I hate this fucking thing to be honest. It's a gimmick, a very dangerous one. They ask in the EULA to not post the output of the AI to social media, but damn i really want people to witness this.
i know diminishing returns are a thing, but GPT-3 looks like a qualitative leap over GPT-2 and its derivatives. GPT-4 will iterate on the mistakes this makes, and get better. i really wonder what the long term solution is
I punched in Heidegger before. It was really fucking something. People will be writing their dissertations this way.
I chose Debord because if he knew we were feeding his words to what is essentially the final form of the spectacle, he'd probably get really upset.
I hate this fucking thing to be honest. It's a gimmick, a very dangerous one. They ask in the EULA to not post the output of the AI to social media, but damn i really want people to witness this.
You absolutely should be sharing this. People need to know the ticks that it has so it's easier to detect.
i know diminishing returns are a thing, but GPT-3 looks like a qualitative leap over GPT-2 and its derivatives. GPT-4 will iterate on the mistakes this makes, and get better. i really wonder what the long term solution is
We'll know soon enough, China just finished one that's 10x as big