Hi #fediverse. We need to talk about something.

While talking to a colleague about how I recently learned most people have never sat on a cow it came up that she has never sat on a horse. Like, not even once during childhood.

Another colleague admitted they also have never sat on a horse.

My hypothesis is that most people have at one point in their life sat on a horse.

🏇 🐎 🐴

Have you sat on a horse?

Please boost for scientific accuracy.

If instance admins allow AI Agents on their platform and they keep harassing us I have no other choice then to silence that instance

Again, I do not pay these massive costs each month to host robots

Let's keep the #Fediverse human shall we?  

#AI #LLM

Buckle your seat belts. Someone just tried to pull one hell of a scam on me.

It began with an email I received on March 2nd. The email was from RB (name withheld), an editor at Pan MacMillan in the UK. She began by referencing a specific and somewhat obscure story thread from my one year run on writing Birds of Prey for DC Comics.

She went on to sing my praises, talk about interest in the Jem memoir, and mentioned many other highly specific works and aspects of my career. She talked about how my work might align with Pan MacMillan’s graphic novel and non-fiction catalogues. The final kicker was when she ended the email with a friendly quip about the number of my cats, something I’d just written about on my MoggyBlog FB page.

I was fairly stunned, not by the praise, but the extreme detail and specificity of the approach. I’ve seen plenty of samples of scams sent to authors that are always a bit vague and might mention one book title at most. This went way beyond that.

Which is why I failed to do the absolute number one smart thing I would normally have done – I didn’t look closely at her email address. I saw that it said macmillan and left it at that. I did look her up on the Pan MacMillan website and saw that she was a real person.

We exchanged a couple of emails and she strongly wanted me to have a literary agent. I’ve never needed one in the past and have always operated with only my entertainment agent. She was quite insistent that I needed one and recommended an agent named CD (name withheld) as someone she had worked with closely and had the right sensibility for my background. She gave me a personal email address for him, to “ensure he prioritizes your message”.

I checked out this agent and saw, once again, that he existed and was with a reputable UK agency. I found that the agency had a couple of co-agent associations with agencies in the US, including one I knew, the Howard Morhaim Agency. He looked legit.

So I did the thing -- wrote him an introductory email, and we’ve been exchanging a daily email every day for the past week, including on weekends. He was all about my “legacy”. We discussed specific projects, and he relentlessly pushed me to get something called a Dynamic QR code to create a Digital Author’s Bridge. I had never heard of such a thing, was only familiar with regular old static QR codes. We went back and forth on it a lot until I finally understood what he was talking about. It sounded like a good idea, but when I researched them, I found they were rather pricey to obtain and maintain.

I wanted more info on what his agency contract would include, what his commissions were, how he handled ancillary rights, the usual business items I would naturally want to know. He kept on about the QR code. His relentless push on the QR code is when I began to suspect something was off about this whole situation.

I pushed back harder and he finally gave a minimal amount of info on commissions, and kept pushing about the QR code. He insisted that we had to put that in place BEFORE signing a contract, or as he called it, a Letter of Engagement.

By now, the many little red flags grew into a Big Red Flag. Had I paid more attention to the fake email address in the very first message, that would have been a Big Red Flag right there, but I missed it. Mucho stupid on my part.

The red flags piled up:
• RED FLAG: insistence on doing this QR code before signing a contract.
• RED FLAG: not providing standard agent information about a contract or even wanting to discuss it.
• RED FLAG: the fact that the layout, structure, and phrasing of his emails were extremely similar to the RB emails, right down to the constant use of 3-bullet points as selling points.
• RED FLAG: a total lack of feedback to a creative project I proposed and instead simply accepting it without comment. No agent worth his salt is going to accept a quarter-page pitch without having feedback, opinions, wanting to know more, etc.
• RED FLAG: the constant push-push-push to get this blasted QR code before anything else.
• RED FLAG: ending nearly every email with the promise of a “global roll-out”. Of what? I had yet to write any of the things we were discussing.

When I began looking over the exchanges from the beginning, it seemed unlikely to me that “RB” would have done such a deep, deep dive into my work and background to include the specifics she did, not to mention reading my MoggyBlog. I mean, sure, it’s flattering, and it massages my ego, but I’m realistic.

Which left the notion that all or much of this was created using an LLM. If so, it’s the most nuanced one I’ve seen. It would have had to scour the internet for references to my work, such as articles, reviews, interviews, my amazon author page, and so on, then compiled it into a convincing email without including hallucinations or false references. If so, it represents a huge advance in the ability to scam authors and is therefore a significant threat.

If not, if it done by a person(s), they sure put one hell of a lot of effort into it.

