dreaming.daxmurray.com
Ellipsus Emboss, Disability, Surveillance, and Proving Humanity.
Sometimes, men with government clipboards sit outside houses and count how many times someone leaves their home, counts how many steps outside they take, and note everything down. They will try to stare through open blinds and determine how often the inhabitant of the home stands up from the couch. All while they have that person's social media feed pulled up on a laptop in the passenger seat. Later, they might ask the neighbors how often they see the person they are surveilling. After, they will go back to the office, put the numbers in a computer, and the computer will determine if the person in question is really disabled.
People with disabilities, no matter where we live, are considered liars until proven otherwise. We can explain until we are red in the face that the photos we uploaded yesterday were taken in 2006, long before our injury. Look, you can even see the date in the corner! I swear, I found an old roll of film and had it developed and scanned these to show online. Here, my receipt from the film developer! Ask my sister, she will tell you, this photo is from 2006, way before COVID wrecked everything! See? Who wears this kind of clothing in 2026 anyway! And I visibly look 20 years younger. What do you mean this is proof I am faking my disability?
Some countries have taken to spending a ton of money on public "awareness" campaigns meant to encourage people to report their neighbors for "fraud" and "cheating the system" by asking them to be extremely watchful of their disabled neighbors and on the look out for signs of faking.
There is an entire subreddit dedicating to calling out "fakers" of disabilities; usually these posts are full of misinformation that only further stigmatize people with disabilities. The people on this subreddit believe they are "helping" the "real" people with disabilities by calling out the "fakers" but are often so misinformed that they couldn't identify malingering even if it tap danced on their nose. Their belief is that unless someone is in the hospital 24/7 and entirely unable to function without a team of medical staff monitoring them around the clock, they are faking it. They also believe that the stigma of disability is so strong that no one in their right mind would ever publicly admit to being disabled. If you go online and say, "I have a disability," they will take this as proof that you aren't _really_ disabled, because no one with a _real_ disability would ever be able to pull themselves out of the societal stigma and _readily_ admit to it. To say you are disabled is to prove that you aren't, according to these self-proclaimed reddit experts.
People with disabilities are used to having our medical records scoured by every new doctor determined to find the small discrepancy that will unravel our intricate webs of lies. A twenty years ago, a doctor incorrectly put your blood pressure down as 40/80 because he was in a rush one day? You can explain until you are blue in the face that CLEARLY those numbers were transposed because HOW would a 40/80 BP even work? But the new doctor just sighs, a small grin stretching across his face as his chest fills with pride at catching yet another awful faker, and erases every single diagnosis from your chart and sends you home saying you are perfectly healthy.
We are used to bosses bringing us in for a one-on-one because they noticed we take at least 3 bathroom breaks per work day, and they all seem to last at least ten minutes. That's a full half an hour! They are determined to find out why you are slacking and put an end to it. You can bring in an embarrassing note from your GI doc explaining that you have IBS. But that note will be filed away in the appropriate place: the shredder.
The key-logging software on your work laptop informs your boss every time your mouse or keyboard isn't active for at least ten minutes. Again, you can explain that you got up to get water and a snack to keep your glucose levels at a functional baseline; but it doesn't matter. You're on company time; can you please leave your dysfunctional body at home with the rest of your private life?
We are used to every aspect of our life being quantified and having those numbers be weighed against some sort of mythical "normal" number.
Which is why, when I first read about Ellipsus Emboss feature, I immediately began thinking of all the ways my writing metrics--due to disability--do not line up with what is "normal" and how that failure to meet the Mythical Normal will cast doubts on my credibility.
## What is Ellipsus Emboss?
Ellipsus is a SaaS (Software as a Service) writing platform. It uses headers to make outlines, has a revision system, and allows writers to organize things into projects and collaborate and share with other authors. Many use it now for getting beta reader feedback, too. It has fun and colorful themes and has a "snippet" feature that transforms snippets of texts into graphics that can be shared on social media.
Many compare it (favorably) to Google Docs in terms of easily accessing writing on the go without tons of "distracting" features. Further, until yesterday, it was entirely free to use.
