Extelec,
@Extelec@mstdn.social avatar

Oh the hypocrisy from Stack Overflow !

gsuberland,
@gsuberland@chaos.social avatar

@Extelec the difference is that it's only banned because moderators pushed really hard for it, against the original wishes of the company, who were entirely OK with having people use it for answers. there was a whole heated thing about it and lots of moderators threatened to resign.

Extelec,
@Extelec@mstdn.social avatar

@gsuberland but haven't they recently gone to selling their (our) soul to AI ?

gsuberland,
@gsuberland@chaos.social avatar

@Extelec the company is trying to, yes. moderators are community members, not employees of the company. and the moderators are the ones enforcing the "no LLM answers" rule.

Extelec,
@Extelec@mstdn.social avatar

@gsuberland Excellent !!!

Shame their policy document has gone from the site though.

gsuberland,
@gsuberland@chaos.social avatar

@Extelec given that's from a year ago I'm not surprised it linkrotted; the policy has changed a lot since then.

gsuberland,
@gsuberland@chaos.social avatar

@Extelec the really unfortunate thing about this openai deal is that it's putting moderators who are against the use of LLMs, and even those who are against the openai deal, in contention with users who oppose the openai deal and want to delete their posts. since you can't delete old posts with accepted answers and points above a certain threshold, all you can do is edit your post, but that harms the quality of the site which moderators obviously aren't happy about.

gsuberland,
@gsuberland@chaos.social avatar

@Extelec you can delete your account entirely, but even then your posts stay behind and are just shown as being written by "deleted user", so that doesn't prevent them being harvested into the openai deal. and, worse still, by removing your name from the work you potentially weaken your position against claiming plagiarism later.

Extelec, (edited )
@Extelec@mstdn.social avatar

@gsuberland and then AI trains itself on its own answers. I cant see how that could possibly cause an issue.

Its just so short sighted.

jwarlander,
@jwarlander@mastodon.nu avatar

@Extelec @gsuberland Yeah.. That whole AI-training-on-AI-output thing is going to be fun, going forward 😂

Anyone who did a few rounds of back and forth computer translation between two languages back in the 90s know how hilariously bad that turned out.

Extelec,
@Extelec@mstdn.social avatar

@jwarlander @gsuberland More consignments of "water sheep" :)

Extelec,
@Extelec@mstdn.social avatar

and their link to their gpt-policy (page https://stackoverflow.com/help/gpt-policy )

  • All
  • Subscribed
  • Moderated
  • Favorites
  • random
  • ethstaker
  • DreamBathrooms
  • InstantRegret
  • magazineikmin
  • ngwrru68w68
  • Durango
  • thenastyranch
  • Youngstown
  • rosin
  • slotface
  • tacticalgear
  • mdbf
  • kavyap
  • khanakhh
  • provamag3
  • osvaldo12
  • GTA5RPClips
  • cubers
  • cisconetworking
  • everett
  • tester
  • modclub
  • megavids
  • Leos
  • normalnudes
  • anitta
  • JUstTest
  • lostlight
  • All magazines