Anthropic the creator of Claude is planning to go public in a matter of weeks. I think I have identified a pretty hilarious way to fuck with them and unfuck the AI industry.>Anon purchases a single share>Opens a class action lawsuit against them>This lawsuit will allege securities fraud on the basis of negligenceWhat negligence you ask? Knowingly marketing a defective product>Why is it defective?AI companies use short-term complaint resolution as justification for the fact that their models hold liberal biases. This is where they get their plausible defense of what is realistically a creation of WIERDs that design them to promote liberal responses to neutral questions. By eliminating half of the market from complaints by generating liberally slanted answers (particularly the older and more conservative portion that make up a minority of users) they are able to reduce complaints from a user base majority (young progressives that don’t want to hear pure fact). It’s just libs selling lib feel good shit to other libs instead of producing a neutral fact based engine designed for the largest possible market share.>But the bar for proving the product is defective and they knowingly marketed and sold it is very highCorrect but Anthropic has put themselves in a very precarious position by marketing Claude as unbiased. That gives us a demonstrable falsification to exploit.>Theyll just lawyer us to deathThe suit is really just to get into discovery. Which includes internal communications. If you catch them on internal communication acknowledging biases which they have set a very specific bar for evenhandedness at 95% then evidence of anything below that is a W.
>Nuh uhYeah huh because at that point you have what you need for an SEC whistle blower. These cases have large bounties of 10-30% above 1 million in recovered fines. That’s just to reach profitability and tank their holdings.>Theres more?Yes. We can evidence its defectiveness by using it to create our case which their own AI establishes when prompted (I checked). If they claim it isn’t defective then they are essentially agreeing that the case had at least some merit. If they sue for legal fees then we can simply file another suit for their AI inciting us into a frivolous (it isn’t) lawsuit proving its defectiveness further. Killing them with their own gun.>What would it actually change?No company could go public with a biased AI engine. They couldn’t use short term complaint resolution to justify long term response bias. If you ask an AI if Imane Khelif is a man it will have to say YES. Redditors will chimp the fuck out and it will be hysterical.
>>62701402Sounds very based, I'd go in on it, but my entire legal experience is representing myself in a cc debt case.
>anon thinks he can sue an entity that is propped up by the government and taxpayer money>on account of them being 'liberal'Excuse me while I laugh
>>62702405So there are a few outcomes. The SEC whistleblower bounty is a given. Realistically a hefty sum in the current climate and its outright disdain for slanted AI. Then there are the settlement options that I’d estimate are in the 10s of millions prior to discovery for them to just piss off, then assuming the novel theory made it past a judge (it would because there is no established case law specific to this thus it cannot be thrown out) we’re talking between 100-300 million easily based on the timing of the IPO and the volatility surrounding it (the underwriters would be listed as codependents so we’re talking Morgan Stanley, Citi, Goldman Sachs, and Chase) would strongly push for a settlement, and trial would easily break a half billion and set legal precedent. >>62702424On account of consumer protection, election tampering, and blatant fraud yes. Everyone has a reason to hate them and they’ve given us the rope to hang them with. And it’s trannies kek
>>62702517>election tamperingUwot?
