I don’t know much about it. How I understand it is that the responsibility of infringement would then also lie with the distributor and not only the uploader.. This kind of liability makes it very unattractive to host user-generated media. In addition, the implementation of this law would require content distribution companies (Tumblr incl) to implement content recognition technology. Nowadays this shouldn’t really be an issue, I guess. YouTube has been doing exactly this with their ContentID for many years. However, herein lies the issue, I think. If an automated process is going to decide what gets flagged for deletion and what doesn’t, there’ll be false positives such as derivative work, and fair-use (parody, criticism, reportage).
Given the magnitude of online distribution, even minor false-positive rates (like 0.1%) is in the millions of posts per year. If the system can be implemented with a safety net, to deal with the false positives, I guess I can sort-of agree with it.
What I haven’t seen is a peer-reviewed analysis of what this kind of system would cost, and how much its benefit would be. Clearly if the benefit is greater than the costs, then it’s worth it. Law shouldn’t be based on principle, it should be based on hard numbers, in my opinion.. ^^
As for me personally.. all the stuff I post on this blog was made by me, so there would be no difference.
but keep in mind I’m a biomedical scientist and not a law student XD. I can tell you all about how DNA gets transcribed to mRNA, or how to analyse lipids on gas chromatography.. but well, it’s not really for this blog..