A screenshot from Ellipsus using Mary Shelley's Frankenstein as an example. This shows various draft versions, word count, and a side bar for comments, spell check, and outlines.
I myself found it interesting and signed up for an account in September or October of 2025. I tried using it for a few weeks and while I enjoyed it, it didn't have all the features I needed. At the time, I was a serious Scrivener user and while I was in search of something that would allow me to use my phone (Scrivener has yet to release an Android version), Ellipsus did not have things like scene metadata, tagging, or worldbuilding features. It also struggled with some of my books that exceeded 100,000 words, causing browser crashes as it loads the entire document at once, unlike Scrivener which only loads the scene you are working on.
But I still saw the appeal. A lot of my friends who had been struggling with writing suddenly found that Ellipsus helped them finally finish that draft or start the project they'd been meaning to.
When someone told me that they thought Scrivener was "too much," I happily pointed to Ellipsus as it claimed to be more privacy focused and promised no generative AI features.
I started using it more as a back up feature than something I actively wrote in, and still saw it as a potential way for me to share my next manuscript when it was ready for beta readers.
Overall, it wasn't for me. But every writer is different. I don't expect everyone to love Scrivener like I do. Every writer is unique. Every writer has their own quirks and habits and flow. My flow involves a tagging system, scene-level meta data, cork boards, labels, color coding, and more. Some people just want a pretty rainbow background and no other so-called distractions. One person's distraction is another person's vital component of drafting.
### Emboss
Yesterday, Ellipsus announced their new paid version, Ellipsus Plus. While Ellipsus already had a number of fun and colorful themes, they weren't very customizable. But with an Ellipsus Plus subscription, you can make your own themes. Your snippet exports can also be further customized with a background image.
And the best part of the announcement was that the monthly subscriptions were more of a "rent to own" model--after a certain number of months / years, your subscription would automatically transition to being a lifetime license. And the monthly price wasn't bad, either, at $6/mo when some writing software goes for $25 - $40 / mo. They also offered a flat $100 lifetime license up front. More expensive than Scrivener at $50/lifetime, but /shrug.
But the most talked about feature is the most controversial.
Emboss.
Emboss touts itself as a sort of insurance against false accusations of AI usage. It's marketed to authors as "here, you can use this if anyone questions your writing!" and advertised to readers as a sort of "If your favorite author can't share their Emboss with you, you can definitely assume they used AI. ONLY accept books written in Ellipsus with the Emboss feature enabled! We PROMISE you our software is more reliable than any other at catching CHEATERS!"
Yeah, it comes across a bit like extortion!
It claims to track your keystrokes and use some (non-AI) algorithms to determine if you write like a "natural" and "normal" human.
Who decides what is natural? Who determined what the benchmark is for normal? None of that is explained in their documentation. But they promise you that their Emboss metrics can definitely tell you if someone used AI.
Emboss lets you create a share link, similar to how you would with Google Docs or MS Office. But the Emboss links allow people to search through your version history, including time and date stamps. While this might be useful for the writer themself to try to figure out what they cut and when so that they can ensure there are no plot holes, etc, during the editing process, it's rather intrusive of a thing for readers to be able to demand.
My version history is often messy. I sometimes do not have names for characters figured out yet, so the minor character who appears in three scenes total might just be referred to as [PLACEHOLDER] in draft one. While I don't mind my author friends seeing this if I share it with them for feedback, it isn't something I want readers to see.
I work hard to revise my stories so that only the best version of it is shown. I was taught in school to show all my work in math class, but to hide it all in lit-lang class.
Further, there are often notes I leave myself, like "note to self: ugh, this sucks, I definitely need to work on the prose here but that's a future!dax problem." Again, I just don't want readers to see that. Those notes are full of anxiety, self-doubt, and often depression, as well as commentaries on my mental state or health state, such as "ugh, I want to keep going but the migraine prodrome is starting; reminder that I want this character to do that when I come back... whenever that may be lolsob."
