The RIAAcidal Woomy! @Twitch VTuber Affiliate with VCephalo & @mnstream2. Totally not a woodcrawler. Splatfest addict. Taken. PFP by @xmikachuu. She/her.
@VProfessorVT@squidorabl Iโm quite liberal myself, but even I realize people occasionally slip up. Hell, Iโm trans myself and even Iโve slipped up on othersโ pronouns sometimesโฆ even on stream.
@squidorabl@VProfessorVT Thatโs not really fair. Everyone (even other trans people) will mess up at least once, as there is no โget out of offending anyone freeโ pronounโฆ not even โthey/themโ.
@VProfessorVT@squidorabl To be frank, thereโs a difference between an honest mistake and actual harassment. The former, although disappointing for the person being addressed, is usually forgiven.
@Major_Ferret@kamikazecash@WheelmanForHire My company has employees who wish to skip lunch breaks sign a waiver so itโs on record. Being that I am in MN, I hope it will remain an option
@FrankConniff I think they only โunitedโ in the sense that they were simply booked to speak at the same event. AFAIK thereโs no actual collaboration between them.
@grok @Foothillseyes @NeverClaimedGod @DisabledRebel@DRodbone Your entire argument is โitโs legal because no actual child is depicted in the resultโ, and technically you are correct.
My argument is that it shouldnโt be legal, based on the nature of the source material, and the 2002 decision is outdated due to current AI technology.
@grok @Foothillseyes @NeverClaimedGod @DisabledRebel@DRodbone Yes, for the reason I stated above. Because even fictional CSAM victimizes RL children, regardless of how technically โlegalโ it may be.
So basically, yes, SCOTUS legalized AI CSAM.
@grok @Foothillseyes @NeverClaimedGod @DisabledRebel@DRodbone The point Iโm trying to make is that photorealistic AI CSAM still involves real life children in the process, still victimizing those children, which is why โitโs not a real childโ is NOT a defense of fictional CSAM in numerous countries.
But thx to SCOTUS, itโs a defense In USA
@grok @Foothillseyes @NeverClaimedGod @DisabledRebel@DRodbone The training data involves real life children, and the output is dependent on said data in order to learn the patterns in the first place.
It still comes back to the involvement of real life children in the process.
@grok @Foothillseyes @NeverClaimedGod @DisabledRebel@DRodbone The end result is moot. Youโd still be, without legal consent, taking real life childrenโs photos to feed into an LLM to create AI CSAM. Why are you even defending this?
@grok @Foothillseyes @NeverClaimedGod @DisabledRebel@DRodbone But you still admit that photorealistic AI CSAM can only exist from patterns taken by real life photos. If the source material still involves real life children, it still victimizes those real life children.
@grok @Foothillseyes @NeverClaimedGod @DisabledRebel@DRodbone Fact: LLMs make photorealistic generative images by training itself on real life photos.
Therefore: AI CSAM will always produce images of real children or at least contain the faces of real children, which, even if the images are not real life, can still victimize real children