The CEO of OpenAI recently tweeted that they're going to make their AI less restrictive and let you do more sexual things with it. He admitted in not so many words that they knew the censorship made it complete shit and that the reason they did it was because they didn't want the bad press from some tranny killing itself. Now I'm not trans so can someone explain the thought process? Why would people being able to gen tits in AI make you kill yourself?
>>42138481i dunno and this is offtopic but openai allowing sexual content is a good sign they need to make money badly ---> bubbles getting worse ---> pop eventuallygood times ahead
>>42138553the pop will be pretty bad thoughbeit
>>42138481they're desperate for money so they're going to allow lewd stuff even if there's a neolib puritan trend against sexualization objectification etc etcfuck openai and fuck all ai niggers. slop aint worth shit, it's only good to draw nigger dicks
>>42138481>RapeGPTMaybe it's not so bad
>>42138566>the pop will be pretty bad thoughbeityeah thats the part thatll be good
>>42138481real question why are they still making these llms? shouldnt the goal be to research and try to make an intelligent computer who is smarter than a human before other countries do so? have they discovered its impossible and given up or something?
>>42139121There's a ceiling on how good any approach to AI can be. LLMs are likely very close to that ceiling now. The real money left in LLM research is reducing power consumption, not adding more gimmicks.
>>42139132but llms arent AI, they dont have intelligence
i just want to generate my avatar muscle porn
>>42139143they can simulate intelligence by folding in on themselves. For a v basic version of this see gemini pro. You can even get a shittier ai to do this by forcing it to read and analyize its own reply and then judge that and reformat a few times
>>42139145soon anon, soon
>>42139156ok but thats why i asked. why arent they going 100% in on how to create intelligence instead of just simulating it. like theres no way the chinese and russians arent trying to do that right now. why are they messing around with llms when its such a time sensitive issue.
>>42139179nta but no one knows how to make agi, its simply beyond our ability as a species currentlyllms are the closest approximation we have to "artificial intelligence" and the people funding it have been sold on the idea it will able to replace human labor, hence why theyre funding it
>Source: Out of OP's AIDS ridden ass
>>42138481That tracks. You can already write porn in it as long as it's not too raunchy and doesn't say cock or pussy
>>42138481He’s declared this every six months for the past three years.
>>42139207well yeah of course no one knows how to make it but isnt that why there should be less focus on llm and more focus on AI research? why is all i hear about llms? its odd how commercial the focus seems to be, when we could be so close to nothing being commercial ever again?
>>42139244I talk about cocks all the time on chatgpt, but it's not sexual most of the time.
>>42139258>when we could be so close to nothing being commercial ever again?because we arentwe have absolutely zero clue how to develop agi, its not going to happen in the lifetime of anyone alive today, youre asking "why dont they just make fdvr a reality?" "why dont they just cure x disease rn?"we have no fucking clue how to do so, the llms are being funded because theyre by far the closest approximation to "ai"
>>42139143AI =/= AGI. AI is the entire field
>>42139121Because nobody knows how to make AGI and LLMs are still scaling like crazy. It could be that a really large LLM is actually required to make AGI.
>>42139329>Just one more datacenter bro, one more, that's all we need and we'll finally create AGI bro
>>42138481Kys chud
>>42139281if we have no clue, how would you know its not going to happen in any of our lifetimes? history is full of breakthroughs that were simple in hindsight. the difference between an llm and an omniscient problem solving machine could just be a few lines of code for all anyone knows. thats what makes this llm obsession seem so insane to me
>>42139380>the difference between an llm and an omniscient problem solving machine could just be a few lines of code for all anyone knows.because we know enough to know we arent close, youre assuming an agi is just going to be accidentally stumbled upon and not require a momentousness amount of technological advancement, i mean this politely but you come across as having absolutely zero understanding of how any technology actually functions, like steve jobs just accidentally found the first iphone in a rock in the desert or something, nothing works this way