If I knew that readers would be pouring over those drafts, I would be even more anxious about leaving those notes in. Some may say those notes aren't necessary; but to me, they are. It's helpful for me when I come back to a draft to see where my head was when I last left off. It can help me figure out "what the heck was I trying to do here?? These words don't make sense?" But then I read the note and see I was at the start of a migraine and ah. It all makes sense.
But it begs the question: how does this revision snapshot help prove authorship? Someone could very well have ChatGPT open in another tab. Maybe they don't copy and paste. Maybe they have the generated text open on Screen A and are typing it into Ellipsus to simulate the authentic author experience.
And what are they supposed to reveal, except that authors often have shit first drafts? Ah, in draft 1 it sounds way too much like AI, and now you can see how they revised it to sound more human? Is that the purpose?
The help page for this feature says one of the intentions is for readers to get an in-depth record of the development of the work. Readers are one of the use cases. This feature pre-supposes that readers are _entitled_ to behind-the-scenes looks at author works.
In a way, I do make some of my work available to my reader before release. But this is often to readers in my newsletter; ie, the people who already trust me. It's even more often locked behind a paywall, too; ie, the people who not just trust me but root for me.
This behind the scenes look at unfinished work has heretofore been the domain of Patreon extras, Ream perks, Kickstarter bonuses, and special edition add-ons. It's been for superfans. It's been for the readers who have already established trust.
Right now, I am releasing my WIP of Shackles and Shards: Shattered as part of a fundraising effort for the Pride: Brighter than the Stars bundle. It's a thank you for those who support me and other queer creatives.
But this feature says that this is not an extra; it's a _right_ that all readers are entitled to. Even the ones who hate you; the ones who want to tear you down; the ones who want a fraction of potential proof to justify a harassment campaign. It gives the people intent on doing harm a look into the private lives of the authors they hate. It gives them societal permission to pour over the author's notes to self and make value judgements, to twist it into a narrative that gives them the verisimilitude of permission to ruin the author's life and career.
> This is disgustingly invasive and wholly unnecessary when all it would take is regulating GenAi and forcing those who use it to label it as such. We are coming up w/ all of these ridiculous things to combat GenAi when simple regulations stomp this all out.
~ Kelley @thebookwyrmshoard2.bsky.social (BlueSky)
We are _already_ seeing this in the art world. Multiple youtube channels have popped up of people claiming they can tell when an artist used AI. And maybe they can, but they then use this as a means to demand the actual psd files so they can go over the layers one by one, so they can critique the entire process an artist goes through, each step of the way.
But that is not enough. These people then demand that the artist record their screens _and themselves_ during the entire process. They want artists to be livestreaming--one camera focused on their screen, one on their body. Artists must expose everything about themselves or be deemed AI.
And some artists are keeping their privacy and saying, "Fuck you, I owe you nothing." But beginner artists? Who are just getting started? Who don't have a decade of proof on DeviantArt already? They are feeling coerced, even if they don't want to, into sharing their lives for fear that not doing so will ruin their ability to pay rent.
This Emboss feature? It's saying that not only is it right and fair that people demand this of artists, but it should be done to writers, too.
It should be done to _nobody._
This is cop mentality. This is putting people in a panopticon; this is assuming people are liars until they can prove they aren't.
A second feature of Emboss is the metrics and statistics it collects. If your metrics meet certain parameters, you can get a Human Authorship Badge. This feature tracks your keystrokes and uses some sort of algorithm to determine if your WPM (words per minute) are "Normal" and your typing flow is "Natural." It measures how often and how long you pause, it looks at how much a sentence changes between revisions, and calculates if you spent either too long or not long enough writing something.
Nowhere in Ellipsus' help documents can I find a source for how they determine what is a "natural" number of "thinking pauses" and what is "unnatural."
Nowhere can I find how they determined that 30 - 70 is a human WPM (especially when touch-typists, like myself, can easily and accurately do more than 120 words per minute, and when some people with mobility and join issues can struggle to do five words per minute unless they have adaptative technologies).
None of their metrics have sources for how they determined what is human and what is not.
But it is especially obvious that they never considered that people with disabilities write.
> I pause a lot because I have to go look up words. Sometimes, I also write by hand and then type it up. Would that look unnatural?
