Conversation

Notices

1. Embed this notice
   In #Flancia we'll meet (flancian@social.coop)'s status on Saturday, 22-Mar-2025 07:38:38 JST In #Flancia we'll meet

   AI is overall good in the long term, like humans -- just sayin'

   • Johnny Peligro likes this.
   • Embed this notice
     anderbill (band@hachyderm.io)'s status on Saturday, 22-Mar-2025 07:38:37 JST anderbill

     in reply to

     @flancian so you claim. it is an empirical question in several ways

   • Embed this notice
     anderbill (band@hachyderm.io)'s status on Saturday, 22-Mar-2025 07:42:07 JST anderbill

     in reply to

     @flancian for me, hope is often an escape, a way to avoid action, and, yes, i often rely on it.

   • Embed this notice
     In #Flancia we'll meet (flancian@social.coop)'s status on Saturday, 22-Mar-2025 07:42:08 JST In #Flancia we'll meet

     in reply to

     • anderbill

     @band yes, I agree! this was a statement of hope. over at work we have this slogan of sorts, "hope is not a strategy" -- but I would posit in some ways, sometimes, it is :)

   • Embed this notice
     iced depresso (icedquinn@blob.cat)'s status on Saturday, 22-Mar-2025 07:50:10 JST iced depresso

     in reply to
     @flancian how so
     it's just another bougie kit billionaires get to own and use against everyone, like all the machines that came before
   • Embed this notice
     iced depresso (icedquinn@blob.cat)'s status on Saturday, 22-Mar-2025 08:53:18 JST iced depresso

     in reply to
     @flancian the AGI probably will unless they succeed in figuring out how to properly lobotomize them.
     
     the little ML toys they call AI are kind of a mess.
   • Embed this notice
     In #Flancia we'll meet (flancian@social.coop)'s status on Saturday, 22-Mar-2025 08:53:19 JST In #Flancia we'll meet

     in reply to

     • iced depresso

     @icedquinn I think in the long term AI will be aligned with humanity, because it is a mirror/lens of humanity (or will have started as one). This was me being optimistic and assuming that humanity also has an overall positive impact on the universe, despite our many shortcomings and looming risks

   • Embed this notice
     dilbert 1 (sun@shitposter.world)'s status on Saturday, 22-Mar-2025 08:56:56 JST dilbert 1

     in reply to

     • iced depresso
     • anemone
     @anemone @icedquinn @flancian ai ethics has just normalized the most extreme form of censorship I can think of, a learning tool that is programmed to defy you
   • Embed this notice
     anemone@ebiverse.social's status on Saturday, 22-Mar-2025 08:56:57 JST anemone

     in reply to

     • iced depresso

     @flancian@social.coop @icedquinn@blob.cat the problem is it's likely to mirror the people creating it who seem to be really shitty people in many cases

   • Embed this notice
     iced depresso (icedquinn@blob.cat)'s status on Saturday, 22-Mar-2025 09:03:22 JST iced depresso

     in reply to

     • anemone
     • dilbert 1
     @sun @anemone @flancian if they are built to properly immitate how the human works they will probably be fine, as most people are only a basic level of evil (peter singer.)
     
     but yes the alignment folk have very much been trying to figure out how to lobotomize something to agree with them, while pretending its for everyone else.