The Good Slop

This week, this standfirst should make some self-deprecating point about this being slop.

I’m broadly of the view that slop is slop, it doesn’t matter who (or what) created it. So is disinformation, and all of that, but nobody bitched at Adobe for creating Photoshop when the unscrupulous people did their unscrupulous things to images.

And nobody took the keyboard away from us B2B marketers, off our faces on leveraging the leading, paradigm-shifting, unique, data-led, and now AI-powered messaging slop of the day.

But now, I understand that we should care if it’s the dark hand of AI behind it.

AI detectors are being put in tools like Substack, and we can apparently rate our writing as what percentage is robot.

LinkedIn has introduced the thinly disguised dislike button called “Seems like AI slop”

Go on, do it, I double dare you.

Today I saw that Anthropic has signed the EU AI Act's Article 50(2) Code of Practice on Transparency of AI-Generated Content, which means it intends to watermark all content it produces, including, somehow, text, as detailed in the Claude support documents.

As geeks, we are all wondering how Claude will hide the little Easter eggs of clues in its suggested edits of our writing, or research, or whatever we are using it for.

I tipped myself down the rabbit hole of how this would work. Go ask Claude, it’s fascinating, if you like statistical token-biasing techniques. And who doesn’t?

This is all because AI-generated content is bad (right?), and we have a right to know, especially now that the robots are wise to their tell of the em dash and the champion of all writers, Ann Handley, has claimed it back for humanity.

I say the EU needs to mandate the em dash and claim it back for the robots.

We don’t do this for other products; nobody really cares whether their Volkswagen is 97% robot-made or whether the T-shirt I am wearing was designed by an artisan and printed by humans. Of course, I would if this were a Rolls-Royce or a handmade suit, but I don’t need legislation to tell me that.

It seems the idea is that the thing we have over the robots is taste, intelligence, and discernment somehow needs help.

So now we have it in our tools and laws in case we can’t figure out that this is bloody slop for ourselves.

Why can’t we?

Because it’s too good?

Or good enough satisfies a need?

Oh, fuck, so if that’s true - if we like it - is it still slop?

BTW - that's a hyphen, not an em dash — I'm not a robot.

And so, does it matter?