If you did get an AI which could convert using pattern-matching between two arbitrary formats, given samples of, eg: songs, in each format… does it still count as “reverse engineering” of any proprietary formats that happen to get passed to it?
Even if the AI is the thing being sold, not the actual conversion algorithm, and there’s no law being broken when it’s converting between open standards?
And who would you sue? The developers of the AI, the person who used the AI to do the reverse engineering or the AI itself?
It’s only theoretical, I know, but all things may come to pass…
“How to resolve difficult court cases:
Create an AI with n randomly-generated conflicting artificial personalities which will interact to arrive at a verdict. n should be a sufficiently high number for a reasonable chance of a fair trial, eg: 12.”
In other random thoughts, I find it amusing that Apple has asked XTunes to change their name. What, the word “Tunes” is branded now? I would have thought they’d worry about the “i”. I’m not so sure that changing XTunes to “Sumitunes” is going to help matters, either…
Even if the AI is the thing being sold, not the actual conversion algorithm, and there’s no law being broken when it’s converting between open standards?
And who would you sue? The developers of the AI, the person who used the AI to do the reverse engineering or the AI itself?
Replace AI with P2P software, and your conversion with distribution and you’ll see that by analogy it’s obvious who will get sued. [1][2] 🙂
As for why they’re worried about ‘Tunes’ and not the ‘i’, easy group have already proven that you have no legal rights to a prefix…
[1] Ignoring the fact that proof by analogy is dubious even when we aren’t talking about law.
[2] Both
What’s the difference between an AI and an algorithm?
You can create an algorithm to convert one thing to another. An AI which can learn doesn’t start off knowing the algorithm; it just has the capacity to learn it, or to work it out.