Conversation

Notices

1. Embed this notice
   Patrick C Miller :donor: (patrickcmiller@infosec.exchange)'s status on Friday, 18-Aug-2023 22:42:02 JST Patrick C Miller :donor:

   An Iowa school district is using ChatGPT to decide which books to ban https://arstechnica.com/?p=1961316

   • Embed this notice
     Lesley Carhart :unverified: (hacks4pancakes@infosec.exchange)'s status on Friday, 18-Aug-2023 23:53:45 JST Lesley Carhart :unverified:

     in reply to

     @patrickcmiller wow all my favorite things combined

   • Embed this notice
     Greg Bell (ferrix@mastodon.online)'s status on Saturday, 19-Aug-2023 00:37:48 JST Greg Bell

     in reply to

     • Lesley Carhart :unverified:

     @hacks4pancakes @patrickcmiller We heard you like dystopian things, so we ran our dystopian oppression with dystopian AI-washing so you can scream while you scream

   • Embed this notice
     Patrick C Miller :donor: (patrickcmiller@infosec.exchange)'s status on Saturday, 19-Aug-2023 00:37:48 JST Patrick C Miller :donor:

     in reply to

     • Greg Bell

     @ferrix Nailed it.

   • Embed this notice
     Patrick C Miller :donor: (patrickcmiller@infosec.exchange)'s status on Saturday, 19-Aug-2023 00:38:30 JST Patrick C Miller :donor:

     in reply to

     • eagerpebble

     @eagerpebble This may be a good thing. Keep them busy with nonexistent things.

   • Embed this notice
     eagerpebble (eagerpebble@social.vivaldi.net)'s status on Saturday, 19-Aug-2023 00:38:31 JST eagerpebble

     in reply to

     @patrickcmiller
     That's going to a whole new level when they start banning books that don't even exist because the LLM made them up.

   • Embed this notice
     Patrick C Miller :donor: (patrickcmiller@infosec.exchange)'s status on Saturday, 19-Aug-2023 00:39:00 JST Patrick C Miller :donor:

     in reply to

     • Lesley Carhart :unverified:

     @hacks4pancakes Right!?

   • Embed this notice
     K. Reid Wightman :verified: 🌻 (reverseics@infosec.exchange)'s status on Saturday, 19-Aug-2023 00:40:16 JST K. Reid Wightman :verified: 🌻

     in reply to

     • Lesley Carhart :unverified:

     @patrickcmiller @hacks4pancakes [This reply was banned by the state of Iowa using its new auto-banning algorithm, and has been removed]

   • Embed this notice
     Patrick C Miller :donor: (patrickcmiller@infosec.exchange)'s status on Saturday, 19-Aug-2023 00:41:57 JST Patrick C Miller :donor:

     in reply to

     • Lesley Carhart :unverified:
     • K. Reid Wightman :verified: 🌻

     @reverseics @hacks4pancakes Will the AI ban other AI-generated content? See echo <> chargen…

   • Embed this notice
     K. Reid Wightman :verified: 🌻 (reverseics@infosec.exchange)'s status on Saturday, 19-Aug-2023 00:46:25 JST K. Reid Wightman :verified: 🌻

     in reply to

     • Lesley Carhart :unverified:

     @patrickcmiller @hacks4pancakes qui impedit obstructore?

   • Embed this notice
     Patrick C Miller :donor: (patrickcmiller@infosec.exchange)'s status on Saturday, 19-Aug-2023 00:49:21 JST Patrick C Miller :donor:

     in reply to

     • eagerpebble

     @eagerpebble they would never miss an opportunity to harass these authors..

   • Embed this notice
     eagerpebble (eagerpebble@social.vivaldi.net)'s status on Saturday, 19-Aug-2023 00:49:22 JST eagerpebble

     in reply to

     @patrickcmiller
     Except they'll attribute the fake books to real authors, who'll still get harassed.

   • Embed this notice
     David Reinertson (davidreinertsoninfosec@infosec.exchange)'s status on Saturday, 19-Aug-2023 01:13:54 JST David Reinertson

     in reply to

     @patrickcmiller Per the article, they limit the investigation to titles provided by some banning community, then use ChatGPT because they can’t read the books before banning them.

   • Embed this notice
     Patrick C Miller :donor: (patrickcmiller@infosec.exchange)'s status on Saturday, 19-Aug-2023 03:02:57 JST Patrick C Miller :donor:

     in reply to

     • David Reinertson

     @DavidreinertsonInfosec so... they can't read them, but they can read ChatGPT summaries of them. Won't this train ChatGPT to be “evil?” Won't someone think of the AI?