remember when google's ai whatever thing recommended people to put glue on their pizza and it was traced back to a 15 year old reddit post written by someone named "fucksmith"
-
remember when google's ai whatever thing recommended people to put glue on their pizza and it was traced back to a 15 year old reddit post written by someone named "fucksmith"
-
replied to enderman0125 last edited by
@[email protected] this is kind of misleading... the thing is that it's not specifically to that comment. it's a common technique in food commercials where they're slicing pizza and want to show the "stretchiness" of the cheese - that's what the glue is for. I mean on one hand that's misleading on the other hand prop masters have to do a lot of weird techniques with food because it's prone to spoilage and you're going to be taking potentially hundreds of shots. Another interesting thing with that - if there's pepperonis on the pizza they have actually screwed down the pizza to the plate. It's one or two slices that they use to lift up from the pizza.
-
replied to Amber (deilannist) last edited by
@[email protected] I mean yes, okay. The comment definitely had some factor in it but I am not sure a single reddit comment could poison the entire dataset like that. At least not for something that's widely documented.
-
replied to Amber (deilannist) last edited by
@puppygirlhornypost2 its funny though
-
replied to enderman0125 last edited by
@[email protected] true i will admit it is fucking hilarious and it's proof that these LLMs are just approximations and not actually using logic.
