Not sure how this would be different from Bing Copilot. I use copilot a bit and it's great for some general queries about current events or things that are written in many articles. For instance, how many children does Drake have or what's up with Eurovision (referenced in dozens of spammy articles).

But I noticed it overly relies on search. If I ask a technical question, it relies on those same crappy search results, compared to gpt which answers correctly without referencing anything external.

I don't know if you can bridge that divide but I'm curious what openai comes up with and how its different from bing copilot