variable without touching the underlying rendering architecture.
Elian says that is what he thought but he wanted to confirm before spending three days on a frontend implementation that hits a backend wall.
Riku says three days is optimistic.
Elian says he knows but he did not want to say five days in the client brief because five days would produce a follow-up call and he is not currently allocating time for follow-up calls.
Riku almost smiles. He says Sable can model the saturation parameter range for optimal depth legibility if Elian wants to send the component specifications. He says it will be faster than Elian running the tests manually.
Elian says yes and sends the specifications and thanks him.
Riku says he does not need to be thanked for being efficient. He says it as though thanking him is a minor social imposition rather than a normal human exchange and Elian has known him long enough to understand that this is accurate and not a performance. He closes the call.
The Sable modeling results arrive forty minutes later while Elian is working through the surgical systems client, which is the more urgent of the two critical remediations because the hospital administrator's follow-up message this morning used the word suspended three times in two paragraphs and each use of the word carried a weight that the previous correspondence had not had, the weight of a person describing a situation they are managing and that they are managing at a cost they have not fully calculated yet.
The surgical systems problem is different from the freight platform problem in a way that matters for the design approach. The freight platform operators are navigating a digital representation of physical space, a spatial model of cargo trajectories that is constructed from data and rendered as a three-dimensional environment. The depth they need to read is the depth of the model. The surgical teams are using an augmented overlay, a digital layer imposed on actual physical space, actual tissue, actual surgical field. The depth they need to read is the depth of reality.
This distinction means that the saturation gradient approach Elian developed for the freight platform cannot be directly applied to the surgical overlay because in an augmented reality context you cannot saturate the overlay elements independently of the physical environment beneath them without creating a visual conflict between the overlay and the reality it is augmenting, and that conflict in a surgical context would be actively dangerous in a way that a poorly communicated cargo depth is not.
He opens a message to Sena because this is the kind of problem that needs the client-facing layer before it needs the technical layer.
He writes that the surgical systems remediation is going to require a different approach from the freight platform work and that the different approach is going to take longer to develop safely and that someone needs to manage the hospital administrator's timeline expectations before he has a solution he is confident in, and that he is not the right person to do that managing because he will tell them honestly that he does not know yet how long it will take and that is not the information they need right now.
He sends it and goes back to the surgical overlay problem.
Sena's response arrives eleven minutes later. She says she will handle the hospital contact and that she needs from Elian one thing, which is his honest assessment of whether the issue is fixable in principle or whether it requires a fundamental redesign of how augmented surgical overlays handle depth communication in a world without black.
He stares at this for a moment. It is a good question. It is the question he has been circling around since he opened the surgical systems file this morning and has not been ready to answer because answering it requires going to the place where the load-bearing questions are and he has been keeping that door closed.
He opens it, briefly, and looks at what is behind it.
Then he writes back. He says the issue is fixable in principle. He says the fundamental mechanism of augmented surgical depth communication does not require black specifically, it requires a visual signal that the human brain in a high-focus operational state will read as spatial information without ambiguity and without cognitive overhead. Black performed that function because it was the most neutral and unambiguous dark value available, carrying no associations in a clinical context except depth and shadow, which are exactly the associations a surgeon needs. The null-percept performs a different function. It is not neutral. It communicates something. He has been trying to determine whether what it communicates can be redirected toward depth legibility in an augmented context without introducing new ambiguities and he does not yet have an answer.
He says he needs two days.
Sena responds in four minutes. She says she will tell the hospital administrator two days and she will make two days sound like a gift rather than a limitation and she will need him to actually have something in two days.
He says he will have something in two days.
He closes the message thread and looks at the surgical overlay component.
The null-percept in an augmented context does not work the way it works in a purely digital environment because in a purely digital environment the null-percept is one value among many, surrounded by other values that give it context and allow the brain to calibrate what it means. In an augmented surgical overlay the null-percept elements are floating above real tissue under real lights in a real room and the calibration context is entirely different. The brain is not just processing the overlay. It is processing the overlay and the reality beneath it simultaneously and trying to build a single coherent spatial model from two different kinds of visual information and the null-percept landing on top of real tissue in real light does something to that process that he has been observing in his test renders for three hours and cannot fully characterize yet.
He sets up a new test environment and begins running systematic variations on the overlay's depth communication system, changing one variable at a time, documenting the output of each variation with the same patient attention he brings to shadow lag problems, the same pre-verbal looking that operates below the level of analysis, waiting for the seam to reveal itself.
The first six variations produce outputs that are technically functional and wrong in ways he cannot immediately quantify. The seventh variation, which introduces a thin geometric border on depth-indicator elements rather than relying on color value alone to communicate their spatial position, is less wrong. He notes this and keeps going.
The sun outside his window, visible through the narrow horizontal gap above his desk, traces its arc across the Martian sky while he works. The sky outside is the color it has always been, rust and butterscotch and haze, and the shadows on the surface below are the null-percept. He stopped noticing them as wrong sometime in the past three days. This concerns him slightly, the speed of the adaptation, but he does not have time to sit with it right now.