> What about when you open a new doc to work something out, and then copy and paste it in.
> This whole thing seems poorly thought out.
~ Kara Jorgensen @authorkaraj.bsky.social (BlueSky)
Various comments I saw in response to this asked some insightful questions:
* What about authors who have children? Children who interrupt them regularly? Will Emboss deem them artificial because their kids are struggling with their homework?
* What about authors who have elderly parents they are caring for? Parents who often need help with things like using the bathroom? Things that can't wait for an author to keep typing until they hit some "human" metric?
* What about authors with cats? Do I even need to say more?
* What about authors with gastro issues? Who are already hyper-monitored at work for too many bathroom breaks? Will they be subject to the same thing here? Will they have to just try to hold it so Ellipsus doesn't think they are AI due to their VERY HUMAN digestive system?
* What about authors who use dictation?
* What about authors who use virtual keyboards?
* What about authors who go to look up a word and then get sidetracked for three hours on an etymological journey to the Akkadian empire?
* What about authors who are white-knuckling it through a migraine until they just can't anymore and then don't write for another week?
* What about authors who can only write sporadically in between the thousand other obligations they have?
What about the authors who are **_human?_**
These metrics leave out people with disabilities; people with ADHD, people with mental illnesses, people with joint pain, people with IBS, etc. And since I fall in that category, that is what I am mostly concerned about here. But it's more than just that.
Humans are messy; unpredictable; **_unmeasurable._ **We all live chaotic lives. We are not machines.
But Ellipsus believes its (non-AI) algorithm--a machine--can determine who is and isn't human.
And it's advertising this ability as a means of building reader trust.
But at the cost of the private lives of the authors.
Someone intent to do harm can see those long breaks and potentially extrapolate that the author they are intent on smearing has IBS. They can see the weeks where the author writes nothing at all and extrapolate that the author might be dealing with depression. They might see a slow words per minute metric and determine that maybe the author they want to defame has some sort of chronic pain issue or joint issue.
> I'm a disabled author, and upon hearing about this whole Emboss in Ellipsus bullshit, I IMMEDIATELY saw the red flags. why are we literally reinventing the FBI but for writers? why are we proposing pressuring authors into using SURVEILLANCE TOOLS???
~ Riley @whamphetamine.bsky.social (BlueSky)
### Emboss and Poverty
One of the points of pride for Ellipsus is the fact that it was, until yesterday, entirely free. But even still today, the free plan still lets you write unlimited words, engage in collaboration, have as many drafts as you want, use themes, etc. None of the "vital" features are locked behind a subscription.
Except. Emboss is locked behind it.
Emboss, which they are trying to sell you as _necessary_ if you want to avoid being the next victim of a false AI witch hunt.
And their marketing towards readers only seems to be aimed at _enflaming_ the next witch hunt.
_Make sure your favorite author shares their Emboss link! Otherwise they might be an AI grifter!_
Already, I've seen readers pestering authors on social media asking if they have their Emboss links. Already, I've seen readers saying they won't read authors who don't have the Emboss link on their carrds or linktrees.
The atmosphere has quickly become: if you _don't_ use Ellipsus Emboss, you are an acceptable target for a witch hunt.
People who want to ruin your career because of a suspicion they have are _entitled_ to pour every pause in your writing routine, scrutinize every edit you made during your months of revision, inspect your writing pace and find it lacking in some way so that they can justify the harm they seek to cause. Your privacy is an acceptable thing for them to demand; it's now the price of admission to the creative fields.
Which serves Ellipsus just fine. It means they might get a ton of authors hoping that Emboss will appease the mob. That's money. On their very first day of offering a paid service.
The Author's Guild already tried something like this. They offered a "Human Made" verification badge, too. $100 for members and $150 for non-members. Per book.
So yeah, the Authentic Human badge you get from Ellipsus just costs a one-time $100.
But it's still extortion.
And for some people, like fanfic writers and hobbyists who seek no monetary gain, why spend money on a hobby they do for the joy of it? Especially in a world where gasoline is $6/gallon and the price of food has gone up 300% since 2020?
