That's how it works in the material world. Data works differently. Data is infinite. You can have as much as you want. Why would anyone buy copies when making new ones is as easy as copy and paste? Only the original copy must be paid for.
So there is in fact a third choice to buying or not buying data: obtain a copy from someone who already has it. Copyright law is all about getting rid of that choice. It's all about making sure that creators monopolize the means to make copies. Copyright is about the right to make copies.
Making copies of data is a fundamental computer operation. We can find this ability in functions such as memcpy, in programs such as cp and HTTP servers. Making copies is trivial. Anyone who owns a computer can do it. Without compromising our freedom to run these programs, copyright law is unenforceable.
So it's not just ideals. In order to actually enforce copyright law, computers must be made to answer to the copyright industry instead of users. As a computer user and programmer, that's not the future I want.