Post

Amyville Hoyrror
@amyhoy.bsky.social

i’m in the game of marketing and could not disagree more. lying sets a bad precedent. if your marketing makes a lot of people realize you can’t be trusted, you’re doing a bad job. and an immoral one.

Amyville Hoyrror
@amyhoy.bsky.social

yes, the “cruel to our models” part. it’s misleading to act as if the models have any kind of experience or that you can be cruel “to” software. you can’t. and they put it that way knowingly and intentionally.

austin (eeek!/ackkk! đź‘»)
@thebadcode.com

i take your point, but i think “cruel” is a description of the state of mind of the actor, not the subject of the cruelty. if you said “ant is deliberately blurring the existence of that question in a way that favors a maximal reading of their product’s abilities” id agree for sure though

0 reposts3 likes

Amyville Hoyrror
@amyhoy.bsky.social

if it were a one-off i would interpret it charitably but it’s not, they anthropomorphize every chance they get, even in the docs / explainers. to say nothing of endless PR to spread their religion/cult fantasy of “rogue” “AI” there are totally valid reasons to dump those chstomers

Amyville Hoyrror@amyhoy.bsky.social

this would be fine and normal if they say "it's bc sometimes humans have to read those logs" or "the things you say to the chatbot change the things it says back to you, and this could create a negative spiral for the user" but no, of course not. they pretend they invented a digital slave. freaks

Hans Gerwitz
@gerwitz.com

I cynically think this is only another (maybe subconscious publicity stunt. But I do like the precedent of banning asshold behavior from customers. Meta could learn from this.