Not everyone can afford Ellipsus Plus. Which they say is fine because you get _almost_ everything on the free plan. Except the service they are selling as a Stamp of Approval Against AI Accusations.
### "This is a Good Start"
Another unfortunate thing I've already seen is people saying Emboss is a "good start" but it doesn't go far enough.
To these people, we need to be asking what people are doing during those thinking breaks that are too long. The author should document for every pause over 5 minutes what they were doing, and then the algorithm should determine if the author took too many bathroom breaks.
It should also capture what other programs and browser tabs the author has open while Ellipsus is open. Like sure, we can assume that if there is a 10 minute break, they probably have either Claude or ChatGPT open in another tab. But by Ellipsus tracking other tabs, we will know EXACTLY which AI the author is cheating with!
Oh! And we should probably have Ellipsus also get root access to the PC so it can scan for macros and other auto-typing programs that might be used to "fudge" human typing. Yeah, some of those are accessibility devices or software used in conjunction with adaptive technologies but c'mon, real authors aren't disabled, we already established that.
OH OH OH! And can we get a feature in Ellipsus Emboss that turns on their webcams, too? That way we can zoom in on the lens of their glasses and see if maybe they have a tablet open too and are running an LLM on their tablet!
It's NEVER going to be enough for the people determined to find evidence ("evidence") of AI usage. The people who start witch hunts are usually intending to do harm. They don't need _actual_ evidence. They need a verisimilitude of evidence; they just need a faint pretext--enough that a random passer-by will go, "Oh, okay, yeah, sounds reasonable," before moving on.
The people excited about this are _bad faith actors._ And bad faith actors will take any tool and turn it into an instrument of abuse.
And Ellipsus advertising Emboss as a tool for catching cheaters is just emboldening the bad faith actors and giving them both cover and reassurance that their bad faith actions are not only justified, but worth pursuing.
### But It's Opt-In!
Emboss is opt-in! If you don't want to share your revision history and writing metrics, you don't have to!
Coercive consent is not consent.
Ellipsus is pitching Emboss as something to protect your livlihood. Ellipsus understands that false AI accusations can absolutely _ruin_ an author's career. And for people who pay their bills with their writing income, something like Emboss can seem like a sort of insurance policy.
So is it _really_ opt-in when you have Ellipsus telling readers to demand the Emboss receipts before buying a book? And is it really opt-in when those readers then start saying "I won't buy your book unless you have the Emboss!"?
That's coercion. That's not true opt-in.
_You must give your worst haters access to every detail of your private life, or they will take your livelihood and harass you out of your community._
That is not truly opt-in.
And worse than that, it's merely the _sharing_ of the data that is opt-in. Ellipsus will still be logging every keystroke, even on the free plan.
### Don't Like Ellipsus, Don't Use It!
Ok, yeah, cool.
But Ellipsus is marketing Emboss as something readers should be demanding of authors.
It's currently the only tool of its kind on the market. Maybe software like Novlr or LoreForge will pick see this feature and decide to add it into their own feature set, maybe to the chagrin of their users who have been waiting for the next version and the promised To Do list feature.
Maybe Literature and Latte will decide they don't want to do that at all, and besides, the next version of Scrivener isn't even on the roadmap yet.
So if all the readers are demanding Emboss, do authors with different work flows need to switch? Will new authors, ones who only started publishing and can't point to a backlog that pre-dates November 2022, feel like they have no choice if they want to get a readership?
I've seen plenty of people who say they simply don't read books published after 2022. And I've seen a few of those people now say, "Unless I can see the Emboss."
I personally do not like Ellipsus, the writing software. Until yesterday, I really liked Ellipsus, the team and company. I would tell writers who said "I don't like all the bells and whistles in Scrivener" that they should give Ellipsus a try!
But the more readers hear Ellipsus say, "Ask for the Emboss and guarantee you aren't reading slop," the more authors are going to feel pressured to move their process into Ellipsus.
Some of those processes might involve dictation, cleaning up the dictation, and then copy and pasting into Ellipsus. Something Ellipsus will mark in the Emboss as "unnatural" and "not normal" and specifically call out, as it tracks how many words were copy/pasted and how many typed into Ellipsus directly. And the more copy/pasted words, the more likely you are to not be deemed human.
