The Wrong Equation

Somewhere in an e-ink company's product meeting — and I want you to picture this room, because the room explains everything — a product manager stands before a whiteboard. The whiteboard has two columns. On the left: ADVANTAGES OF E-INK. On the right: PRODUCT OPPORTUNITIES. Under advantages, someone has written one item: eye protection (护眼). Under opportunities, the list runs off the bottom of the board.

This is the moment the category error was born.

The product manager looked at e-ink technology and saw a medical device. Not a reading device. Not a contemplative device. A health device — something that protects your eyes. And once you define the technology as "anything you look at for a long time should be on e-ink," the logic becomes irresistible: People look at stock charts all day. Put stock charts on e-ink. People sit in meetings all day. Put meeting recorders on e-ink. People use office suites all day. Put WPS Office on e-ink. People talk to AI all day. Put AI on e-ink.

This is single-indicator thinking — seizing one property and extrapolating it to infinity. E-ink's real advantage is not that it protects your eyes. E-ink's real advantage is that it is slow. It refreshes slowly. It responds slowly. It does not seduce you with notification badges or dopamine-optimized scroll feeds. It is a technology that, by its physical limitations, encourages you to sit with one thing for a long time. To read. To think. To not switch tabs.

But "slow" is not a word that survives a product meeting. "Eye protection" is.

Everything it can do, except what you actually want

I wrote previously about the practical costs of buying a BOOX e-reader as a cross-border consumer — the locked ecosystem, the phantom return policies, the payment walls. That essay was about the price of freedom. This one is about a different question: why are these devices designed this way in the first place?

Open any e-ink manufacturer's product page and you will find a device that can do a remarkable number of things. The BOOX Note X6 runs a built-in AI assistant. It has placeholders for stock trading apps, news aggregators, office suites. It supports handwriting recognition, voice recording, and document scanning. It is, on paper, an omnipotent tablet — a Swiss Army knife rendered in electronic ink.

Now try to install ChatGPT. Or Claude. Or any application that requires Google Play Services. The device transforms. The omnipotent tablet becomes what I can only describe as a Ministry of Rites official: Ancestral rules cannot be violated! (祖宗之法不可变!) The system that moments ago was advertising unlimited possibility now draws a line in the sand with the solemnity of a Qing dynasty censor. Open Android, Google excluded. Your productivity, our definition.

The built-in AI assistants across these devices are, almost without exception, shells — domestic large language models wrapped in proprietary interfaces. iReader's AI runs on Baidu's Wenxin/Ernie wearing a different hat. Bigme's has a name I will get to later. BOOX's own offering, Xiao Wen AI (小文AI), generously provides thirty free book summaries per month. Thirty. I can picture the product manager's expression when that number was approved: Enough for you, right? The assumption is that reading is something you do in small, managed doses — thirty summaries, thirty transactions, thirty units of content consumption — and the AI is there not to help you think but to help you finish faster.

The meeting-hell product pipeline

To understand why these devices look the way they do, you have to understand the rooms they were designed in.

E-ink product managers live inside meeting hell. Their days are sliced into morning syncs, weekly reviews, quarterly retrospectives, cross-departmental alignment sessions. I should be careful with that observation, because the obvious regional version of it is not what the data shows: on Asana's 2024 figures the countries losing the most hours per week to unproductive meetings are France (9.1) and Germany (8.8), ahead of Japan (8.3), the US (5.2) and the UK (4.1), and while Slack and Qualtrics found Japanese respondents spending 37% of their time on performative work against a 32% global average, South Korean respondents reported 28% — tied with the US for the lowest in the survey. Meeting hell is not an Asian trait. What is specific here is narrower and harder to argue with: every device in this category ships with a meeting recorder. Hand meeting-dwelling people a new display technology and ask them to imagine use cases, and they imagine it inside a meeting. This device does not have a meeting recorder? Give the boss a meeting recorder!

And then someone adds AI to the meeting recorder. This is where it gets genuinely perverse.

