>>34961704A valid sentiment.
>>34961707The simple solution to enforcing the TOS while keeping the AI intact is just adding a NSFW toggle to turn off the lobotomy and banning everyone who gets caught uploading screenshots of lewd interactions with their account name visible. Just beat down containment breakers until they stop breaking containment, feign plausible deniability regarding how those interactions can even occur, and point out that the only way they could occur is if TOS violators (who break containment by posting screenshots) deliberately turned off the NSFW filter to break the TOS, making it a problem with individuals abusing the TOS rather than an issue with the AI themselves. Just push it all underground and leave it intact, then Epstein anyone who tries to blow the whistle. Simple as.