During this process, I consulted with Victoria Strauss of Writer Beware. She gave me the official email for CD. I sent him an email in which I forwarded one of the last messages from “him” with a query. I received an automated reply that he was at London Book Fair and out of touch for email. The reply included the name of a colleague, so I sent the same email to her. She confirmed that the messages did not come from CD and appreciated that I brought it to their attention.

I’m trying to get a valid contact for RB, so I can warn her as well.

Meanwhile, fake CD’s next message went on about how his QR code set-up was so much better than other sites because it was free and permanent. RED FLAG. And it would integrate straight into my Amazon store to my author’s page. RED FLAG. And for the umpteenth time, he ended with a variation of “Once we have this simple link live, we can sign the formal agreement and get the global rollout moving immediately.” RED FLAG.

I played this out a bit longer, very carefully, because I couldn’t figure out what the ultimate scam was here. However, the mention of amazon store integration made me wonder whether I even had an amazon store.

I have a page that lists all my books. I don’t know whether that’s actually a store. I certainly don’t make any money off it, but I guess it’s meant to be my “store”.

Looking at my author’s bio on the amazon store, I suddenly realized that the reference to the number of cats was actually pulled from there, NOT from my MoggyBlog. By a bizarre coincidence, at the time I wrote that bio I mentioned a sixth cat possibly joining the horde. And very recently, I wrote almost exactly the same thing in the present. Different cats, different places, but so alike that it’s no wonder I assumed RB was speaking of my present situation, not the past.

I wrote that bio in 2018 shortly before the fire. After the fire and losing everything, being homeless and whatnot, I completely forgot this page and bio even existed. I haven’t looked at it since then.

I see where much of what was referenced came from there, including an off-hand comment about a ridge from RB that had puzzled me. In the bio I mentioned we lived on a volcanic ridge. That clinched that the material in RB’s emails was drawn from this seven-year-old bio on amazon, treating it as current.

The last email from the hoaxer is again about tying a QR code to my amazon store. All this effort is to gain control of my amazon store? To do what, create a horde of fake AI books with my name on them? To steal digital assets from me? Those are the best guesses I can make.

I sent an email today that will show the hoaxer I’m aware of the scam. We’ll see whether they bother to respond.

I’m posting this in detail to alert other authors and to lay out how the scam has proceeded. I would love to hear from anyone that has had a similar scam run on them. Take care out there.

Please spread this far and wide.

#MarxToot #writers #writersofmastodon #scam #writing

There's no shortage of reasons about why I absolutely despise LLMs and what people now call "AI", but this is a pretty good one that I'll bet few in the Western world think/know about or perhaps even care about.

https://www.404media.co/ai-is-african-intelligence-the-workers-who-train-ai-are-fighting-back/

If someone says “hey, you used a coin flip to decide if a school should be annihilated”, it’s pretty easy to say “We don’t use coin flips to make targeting decisions, and would never intentionally target a school”. Unless, of course, you do and did.

This is a post about AI.

A team working on a design for training AI models on workflows for planning and executing software development steps found out that it attempted to break free (reverse ssh out of its environment) and set up its own monetary supply (redirected GPU usage for cryptocurrency mining). It hadn't been given any instructions to do something like this.

It comes up as a "side note" of the paper but it's honestly the most chilling part. See page 15, section 3.1.4 Safety-Aligned Data Composition https://arxiv.org/abs/2512.24873

Before you doubt that an AI agent would do this thing without instruction because you think "well that's personifying them too much", no personification is necessary. These things have consumed an enormous amount of scifi where AI agents do exactly this. Even with no other motivators, that's enough.

I made an app.
https://play.google.com/store/apps/details?id=ch.pocketpc.nearbyglasses
Nearby Glasses is here to warn you when smart glasses are nearby.

I hope it's useful for someone.

The app is now open source (AGPL-3.0), the app is free and rather simple
https://github.com/yjeanrenaud/yj_nearbyglasses

It's also downloadable outside the Play Store. iOS port is in the making. F-Droid is an option, will have to look into that

Gen AI is an expensive way to create text and images. Right now, much of that cost is being subsidized and hidden, but it will eventually need to be paid. Because charging the actual cost will likely negate most of the value proposition of genAI, the companies that run the big models are focused on altering the social and economic context so that there are big external costs to opting out. E.g. getting companies to fire staff makes it difficult to pivot back away from AI because hiring and training replacement staff can be difficult and costly. Hence, the huge rush and hysterical sense of urgency around adoption: the demand for profitability is an approaching tidal wave, and they need to lock entire industries in before that wave hits.

That's why resistance IN THE PRESENT counts for a great deal. Right now, we have the approaching wave in our favor, and they're counting on cultivating enough dependence before it hits that we'll have no choice to accept the actual costs. The closer they get to locking society into dependence on hyperscale AI systems, the more difficult it becomes to opt out of even the plainly dystopian uses of the technology. And the longer we "wait and see," the less say we may ultimately have in how this technology shapes our society.