I mean, ChatGPT isn't even innovative for OpenAI's team. In terms of raw language capabilities it's not much more effective as GPT-3, which passed the HN Turing

I mean, ChatGPT isn't even innovative for OpenAI's team. In terms of raw language capabilities it's not much more effective as GPT-3, which passed the HN Turing test almost a year before. Some of the credit goes to the fact that Dall-E 2 got everyone looking at OpenAI (and wrt Dall-E I'm personally not aware of any comparable work done anywhere else), a lot of the credit goes to the fact that OpenAI was willing to put it out there, and the work they did to make ChatGPT much safer really doesn't get acknowledged enough (on release, ChatGPT would refuse to produce potentially hateful/harmful text, and eventually a lot of the workarounds got patched as well). It's probably not fair to characterize the discourse around ChatGPT as treating it as a revolutionary bit of tech, lay people are taking note because 1) people they know are actually using it and 2) the interface really is intuitive - literally anyone knows how to use a chatbot.