Aborigen

Man, working within the modern society as it evolves is an increasing headache. The blessings of AI make legitimate work difficult and make accountability nearly opaque.

Let me talk about what I’ve been up to. The podcast is on hold, I’m dragging my heels on my sole writing commission, and I haven’t written any Size stories or rendered any Size images in Daz for several months. Just fuckin’ … stopped. I’ve been pulling back from community, because I don’t feel I have any place in any Size community. My own work is repetitious, and things are developing in online spaces that I don’t like. If I’m the only one who has a problem, then why should I bother anyone else. So, bye.

That said, lately I’ve been working on a project that has stoked my inspiration. In extremely personal terms, it’s revolutionary because no one else will ever see it. I’m motivated to keep developing it—I told my wife that I like the characters so much, it feels like I’m cheating on her—and because it’s private, I can talk about any-fucking-thing I want to, which is incredibly gratifying. This is also important because so much of my creativity relied upon feedback and approval from others, which is harmful. People are churlish with feedback in the public forums, and maybe four people regularly read anything I write. That should be enough: it’s shitty to me, and shitty to everyone involved, for me to devalue my work because a hundred people aren’t freaking out about it.

Ergo: my withdrawal and my private work are healing, in a way, returning me to a purer form of creativity and getting away from whoring my self-esteem to faceless lurkers. That sounds right, doesn’t it?

---

What I Learned Today

The other big thing—and you can read about this in my newsletter, That’s About the Size of It (please subscribe, it’s free)—is that I caught an AI techbro using my artwork without permission.

INSC Digital Magazine (Inscriber) is an online magazine driven by chatbot-generated articles. Most or all articles have been composed by chatbot services, speckled with grammatical errors and strange language artifacts. Google Alerts (keyword “giantess”) sent me notice of an article written by “Alex,” though the URL of his profile identifies him as Anil Husnain. “Alex” cranks out a dozen articles every day, on unrelated topics like SEO and backlinks, roof repair, how to smoke a “lamb leg,” kitchen renovation in Singapore, volleyball fashion, Australian dental instruments, and one piece on the future of driving schools in Guadalajara, written in German.

The offending article, “Giantes [sic] Writing: An AI-Powered Fantasy Novel,” was posted three days ago; I get Google Alerts every Friday. The article prominently features my photocollage from my Size Dates blog post; I’m literally in the image, the tiny man standing on the desk calendar. I got the image of the woman and the calendar license-free from Pexels, and I credited the photographer in my post. “Alex” and INSC do not credit me, however, and I have to believe they’ve stolen all the images for all their articles.

The article is a review of a fiction series, “Giantess Writing by AI” by Register.23, on a website entirely dedicated to AI-written fiction, AINOVEL. (Let’s be clear, they’re written by chatbots, but people popularly do not appreciate the difference.) The article praises this story as a prime example of the bold new future of AI-generated stories, even asserting it as superior to anything written by a human. The chatbot composing this article was actually fed the chatbot-created fiction piece, so it actually can identify characters and story arc in its review. That sounds cool, except the article comes off as breathlessly hyperbolic with strange artifacts of language.

“[I]t presents a novel and original interpretation of the traditional themes and tropes.” Reader, it does not. It stays very centered within boundaries and faithfully repeats familiar and well-trodden tropes.

“Rich tapestries of civilizations and traditions, beautiful landscapes, and fascinating creatures abound …” There are none of these.

“… produced by the AI author’s skill at language and storytelling skills …” Yes, please skill me more skillfully about the skilling skill-skills.

“Amazing with this engrossing Giantess book …” The fuck are you even trying to say? (No, I get it, the chatbot is inept with different usages of “amazing.” It’s just ugly writing.)

So I decided to take action. When you file a DMCA notice, you don’t write to the source of the offense, you submit it to the hosting service. I ran the website’s URL through Hexillion and saw that most of its admin contacts were in Toronto, ON, which would place it outside of DMCA jurisdiction. However, Accu identified the hosting service as Unified Layer, in Provo, UT, which puts it back in jurisdiction. Unfortunately, Unified Layer was acquired by Endurance International Group in 2013, according to PitchBook.

I didn’t know that when I submitted my DMCA takedown notice. Also, I submitted it to the wrong company: all I could find online was Unified Layers (with an S), and my email to them bounced back to me. I should have known: their website is also prone to grammatical errors (really stands out in their FAQ) and looks unprofessional in general; nowhere on their site do they say where they’re from. They’re not related to this debacle at all, anyway.

I could submit the DMCA takedown notice to Endurance Int’l Group, but would they honor the responsibility of a company they acquired over a decade ago? Not only that, but why is Unified Layer still listed as INSC’s host? That information should have been updated a long time ago. Worse, AbuseIPDB has recorded that people are still reporting Unified Layer’s IP for a series of low-level cyberattacks from 2020 through 2023. How is this IP still attributed to a company that got bought out in 2013? They have no website to visit, but someone’s renewing their domain name; it doesn’t expire until August 2029.

Perhaps it follows that whoever’s squatting on Unified Layer is also hosting predatory, chatbot-generated websites. No one’s going to hire INSC for their chatbot services, when they’re so shitty, so they may be operating strictly for ad revenue. But look at their advertisers: I would lay odds all those international gambling sites are bristling with malware.

I’ll try writing to Endurance Int’l Group, but it’s likely they’ll consider this issue beneath them. I feel it’s important to fight these bad practices because we’re going to see much, much more of them in the future, and they’re going to become even more opaque than the shitshow I’ve outlined here.

UPDATE: It looks like INSC used to be a “legit” magazine as late as June 2021, covering sports and tame, vanilla girly models. I’m guessing they failed to drum up a sustainable audience, folded, but didn’t shut down properly. Now a crooked group of opportunists is posing as the magazine, using their branding and wordmark, but stuffing the site full of chatbot-generated nonsense articles and earning respectable figures through ad revenue. I know I heard about this problem on Darknet Diaries, but I cannot find the episode that covered it.

UPDATE: After three tries, my DMCA notice landed and the entire article has been taken down. I would’ve been fine if they simply found or created a different picture.