July 6Jul 6 11 hours ago, Dave Moss said:No one here is gonna fall for that.No, lots of dumb fringe conservatives and Joe Rogan libertarians will. Liberals and leftists hate him and will never go that way. Nor will any of the "people in the middle" here. Dems are on life support and Elon is rolling in like I got ya.
July 6Jul 6 33 minutes ago, Diehardfan said:No, lots of dumb fringe conservatives and Joe Rogan libertarians will. Liberals and leftists hate him and will never go that way. Nor will any of the "people in the middle" here. Dems are on life support and Elon is rolling in like I got ya.If it was between Elon and Trump I might go Elon, but I find him inscrutable
July 6Jul 6 5 hours ago, Dave Moss said:If it was between Elon and Trump I might go Elon, but I find him inscrutableRight, but all he is doing is splitting the ticket. Nobody on the left is going sign me up. The people in here who claim to be in the center will vote for Harris 2.0 again like last time. As you said not many people will fall for it, but it saves the Dems from their disarray.
July 6Jul 6 3 minutes ago, Diehardfan said:Right, but all he is doing is splitting the ticket. Nobody on the left is going sign me up. The people in here who claim to be in the center will vote for Harris 2.0 again like last time.I think generally in here you’re getting college educated white males (the majority who grew up in South Jersey or Eastern PA). Yeah, there’s a few exceptions, but that’s why the typical poster is center-right politically. I think it’s become skewed on here because Trumpism appeals to working class folks.
July 7Jul 7 10 hours ago, Dave Moss said:Right???The best part was, everyone who wasn't a Trumper knew it was all a con, everyone apparently except for the super wealthy genius who didn't realize he was being used....
July 7Jul 7 At the end of the day, Trump is nothing more than a carnival barker. Everything he does is just theater. It's why Trumpbots spray their shorts whenever he pulls a stunt like this and the rest of us normies roll our eyes.
July 8Jul 8 NBC NewsGrok, Elon Musk's AI chatbot, seems to get right-wing updateIn some answers, Musk’s chatbot on X answered questions in ways that removed nuance.
July 9Jul 9 All the things they accused Google of doing, Musk himself is doing with grok. What a plot twist.
July 9Jul 9 It's funnier the more you think about it.Picture a dumb intern over at Twitter spitting out his coffee and hitting a comically large red "panic" button that triggers a damage control alarmPicture spending hundreds of millions of dollars to develop your own AI, designed to uphold your ideals and support views that form the foundation of your morality...and then it almost immediately starts praising Hitler
July 9Jul 9 17 minutes ago, Dave Moss said:man, they certainly didn't do a good job in the test environment before releasing grok. 🤣
July 9Jul 9 1 minute ago, Phillyterp85 said:YesThat is......something. Not even sure how you train a model to be this way so clearly.
July 9Jul 9 9 minutes ago, Paul852 said:That is......something. Not even sure how you train a model to be this way so clearly.Best analogy I can think of is similar to a runaway chain reaction in a nuclear explosion. As you probably know, LLMs are largely a function of the scope and quality of data fed to them, and this one is probably trained on datasets from multiple sources, but if someone were to overweight Twitter itself as a dataset (knowing the user base of Twitter these days) it's not a stretch to see how this could happen. Especially when the teams that were in place to monitor and perform QA were all axed as a cost cutting measure.And not to defend Twitter, but Microsoft even had this happen a few years back where they had to shut theirs down for morphing into an overtly racist chatbot almost overnight.
July 9Jul 9 3 minutes ago, we_gotta_believe said:Best analogy I can think of is similar to a runaway chain reaction in a nuclear explosion. As you probably know, LLMs are largely a function of the scope and quality of data fed to them, and this one is probably trained on datasets from multiple sources, but if someone were to overweight Twitter itself as a dataset (knowing the user base of Twitter these days) it's not a stretch to see how this could happen. Especially when the teams that were in place to monitor and perform QA were all axed as a cost cutting measure.And not to defend Twitter, but Microsoft even had this happen a few years back where they had to shut theirs down for morphing into an overtly racist chatbot almost overnight.Fair. I just feel like you have to actually try to make it this bad Did they even test it before deployment?
July 9Jul 9 Just now, Paul852 said:Fair. I just feel like you have to actually try to make it this bad Did they even test it before deployment?Testing is for companies who care about their customers. Plus they were probably a bunch of libs so Musk sent them packing the morning he assumed control. He's a high tech super genius, how dare you question his expertise?
July 9Jul 9 4 minutes ago, Paul852 said:Fair. I just feel like you have to actually try to make it this bad Did they even test it before deployment?We are the testers
Create an account or sign in to comment