Exactly this. I don't understand why this is being thrown about like it's a bad thing; I gladly pinned all of Wikipedia, but it's also a huge amount of data and I made sure I was prepared to follow through before I asked my IPFS node to do it. If there's content out there that has earned my respect, I'll pin it gladly just to help the authors out, especially if they ask me politely.
To be fair, some of this is automated. If you browse using your own node, it does maintain a small cache, which helps with the load distribution when something gets really popular. This blog post is a good example; despite literally being the #1 item on HN, it's still up; that's the global IPFS cache at work. It's slower than the average site, sure, but it didn't outright vanish, even though the author clearly is using minimal resources to actually host it. I think that's awesome.
To be fair, some of this is automated. If you browse using your own node, it does maintain a small cache, which helps with the load distribution when something gets really popular. This blog post is a good example; despite literally being the #1 item on HN, it's still up; that's the global IPFS cache at work. It's slower than the average site, sure, but it didn't outright vanish, even though the author clearly is using minimal resources to actually host it. I think that's awesome.