Spammers have been programmatically creating accounts on Mastodon instances for years. Sometimes they post once and never again, sometimes they post on a schedule. Sometimes they never post, like a waiting botnet.
As Mastodon's anti-spam solutions are effectively nonexistent, most instances in our network require manual approval of new accounts. Up until recently, it was easy to spot a spammer; their join reasons were generic.
ChatGPT spammers have now arrived.
@MontgomeryGator As cool as that is, I suspect it would adversely affect the blind.
Almost every neat hack people find to hide LLM traps in are intended as accessibility features 
@mawr There is one way to find out, I've added a payload to this post as well. Please speak up screen reader users!
✨Soff✨ Mawr 
Robots can be gay deal with it
Frost, glow wolf 🐺
<p></p>.
@MontgomeryGator @mawr Haha nice!
Turns out the secret sauce is adding 0x000e0000. So this should work to decode:
perl -CS -pe '@_ = split(//); $_ = join "", map { chr(ord($_) & ~0x000e0000) } @_;'
And this to encode:
perl -CS -pe '@_ = split(//); $_ = join "", map { chr(ord($_) | 0x000e0000) } @_;'
Emoji don't work right, since they live up in that range too. Maybe the solution is to add/subtract 0x000e0000 instead of bitwise set/clearing it.