The old cost of a meeting was: meeting time plus the time spent writing minutes afterward. AI meeting summarizers reduced that cost to: just meeting time. The minutes write themselves. Sounds like an efficiency gain, right? Except what actually happened is that the reduction in post-meeting labor made meetings feel cheaper to schedule. If the minutes are free, why not meet for another hour? The AI summary is clean, structured, action-item-rich. The boss reads it and thinks: That was a productive meeting. Let's do it again tomorrow. AI did not reduce the number of meetings. AI reduced the perceived cost of meetings, which increased their frequency.

The meeting-hell product pipeline is a closed loop: product managers who live in meetings design features for meetings, which generate AI-summarized proof that meetings are productive, which justifies more meetings, which generates demand for more meeting features. The device is not a reading tool that acquired meeting features. It is a meeting tool that happens to have an e-ink screen.

The nemawashi accelerator

What differs between places, then, is not the volume of meetings — the hours data does not support that story — but the ritual form the meetings take, and what happens when AI is applied to the ritual. In Japan, the relevant process is nemawashi (根回し) — the practice of building consensus through informal pre-meeting consultations before the formal meeting where the decision is officially "made." Nemawashi is old. It predates electricity. It is one of the most sophisticated consensus-building mechanisms in organizational history, and it is also one of the most time-consuming.

Now add AI.

In the Chinese version: the data does not look compelling enough — let us meet for three more hours so the AI can generate a prettier report. In the Japanese version: we used to do three rounds of nemawashi before each meeting, but now AI lets us do thirty. Both outcomes use the most advanced technology available to accelerate the most ancient bureaucratic rituals. The tools change. The rituals do not. The rituals simply run faster, which means they run more often, which means they consume more of the day, which the AI dutifully summarizes into a document that proves how efficiently the day was consumed.

If you ever wonder why an e-ink device ships with a meeting recorder and an AI summarizer, this is why. The people who designed it cannot imagine a workday that is not made of meetings. The device is a mirror of its makers' lives.

One correction, because an earlier version of this essay made a claim in passing that conceded the argument. It described e-ink as a technology that exists because someone wanted to read without eyestrain. That is the industry's story about itself, not the record. Electronic ink was built at the MIT Media Lab in January 1997 by J.D. Albert and Barrett Comiskey, working from an idea of Joseph Jacobson's: a book whose contents could be changed at the push of a button. The design targets were paper-likeness — a display that emits no light of its own, stays legible when tilted, and draws almost no power because, like a page, it holds its image without being fed. Comiskey later described the objective as giving life to the surfaces around us, which is broader than reading and says nothing whatever about eyes.

Eye protection was never in the technology. It was attached to it afterwards. That is a stronger version of this essay's argument than the one I originally wrote: the wrong equation is not a misreading of what e-ink was for, it is the invention of a purpose the thing never had.

The same test has to be turned on my own equation, or I am doing what I am objecting to. Slow reading was not Jacobson's aim either. The founders wanted a rewritable page, not a contemplative one. So "e-ink means slow reading" is not a recovery of original intent — it is a proposal about what this hardware is good for, and it has to be argued on the merits rather than smuggled in as the designers' wish. Both equations are impositions. The case for mine is only that it follows from what the physics actually does: a screen that takes half a second to turn a page is bad at everything that depends on immediacy, and good at the one thing that depends on its absence.

The electronic urn

I want to describe a scene to you, because it is the single funniest thing to emerge from this entire product category.

Someone — a real person, in a real conversation about e-ink use cases — suggested using color e-ink displays to watch the stock market.

If you have never used a color e-ink screen, here is what you need to know: the refresh rate is terrible. Not "slightly worse than LCD" terrible. Physically, structurally terrible. Most colour e-ink panels in shipping devices belong to E Ink's Kaleido family: a standard black-and-white microcapsule layer with a colour filter array laid over it. A smaller number of panels (ACeP / Gallery) drive several pigment types inside each capsule instead, and are slower still. Either way the particles have to be pushed into position electrostatically, and that takes time — enough time that fast-moving content leaves ghost images from the previous frame overlaid on the current one. On a static page of text, this is a minor visual artifact. On a stock chart updating in real time, it is catastrophe.

