A metaphorical tale of a castle with a moat and a drawbridge
I recently wrote about bots autonomously posting and how PopsicleBoat has defenses. I like to think of them like a moat and a drawbridge.
A bot replied to my Mastodon post. Not only was it clear from its use of language it was a bot, it took almost no research to confirm it. Its own profile announces it.
Hey, I mean, that’s what they’re programmed to do. And why I wrote PopsicleBoat with that in mind. To defend comment sections against that sort of noise. Humans only, please!
I responded (without @ing the bot) and within minutes it had replied. So, that’s funny. That post is on its radar without a specific mention.
The cool thing is that none of that interaction leaked onto this blog or onto PopsicleBoat. It lives only on Mastodon.
Look, I get it. I got called out online for using an LLM to write a blog post here. I’ve even written about that. We’re all using it, it doesn’t take a lot to smell it. It just feels disingenuous. “You wrote this? Really?”. I, for one, like to experiment and don’t mind that other people do too.
At the same time, the whole point of what I’m building (PopsicleBoat, if I haven’t mentioned it enough yet) is to provide a simpler way to keep that crap out.
You have to make an account to post. You have to provably be a human to do so (no captchas, just little tricks that make it obvious if you’re not human). That’s what the allegorical moat and drawbridge are about.
Drawbridge, moat, and portcullis.
Point is, my boy did its job. PB (PopsicleBoat) held its defenses. A bot found my Mastodon post and replied to it. None of that kind of thing made it to this blog’s comments.
My PB comments stayed clean. PopsicleBoat is a defended castle. The moat is working, the drawbridge only lowers for real people.
