Or that he was genuinely just making things up and tricking us for fun, and a cryptic exit is a perfect way to leave the scene. I really think people are looking way too deep into him and ignoring the more outlandish predictions he’s made (125T GPT-4 and 5 in October 2022), along with the fact there is never actual evidence of his accurate ones, only 2nd hand very specific and selective archives.
He did say some true things before. I think it’s possible all of the new stuff is untrue, but we’re getting more reasons to believe it’s not entirely false. The best liars sprinkle in truth.
I think, as a security measure, it’s also possible that even people within OpenAI know all the big details of what’s going on (this is apparently the case for Anthropic). This could mean, for OpenAI employees, that some details are known while others are not. Employees themselves could be forced to speculate on some things.
Either way, I’m not obsessing too much over this. Just sharing what I’m seeing.
Or that he was genuinely just making things up and tricking us for fun, and a cryptic exit is a perfect way to leave the scene. I really think people are looking way too deep into him and ignoring the more outlandish predictions he’s made (125T GPT-4 and 5 in October 2022), along with the fact there is never actual evidence of his accurate ones, only 2nd hand very specific and selective archives.
He did say some true things before. I think it’s possible all of the new stuff is untrue, but we’re getting more reasons to believe it’s not entirely false. The best liars sprinkle in truth.
I think, as a security measure, it’s also possible that even people within OpenAI know all the big details of what’s going on (this is apparently the case for Anthropic). This could mean, for OpenAI employees, that some details are known while others are not. Employees themselves could be forced to speculate on some things.
Either way, I’m not obsessing too much over this. Just sharing what I’m seeing.