Even people without disabilities might draft in something like Scrivener and then edit in Ellipsus. Or maybe they draft by hand and then use an OCR (super ancient tech that vastly predates AI, btw) to get a digital text file and then paste that into Ellipsus.
Some people purely use Ellipsus as a "back up" or for getting beta readers feedback and draft, edit, and revise in something entirely different!
But it's really kinda convenient that the marketing for Emboss kinda makes it seem like anyone who doesn't draft, revise, and edit in Ellipsus in a very mechanical way isn't really an author and shouldn't be trusted and probably is a devious monster using AI. Just when they come out with a paid plan.
I'm not casting aspersions, but it is very interesting to me how their marketing is playing out!
I've always said that the best writing software / platform / app / etc is the one that you actually want to use. It's not Scrivener, it's not Word, it's not Campfire, LoreForge, WorldAnvil, StoryFyre, Novlr, etc. It's whatever you like. And maybe, the best writing sofware _for you_ is Ellipsus. I'm not saying don't use it. But right now Ellipsus' marketing _is_ positing itself as the only _ethical_ writing software and I think that's gross. They aren't saying "Don't like us, don't use us." They are saying, "Don't like us? Enjoy your justly earned AI accusations."
The more I write this the more it sounds kinda like extortion. Which is what people accused The Author's Guild of, too, when they tried to sell Human Authored badges.
## The Marked and the Unmarked
Once upon a time, English had wereman and wifman. Neither gender was "marked." But then wereman became just man, and wifman became woman.
The marked and the unmarked is a concept in sociology by which humans describe ingroups and outgroups; by which we say which thing is the default, and which is the deviation.
It's like when you ask your aunt if her friend is cis or trans and she says, "She's just normal??" Similar to how people once said, "I'm not gay, I'm just normal." Or instead of saying, "I'm allosexual," people used to say, "I'm not asexual."
Gay, trans, and asexual were the "marked" states. And, in some cases, still are. But people are recognizing that having default and deviation categories for _human beings_ is really not great, and are using terms like cis, straight, and allosexual not just for accuracy but because there really isn't a "default" human.
The marked category is announced as different from the start; whether that is a good thing or a bad thing depends on how the marked and unmarked states were determined.
In a world where there is AI writing and there is human writing, the normal has been and _should be_ **human writing.__** It should be the unmarked state.
Consider imitation crab. Regular crab is just crab. But laws and regulations say that you cannot sell _imitation_ crab unless it is **marked** as such.
AI is imitation crab. It should be marked. It is the deviation.
The Ellipsus Emboss feature accepts and is preparing for a future that techbros want: one where genAI is the default, it is the standard, it is what should be expected when opening a book. The Emboss feature and its stamp--as well as the Author Guilds badge--presuppose a future where humans are the _marked_ category. We are what needs to be called out.
I refuse to except this as the inevitable future. We humans should not be marking ourselves; we should not be accepting the premise that AI is the new standard just because billionaires might lose money otherwise.
> At the end of the day, we're still using a machine to determine who is human and we should not be the ones having to do that.
>
> we were the ones here first.
~ Laika L. Gagarin @laikalgagarin.bsky.social (BlueSky)
## Proving Humanity
And this brings me back to disability. Disability is a marked state. Some of us use "Disabled." Some of us use "People with Disabilities."
But we are still marked. We are still asked over and over to not just prove our humanity, but prove that our humanity is worth preserving.
At least in the US, there were senators calling for hospitals to let the elderly and disabled just die from COVID-19 so that people who "actually contribute to society" can get the resources to survive and live to help someone's profit another day.
Since we don't "contribute," we are forced to prove that we are humans worth dignity, respect, and consideration. So many writers live with disabilities and are used to being surveilled; used to having people pour over metrics calculated by machines. Metrics that will be then used to determine if we _deserve_ accommodations, deserve medical care, deserve to survive another day in this capitalist hellscape.