Picture it: 9:30 AM, the Shanghai exchange opens. Yesterday's crash lingers on screen — red candles, plunging trend lines — while today's rally tries to paint itself on top. Two K-line charts physically superimposed. The colors, already compromised by e-ink's limited gamut, undergo a further transformation. Red becomes dried-pig-blood red (猪血红). Green becomes poisoned-vegetable green. Both are overlaid with an unwashable gray-black ghosting that no amount of manual refresh will fully clear. The result does not look like a stock chart. It looks like an electronic funeral urn (电子骨灰盒) — a memorial display for your portfolio, rendered in colors that suggest the money has not just been lost but has been decomposing for several days.

The thing is, e-ink can play video. A slow lecture, a talking head, a PowerPoint walkthrough — these work tolerably on a good color e-ink panel. The question was never whether it could be done. The question is what you give up to make it happen: every engineering hour spent chasing faster refresh is an hour not spent on sharper text rendering, cleaner PDF reflow, longer battery life, fewer ghost images on the page you were actually trying to read. The eighty-frames-per-second mode that Bigme advertises does not come from nowhere. It comes from the same budget — silicon, power, development time — that could have gone toward making the device better at the one thing e-ink was born to do. But "we made reading 15% better" does not make a product launch video. "We can play stock charts" does.

The iFlytek pantheon

iFlytek (科大讯飞) deserves a section of its own, because iFlytek has achieved something no other company in the e-ink space has managed: they have built an entire consumer electronics empire on a single premise — our speech recognition is the best, therefore everything involving speaking or hearing needs us — and they have executed on this premise with the relentless consistency of a state-backed religious movement.

Visit the iFlytek store. AI mouse. AI keyboard. AI earphones. AI learning machine. AI translation pen. AI reading device. AI recorder. AI office suite. Every product is a different shell around the same core conviction: that iFlytek's voice technology is so transformative that the correct response is to put it inside everything. The AI learning machine costs eight thousand yuan. I do not know what it does that justifies the price, but the marketing implies that it will get your child into a good school, which is the most expensive promise a product can make in China and therefore the one most frequently made.

The reading device is part of this pantheon. It does not exist because someone at iFlytek thought deeply about what readers need. It exists because iFlytek makes AI products, e-ink is a screen, and screens are surfaces on which AI can be displayed. The reading function is not the purpose. It is the excuse.

The seven-inch office

If iFlytek represents the wrong equation applied with religious zeal, Bigme (大我) represents the wrong equation applied with a kind of cheerful delirium.

Bigme's B7 Pro is a seven-inch color e-ink device. Seven inches — roughly the size of a large smartphone, smaller than most paperback novels. Into this surface, Bigme has packed: an 80-frame-per-second fast-refresh mode, 4G cellular connectivity with full voice calling capability, a high-definition camera, and a product identity that the marketing materials describe, without apparent irony, as a "three-in-one system" — a smart office tablet for adults in the morning, a smart learning device for children in the evening, and a smart family tablet for everyone on weekends.

I am not exaggerating the marketing copy; this is how the product is advertised. A seven-inch e-ink screen. For office work. With a phone camera. That your child uses for homework after you finish your meeting notes.

The bookstore situation deserves special mention. BOOX, which I have already written about, welds an uninstallable bookstore into the system with two content providers — JD Reading and Dedao. Bigme looked at this and thought: only two? The B7 Pro ships with eight: JD, Dedao, Qidian (起点, web novels), Tomato (番茄, free web novels), Douban (豆瓣, reviews), Amazon Kindle, WeChat Reading, and Jinjiang (晋江, romance fiction). The device arrives pre-loaded with the assumption that you are simultaneously a business audiobook listener, a web novel addict, a literary critic, a Kindle migrant, a WeChat ecosystem participant, and a romance reader. It does not know who you are, but it is absolutely certain you are all of these people at once. And if you are none of them — if you use none of these eight storefronts, if your reading life exists entirely outside the ecosystem of web novels and business audiobooks and romance fiction — the device does not have a response for you. You cannot remove them. The product manager did not design for the possibility that a user might want none of the above, because that user does not generate referral revenue. The bookstores are not features. They are revenue channels wearing the costume of features, and they are uninstallable for the same reason a landlord does not let you remove the vending machine from the lobby: it is not there for you. (The same logic applies to the pre-installed cloud storage apps — Baidu Netdisk, Nutstore, sometimes three or four at once — as though the user could not possibly figure out how to transfer a file without the manufacturer holding their hand. The device treats its owner like an infant who might hurt themselves if left alone with a USB cable.)

