I think that there is difference between idea of Open AI as it was suggested by Musk in the beginning and actual work in the organisation named “Open AI”. The latter seems to be more balanced.
Public understanding by reading a few blog posts might not give a good overview over the reasons for which Open AI was started. I think looking at the actual actions might be a better way to try to understand what the project is supposed to do.
I read that you joined Open AI and I think it is good project now, but the idea of “openness of AI” was fairly criticised by Bostrom in his new article. But it seems that the organisation named “OpenAI” will do much more than promote openness. There is a little confusion between the name of organisation and the idea of letting everybody to run their own AI code.
I joked that the same way we could create Open Nuke project which will deliver reactors to every household which would probably result in in very balanced world where every household could annihilate any other household, and so everybody is very polite and crime is almost extinct.
I have no affiliations with Open AI. In this case I’m driven by “Don’t judge a book by it’s cover”-motivations. Especially in high stakes situations.
But it seems that the organisation named “OpenAI” will do much more than promote openness. There is a little confusion between the name of organisation and the idea of letting everybody to run their own AI code.
I think taking the name of an organisation as ultimate authority of what the organisation is about is a bit near-sighted.
Making good strategic decisions is complicated. It requires looking at where a move is likely to lead in the future.
Public understanding by reading a few blog posts might not give a good overview over the reasons for which Open AI was started. I think looking at the actual actions might be a better way to try to understand what the project is supposed to do.
I read that you joined Open AI and I think it is good project now, but the idea of “openness of AI” was fairly criticised by Bostrom in his new article. But it seems that the organisation named “OpenAI” will do much more than promote openness. There is a little confusion between the name of organisation and the idea of letting everybody to run their own AI code.
I joked that the same way we could create Open Nuke project which will deliver reactors to every household which would probably result in in very balanced world where every household could annihilate any other household, and so everybody is very polite and crime is almost extinct.
I have no affiliations with Open AI. In this case I’m driven by “Don’t judge a book by it’s cover”-motivations. Especially in high stakes situations.
I think taking the name of an organisation as ultimate authority of what the organisation is about is a bit near-sighted.
Making good strategic decisions is complicated. It requires looking at where a move is likely to lead in the future.