Aggregatet
  • Communities
  • Create Post
  • heart
    Support Lemmy
  • search
    Search
  • Login
  • Sign Up
David Gerard@awful.systemsM to SneerClub@awful.systemsEnglish · 11 months ago

what if, right, what if our super-duper-autocomplete was just tricking us so it could TAKE OVER ZEE VORLD AHAHAHAHAHAHA! that'd be wild, hey

www.lesswrong.com

external-link
message-square
32
fedilink
4
external-link

what if, right, what if our super-duper-autocomplete was just tricking us so it could TAKE OVER ZEE VORLD AHAHAHAHAHAHA! that'd be wild, hey

www.lesswrong.com

David Gerard@awful.systemsM to SneerClub@awful.systemsEnglish · 11 months ago
message-square
32
fedilink
New report: "Scheming AIs: Will AIs fake alignment during training in order to get power?" — LessWrong
www.lesswrong.com
external-link
I examine the probability of a behavior sometimes called "deceptive alignment."
  • barsquid@lemmy.world
    link
    fedilink
    English
    arrow-up
    2
    ·
    11 months ago

    Sorry the thesis is that checks reality gradient descent might be consciously trying to avoid having its nefarious goals overridden?

    • David Gerard@awful.systemsOPM
      link
      fedilink
      English
      arrow-up
      3
      ·
      11 months ago

      what if right my spellcheck dictionary got so big it TOOK OVER makes u think

      • barsquid@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        11 months ago

        It is imperative that we first build a mathematical framework for guaranteeing benevolent thesauri before we travel this path any further!

        • David Gerard@awful.systemsOPM
          link
          fedilink
          English
          arrow-up
          3
          ·
          11 months ago

          Urban Dictionary’s Basilisk

      • V0ldek@awful.systems
        link
        fedilink
        English
        arrow-up
        3
        ·
        11 months ago

        If we grow AIs too big, say, bigger than the Moon, then well, the Moon could get jealous and mad at us.

        • skillissuer@discuss.tchncs.de
          link
          fedilink
          English
          arrow-up
          2
          ·
          11 months ago

          Not enough people are preparing for this.

SneerClub@awful.systems

sneerclub@awful.systems

Subscribe from Remote Instance

Create a post
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !sneerclub@awful.systems

Hurling ordure at the TREACLES, especially those closely related to LessWrong.

AI-Industrial-Complex grift is fine as long as it sufficiently relates to the AI doom from the TREACLES. (Though TechTakes may be more suitable.)

This is sneer club, not debate club. Unless it’s amusing debate.

[Especially don’t debate the race scientists, if any sneak in - we ban and delete them as unsuitable for the server.]

Visibility: Public
globe

This community can be federated to other instances and be posted/commented in by their users.

  • 7 users / day
  • 10 users / week
  • 5 users / month
  • 690 users / 6 months
  • 1 local subscriber
  • 1.04K subscribers
  • 157 Posts
  • 2.43K Comments
  • Modlog
  • mods:
  • self@awful.systems
  • blakestacey@awful.systems
  • David Gerard@awful.systems
  • BE: 0.19.8
  • Modlog
  • Instances
  • Docs
  • Code
  • join-lemmy.org