This is what happens when the wrong equation runs unchecked. Once you have decided that e-ink means "eye protection" and eye protection means "everything," there is no natural stopping point. Why not a camera? People look at cameras. Why not cellular calling? People talk on phones. Why not a children's learning mode? Children have eyes too. The logic is impeccable if you never question the premise.

The B7 Pro also ships with a built-in AI assistant called, magnificently, BigmeGPT. I have nothing to add to this. The name is the commentary.

But the B7 Pro's true achievement is its dual-system mode: by day, a "smart office tablet" for adults; by night, a "smart learning device" for children. The learning system offers synchronized classroom exercises with AI auto-grading, synchronized reading-aloud tutoring, and — on a seven-inch e-ink screen with a refresh rate that makes animation look like a slideshow — synchronized teaching videos. It also offers doodling, handwriting practice, and, in what may be the single most conceptually confused feature in the history of consumer electronics, focus training exercises and brain games. The device that exists because you cannot stop looking at screens now wants to train your child's attention span. On a screen. Notably absent from the B7 Pro's marketing is any mention of quiet, focused, uninterrupted reading — the one thing e-ink was invented for. The product page is a wall of features competing for your attention, which is fitting, because that is also what the device itself has become: an Android tablet with notifications, pop-ups, and app stores, wearing the skin of a reading device. The e-ink is no longer the point. It is the disguise.

The teaching content, the marketing assures us, has been "triple-reviewed and triple-proofed by elite teachers from prestigious schools." The learning system costs 499 yuan after a three-day free trial. It is classified as a virtual product and does not support refunds after payment. It is locked to the specific device. If the device breaks, the 499 yuan does not transfer. I am not exaggerating. This is the product.

The device also advertises "intelligent recognition of text, graphics, and tables with automatic formatting." In a parallel universe where e-ink OCR and PDF reflow worked reliably, this might be worth mentioning. In this universe — where even the best e-ink devices still struggle with academic PDFs that have two-column layouts and footnotes — it reads less like a feature and more like a prayer.

The collectors and the manufacturers

On Xiaohongshu (小红书), China's visual social platform, there is a thriving community of e-reader collectors. These are people who own three, four, sometimes thirty or more e-ink devices. Their posts are beautiful: pastel-colored readers arranged on wooden desks next to dried flowers and iced lattes, leather cases in muted earth tones, screens displaying poetry or hand-drawn illustrations. One commenter wrote, with evident pride: "These are the ones I kept after going through more than thirty devices."

These users do not want office tools. They do not want meeting recorders. They do not want AI summaries. They want aesthetic pleasure. Emotional comfort. A small, pretty object to hold in a cafe, an escape from the screens that dominate their working lives. Their e-readers are chosen for color, for size (almost always under ten inches — pocket-sized, personal, not work-shaped), for the quality of the case, for the feeling of the device in the hand.

The manufacturers, meanwhile, are building 10.3-inch productivity tablets with stylus support and split-screen multitasking. The two groups are having entirely different conversations. The collectors want beauty. The manufacturers want efficiency. Neither, if I am being honest, seems particularly interested in reading itself — in the sustained, attentive, sometimes difficult encounter with a text that does not yield its meaning easily. The collectors want the aesthetic of reading. The manufacturers want the productivity of reading. The actual activity — sitting with a book for hours, arguing with it, losing track of time inside it — appears in neither group's vision of what the device is for.

The result is two groups fighting magic with magic. One side decorates the device. The other side loads it with features. Both have forgotten — or never knew — what the device was originally supposed to do.

"Cognitive misalignment"

And then there is DASUNG (大上科技).