His stomach reminds him at some point in the afternoon that he has not eaten since the coffee he made at seven. He gets up, crosses to the nutrition dispenser, selects the option he has selected three times already this week because selecting it requires no decision-making and decision-making is a resource he is rationing carefully, and eats standing at the counter while reading the modeling results Sable sent for the freight platform saturation gradient.
The results are good. Better than good. Sable has identified an optimal saturation range for depth legibility that is narrower than Elian would have chosen manually and has included a secondary parameter set for operators who have already begun developing the extended null-percept spatial perception that the early medical studies flagged. Two parameter sets, one for standard perception and one for extended perception, both running from the same underlying variable, both compatible with the existing rendering architecture. It is elegant in the way Riku's work is consistently elegant, structured from the inside out, starting with what the system needs to be.
He sends a message to Riku that says only this works and Riku sends back a message that says he knows.
By evening he has eleven variations of the surgical overlay depth system documented and tested and the eleventh is close enough to right that he can see the shape of what right looks like from where he is standing. The geometric border approach needs to be combined with a luminosity differential, not the saturation approach he used for the freight platform but something narrower, a brightness offset applied specifically to the depth-indicator elements that lifts them fractionally above the visual plane of the underlying tissue, not by much, just enough for the brain to read them as occupying a different spatial layer without reading them as competing with the reality they are augmenting.
He runs the twelfth variation.
It is not wrong. He checks it again. Still not wrong. He runs it at three different simulated lighting conditions, the bright overhead lighting of a standard operating theater, the adjusted low-light conditions used for certain procedure types, and the mixed-light environment of a minimally invasive setup with a camera feed. Not wrong in any of them. Better than not wrong. In the mixed-light environment with the camera feed it is actively good, the depth indicators reading clearly and unambiguously, the null-percept present in the system but not as a depth communicator, as a background value against which the depth indicators are visible, the threshold quality of the null-percept serving for once as a feature rather than a liability, the far end of the depth field reading as the limit of the camera's spatial knowledge rather than as a wall.
He sits back in his chair.
Then he opens the documentation file for the surgical overlay component and begins writing the technical specification for the twelfth variation. He writes it the way he writes all technical documentation, clearly and specifically and without assuming that the person reading it already understands what he understood when he wrote it, because documentation written for people who already understand is not documentation, it is notation, and notation is only useful to the person who made it.
He writes for two hours. The documentation is longer than the implementation because the implementation is a set of values and the documentation is an explanation of why those values are right, and why requires more words than what.
When he is done he opens the message thread with Sena and sends the specification. He writes that this is what he has for the surgical systems overlay. He writes that it is not the same approach as the freight platform remediation and that the reasoning is in the specification if she wants to understand it before the client call. He writes that in his assessment it is safe to deploy and that he recommends a two-week monitored pilot with a small number of non-critical procedures before full deployment, not because he is uncertain about the technical solution but because any change to a surgical environment should have a monitored pilot regardless of how confident the developer is.
He sends it. He looks at the clock. It is past ten. He has been working for fifteen hours.
He gets up and makes coffee, his fifth of the day, and carries it to the window. The Martian sky outside is dark, or rather it is the null-percept, the same quality of threshold and openness that the deep space background renders in the THRESHOLD project, the same color that his coffee cup has been for a week and a half now and that he no longer looks at and thinks is wrong.
He thinks about what he noticed today, about the saturation gradient and the geometric border and the luminosity offset, about the specific properties of the null-percept that he has been working with intensively enough to begin to understand not as a replacement for black but as a thing with its own characteristics and its own logic and its own possibilities that black did not have.
He thinks that designing around the null-percept is not the same as designing with black. He has known this since the first remediation session but he has been keeping it at the edge of what he is willing to articulate because articulating it requires going further into the question of what the null-percept is and why it has the qualities it has and what those qualities mean for the people who are beginning to use it as a genuine perceptual tool rather than as a broken replacement for something they lost.
He writes one line in his investigation file before he goes to sleep.
The null-percept is not a wound. It is not a healing either. It is something that does not have a word yet because it has not existed long enough for a word to accumulate around it.
He closes the file. He finishes his coffee. He looks at the cup in his hand, the ceramic one that started all of this, the one that used to be black and is now the color with no name, and he turns it in the Martian light and thinks that it is strange how quickly the most fundamental disruptions become ordinary. How quickly the eye adjusts. How quickly the brain finds a new baseline and calls it normal and moves on.
He wonders if that is resilience or if it is something else. If there is a version of adjustment that is not adaptation but erasure, not learning to live with the change but simply forgetting that the change occurred.
He does not write this in his investigation file. It is not a technical observation. It is the kind of thought that belongs in the file but that he is not ready to commit to yet, because committing to it would require deciding what it means, and deciding what it means would require knowing whether the forgetting is a feature or a failure, and he does not know that yet.
He sets the cup on the windowsill.
He goes to sleep.
In the morning there will be nine more clients in the remediation queue. He will open the first file and find the seam and begin, again, the patient work of fixing it.
The Martian sky outside his window shifts through its colors in the dark, rust and amber and the slow change toward whatever the sky here is becoming, and the shadows on the surface below are the color with no name, threshold rather than terminus, the far end of things opening outward rather than closing.
Still there.
Still strange.
Just slightly less strange than yesterday.