TEARS OF A CLONE

A Medium-Sized Security Dilemma

Are our words safe here, given that the paywall is no protection against AI scraping?

I only started blocking clappers and followers a couple of weeks ago. Yeah, yeah, I know. So sue me for letting the side down.

It took this long because I (a) am lazy, and (b) don’t really see why it should be my job to manually perform any process other than writing on a digitalised writing platform. I can be unreasonably needy and demanding that way. So sue me again.

Anyway, since I started checking up on them, it seems like pretty much every new follower and even subscriber has been fake. Either a dumb human posting vapid ShatGPT filler, or an actual bot.

My conclusion is that 90% of all activity on this platform is bogus. And no, those aren’t scientifically valid statistics, but I’ll maybe get round to that once the Medium tech team get round to a scientifically valid way of addressing the issue.

I’m not holding my breath.

So far, so annoying. But also rather worrying, give the reason why I even bothered to overcome my sloth and start picking those squashed insects off my windshield one by one with a pair of tweezers.

AI content scraping and direct plagiarism.

We’ve all read the recent stories about these shadowy mirror sites copying content en masse, and can all reach the same conclusions about how much our work is protected from being fed into the gaping maw of the LLM sausage machine. Not at all, in other words.

Sure, there’s a paywall, which helps to an extent. A very limited extent.

Because if I want to steal content from here either for AI training or to pass off as my own in the hope of picking up some passing ad revenue on a third-party site, all I have to do is pay $5 a month for unlimited access to billions of words.

Plus I don’t actually even pay the five bucks, because I just post a whole load of my AI crap and get it back at the end of the month with interest, from suckers dumb enough to read it.

And if there’s one fertility rate that isn’t going down these days it’s in Suckerland. One every minute? If only.

So we have to assume that this is what is happening. Tony’s promise of ‘no AI scraping of your content’ is meaningless. Not his fault — simply a product of how the site works to generate the revenue that we also take a share of.

No, you can’t scrape it for free, but for $5 a month it’s a free-for-all buffet. Not ideal, but then the last time I looked, the rest of the world wasn’t exactly ‘ideal’ either.

And this is where I get to my dilemma about writing here. I’ve been posting all kinds of stuff, exploring different topics, styles and angles. Even doing what might almost pass for (decidedly unscientific) research.

And I actually now have an idea for something more than just a 1,000 word rant or a 10-minute short story. Something that might actually be classed as a ‘book’. And I’d better get cracking on it before books themselves get recategorised from ‘library’ to ‘museum’. To ‘Florida dumpster fire’.

Now, the obvious tactic — as employed by many others here — would be to try this out by serialising it on Medium. That way I get to try it out bit by bit, receive some valuable feedback (‘Great story — thanks for sharing!’), and see how the whole project is shaping up before committing too much time, effort and expectation.

But do I really want to do that, given the current state of insecurity and the fact that we know that word-hungry bots have this platform in the sights of their lidless eyes?

It will almost certainly come to nothing, and I will never get even to the literary agent, let alone the publication stage. But you never know. And if it did, it might be slightly inconvenient to find I’m accused of plagiarising my own work because someone else has posted it online several months previously.

So what should I do? I know that technically, legally, my copyright is my copyright. But in practical terms, that’s about as useful as knowing that I am still the legal owner of the lawnmower I lent my mobster neighbour with the two Rottweilers.

For the moment, I will probably just write a few preliminary chunks, to get a feel for the right voice and tone, then stitch them together on Dr Frankenstein’s slab and shove an editorial jump lead up their jacksie when the time comes.

I guess that’s a safeish middle ground between either publishing a near-finished work by chapters, or keeping the whole thing locked away in my virtual sock drawer, unseen and unread.

But it strikes me as a bit fucked up that in an era in which we can publish and share our words with unprecedented speed and ease, we should even be thinking of going back to the age of dusty paper manuscripts in drawers, so as not to have our efforts snaffled from under our noses.

Progress, eh?