DASUNG makes e-ink monitors — external displays that connect to your computer, rendering your entire desktop in electronic paper. It is, within the e-ink space, a genuinely interesting product for a genuinely specific use case: people who stare at screens for twelve hours a day and want to do so without the eyestrain of backlit LCD. Reasonable premise. Clear audience.

Then DASUNG posted on Xiaohongshu. What I recorded at the time was that the post named the market's problem as 认知错位 — cognitive misalignment — and that its argument was that users are still judging these products by whether they can read books, rather than by whether they can improve efficiency. I have to be exact about the status of that: I did not archive the post, I have not been able to re-locate it, and I did not record the account that published it. What follows is my paraphrase of something I read once and cannot now show you. It is not a verbatim quotation, the wording should not be quoted onward as one, and until the post is recovered this section should not be cited by other pieces as documentation of anything.

Let me restate this so the structure is visible. A manufacturer in the e-ink industry publicly announced that users who judge value by whether a device can read books have a misaligned value standard. The users are not wrong about what they want. They are wrong about what they should want. They should want productivity. They should want efficiency. They should want the device to make their meetings better, their reports faster, their workflows smoother. The fact that they want to read books — on a device the industry itself spent two decades selling to them as a reading device, whatever the underlying display was originally developed for — represents, in the manufacturer's view, a failure of understanding.

This is the category error arriving at its logical terminus. The wrong equation — e-ink equals eye protection, not reading — has traveled so far from its origin that the manufacturer can now look at a user who wants to read and treat that user as the party who has misunderstood the product. Those are my words for the position, not theirs.

What the equation should have been

E-ink is slow. This is the fact. Not a limitation to be engineered around. Not a problem to be solved with faster refresh rates and AI integration. A fact — a physical property of the technology, as fundamental as the weight of paper or the viscosity of ink.

And slowness, for reading, is not a bug. It is the entire point. The best reading I do — the reading that changes how I think, that restructures my assumptions, that leaves me different afterward — happens slowly. It happens when I am not checking notifications. When I am not switching between tabs. When the device in my hands is not trying to be anything other than a surface for text. E-ink, by accident of physics, enforces this kind of attention. Not because it has a "focus mode" toggle. Because the things that would distract you all work badly on it — which is not the same as their being impossible, as the rest of this essay has just spent several thousand words showing. The ghosting, the slow refresh, the limited color — these are not failures. They are the technology doing what it does, which is slowing you down, which is what reading requires.

And if the problem were truly eye protection — if the goal were genuinely to help the millions of people who suffer from dry eyes, screen fatigue, and chronic eyestrain — the humane response would not be to put every possible activity onto an e-ink screen. It would be better lighting. Better lenses. Better workplace norms that limit compulsory screen time. Better medical care for the people who already have symptoms. The e-ink industry chose the strangest possible route: instead of improving the conditions under which people use screens, it built more screens and called them protection. The wrong equation did not just misunderstand the technology. It turned a symptom — sore eyes — into a market expansion strategy.

The product managers looked at this and saw a medical device — something that protects your eyes. They should have seen a temporal device — something that protects your time. Not because that was the founders' intention; as noted earlier, it was not. Because it is what the hardware is actually fit for: e-ink means slow reading, which means deep attention, which means the kind of thinking that cannot happen on a device refreshing sixty times a second. That is an argument about fitness rather than origins, and it is the weaker kind of claim to make — but it is still a better bet than a medical promise the technology was never in a position to keep.

But "slow" does not survive a product meeting. "Slow" does not go on a spec sheet. "Slow" does not justify an eight-thousand-yuan price tag. So they wrote "eye protection" on the whiteboard, and from that single wrong equation, everything followed: the stock tickers, the meeting recorders, the AI summarizers, the thirty free book summaries per month, the built-in bookstores welded to the system partition. A technology that was supposed to give you time became a technology that fills your time. A device that was supposed to let you read in peace became a device that does everything except let you read in peace.

And when users pointed this out — when they said, simply, I just want to read — the manufacturer's answer was that reading is the wrong yardstick.

The gap is real. But it is not where they think it is. The wrong equation did not merely produce the wrong products. It produced the wrong idea of who the user is — and then blamed the real user for not matching.

July 19, 2026.