No one knows what really happens at OpenAI rn. Over the weekend there was a series of rumors, and I couldn’t find any sources for them, so I consider them null and void. ChatGPT also has no idea:
But since I was talking with ChatGPT I took the occasion to see what it will say about processing copyrighted data, like websites, books, without the owners permission and compensation.
Because OpenAI and other companies think it’s ok to assume they can process copyrighted material without owner permission. To withdraw the assumed permission you have to explicity tell it with a tag.
Probably the lack of legislation in that matter is the cause AI developing companies push for legislising it. But that’s my guess. These companies of course act according to the law and have great intentions.
So what ChatGPT thinks about it?
ChatGPT argues that it’s not ok to mix music by a DJ without paying to the music authors. But mixing data is complicated…
Some would argue that a DJ and AI are two different things. Because AI is built in a way attributing original authors is hard if not impossible.
Let’s see what ChatGPT thinks about it:
ChatGPT basically tells that if you build a complex system of processing data that loses the information about authors or copyrighted material it doesn’t mean you can use it as an excuse to not respect copyrights and not compensate authors.
It’s an interesting view, because AI is a complex system that actually loses the information about the origin.
What’s more interesting, it’s not caused by the deficits of the technology. Technically it’s possible to…