Even people without disabilities got a taste of this during the initial work-from-home phase at the start of the pandemic. Workers were asked to install eye-tracking software; mouse tracking software, leave their webcams on at all times, etc.
And as the return-to-office movement started, many were hoping to use the metrics obtained during mass worker surveillance to justify return-to-office. They were hoping to use those metrics as a cudgel. They wanted the metrics to say workers were less efficient at home and therefore must once again endure long commutes, high commuting costs, and less time to themselves or with their families.
People who use metrics are usually using them as a veneer for causing harm. These metrics are used to ostracize people with disabilities; drive them out of jobs, force them out of communities, and harass them out of the public until they die alone.
So, when I read about Emboss, that was the first thing I thought off. It is surveillance tech just the same as the many that are used against people with disabilities. And like those techs, it is being used to determine who is and isn't a human being.
> My disability is inextricable from my writing process and I really don’t need one more way people can judge me for that, thank you.
~ Anna Holmes @annabookwriter.bsky.social (BlueSky)
AI is eroding so many things we took as part of the social contract. Included in that is trust. And Ellipsus posits that the solution to that erosion of trust is cop mentality surveillance tech of every single creative.
Not just cop mentality surveillance tech, but cop mentality surveillance tech meant to appease bad faith actors; a group of people that is prides themselves on never being appeased.
We are already seeing this in art spaces, with some saying that they will assume all art is AI unless the artist livestreams themselves from start to finish. There is no way for authors or artists or other creative workers to sufficiently prove their humanity. And attempting to find ways to prove humanity is ceding the premise.
And I refuse to do that. I am choosing to continue using my weird writing flow that incorporates Plottr, Scrivener, LoreForge, and a gratuitous amount of notebooks and gel pens. If anyone asks for my Emboss for my next release, I'm going to tell them, "Fuck you."
But I debuted in 2017. I have several creative works in the forms of novels, novellas, poetry books, and short stories that predate LLMs. In that sense, I have a ton of privilege that authors debuting this year will never have. I have a history I can point to; I have proof that I could put together a story long before AI. (But even that is not foolproof! Coral Hart once wrote her own books, but now she peddles courses in writing with AI!)
But I am no stranger to surveillance and having it weaponized against me no matter how well or how badly I measure up against an arbitrary number.
I know that my yet-to-be-determined gastro issues would show up as too-long pauses. I know that the days I use dictation will mean pasted words. I know that my frequent breaks will expose the migraine's I am trying to white knuckle through because I don't want to lose _another_ day. I know that the weeks where I don't touch my drafts could expose hospital stays. I know, already, without Ellipsus telling me, how my writing process is bound to my disabilities. I don't need other people knowing that, too.
I know what my revision history would show, too. Sentences cut off because I can't hold myself back from puking any longer. Notes to myself mentioning that my words might get a little weird later cause I just took my Spravato so please come back to edit this more thoroughly when a bit more lucid. Yes, I write down sometimes when drafting when I took my meds; it's helpful for me. But it's not going to be helpful for anyone else, especially not for the purpose of proving humanity.
I fear for other writers with disabilities and what their Ellipsus Emboss metrics will expose if they feel pressured and coerced into sharing it with the people determined to do harm.
Listen. I despise generative AI, too. It's destroying the planet. It's driving up water and electricity prices. It's straining our electric grids. It's built on theft (including theft of my 2017 novel!). It was built by fascists trying to destroy democracy with it and bring about technofeudalism and a network state. But fighting AI can't be solved by subjecting everyone to a panopticon.
In my opinion, the best option we have is waiting for the bubble to pop. Already, tech companies are saying their tokens are too expensive and they are cutting down on AI use and (trying to) re-hire actual humans. As soon as tech companies start charging the actual cost of AI, it will no longer be profitable to write with AI or use AI to generate cover art. Sure, some may run a local LLM on their (very expensive) computer, but that will be a very small percentage and they will likely be using less "human-like" models. It _will_ be obvious, just as it is now usually.
When you encounter a book you suspect was AI, simply stop reading it and move on. No need to draw attention to it. Let the grifters die in obscurity. Eventually, the grift won't be profitable anymore.
Uplift those who you admire instead.
