Web search engines claim fair use for snippets, and MS claims fair use for Copilot. However, a web search engine result page is only an aggregate of the excerpted pages, and it refers back to them; unlike software authored using Copilot, it does not draw those snippets together into a coherent purposeful whole, strip all reference to the original author. and claim complete originality.
That MS went ahead with training Copilot on open source code authored by people who are clearly not OK with it is why they are so short on goodwill.
Why is Copilot any different from a human doing essentially the same thing? Complete originality doesn't exist. I think the people who are complaining about Copilot have likely done the same thing themselves. I can't erase the memory of all the code I have read and written that somebody else 'own' the copyright to, so my brain will subconsciously use copyrighted code when writing software.
Because copilot is not a human and can do things that humans can't. Moreover I encourage you to learn music pieces by listening to them and then selling/giving away snippets. You will have big corporations on your heels fairly quickly even though you are a human.
That MS went ahead with training Copilot on open source code authored by people who are clearly not OK with it is why they are so short on goodwill.