Well, with models like SD at least, the datasets are large enough and the employees are few enough that it is impossible to have a human filter every image. They scrape them from the web and try to filter with AI, but there is still a chance of bad images getting through. This is why most companies install filters after the model as well as in the training process.
Broken pasta is good for soups where the noodles should be small, such as chickpea or lentil soup. Otherwise, if you have a pot large enough to hold the noodles, there is no reason to break them.
Even with automating shitty jobs that no one wants, you’re still getting people out of a job and the only way they have of making money. This is kind of how people reacted when Boston Dynamics showcased its warehouse robot. It seems that we need a universal basic income first, but no politicians are willing to do that at least until unpleasant jobs are automated. There’s a bit of a chicken and egg problem there. And, on top of this, companies don’t care that much about automating shitty jobs because the people in them get low wages, so they don’t cost the company much to employ.
As someone who has tried out a lot of open source image AIs, I would say that ‘original art’ is something it can’t do. It can make a lot of stuff, but if you deviate too far from the topics it knows it just stops giving you what you asked for. In addition to this, most of the originality the generations do have comes only from the prompt.
toxic help forum (lemmy.world)
Every base is base 10 (lemmy.world)
old piston sound (lemmy.world)
FBI Arrests Man For Generating AI Child Sexual Abuse Imagery (www.404media.co)
Conservatives are freaking out because they learned that some animals are gay (www.lgbtqnation.com)
Anon wants to ride a zeppelin (sh.itjust.works)
me_irl (lemmy.world)
Nope. No. (lemmy.world)