Notice: You have been identified as a bot, so no internal UID will be assigned to you. If you are a real person messing with your useragent, you should change it back to something normal.
Anonymous A (OP) double-posted this , 3 minutes later[^][v]#1,469,075
"Woah, we have a brand new super intelligent AI and we used 10,000 copies of it and spent millions of dollars to come up with a counter example to show that a fluid simulation method that assumes atoms don’t exists fails in some rare situations and can produce a region of infinite density in a finite amount of time, this changes everything"
Anonymous C joined in and replied with this , 46 minutes later, 16 hours after the original post[^][v]#1,469,206
I’ve had it hung up because it can’t execute a command and keeps trying. Just like other software. It is, however, easier to escape most loops and ask it what went wrong and to abandon the process.
Anonymous D joined in and replied with this , 25 minutes later, 16 hours after the original post[^][v]#1,469,215
I just wish you could interrupt it sometimes while it's thinking. You know when you notice it misinterpreted you or is missing a key piece of info as it spits out its "thinking" stage, and you're like shit it's already spent ages getting to this point so I don't want to fully stop it. It would be great if you could quickly type something for it to consider and tack onto the context or whatever and reprocess. I don't know if that's feasible given how these models actually work but maybe one day they'll implement something like that.