While it’s probably a very unpopular opinion here, I personally don’t think that open-source works for everything. In this case OpenAI had the advantage of moving first. I believe that brings with it some extra considerations.

The technology surfaced very quickly. There were obviously other companies working on this kind of tech behind closed doors, but OpenAI was the first to make it publicly available and accessible. Personally, I’m not sure if I would like to be responsible for non-friendly actors quickly cloning the code and removing all the safety stops. I would say for most of the public AI really wasn’t on their radar. Then one day that’s what everyone is talking about. There was space then, while nobody knew what hit them, for bad actors to use the technology for some really bad stuff. I do think there would be space for open-sourced models now (which is what we are seeing) but because they were “the first” I am fine with how they did it. There will still be competition. Other companies are being forced to come out with similar products which I believe is better for everyone. Afterall, I don’t think we can treat this like some random program. This is something very different and it should be treated a bit more carefully in my opinion. I think “being careful” in this instance would be to keep it closed until people get used to the idea.