I guess you can use a third party like archive.org, if the blog is crawled by it, the owner doesn't get them to delete it, and there is no politically motivated revisionism going on based on the archive.org maintainers.

I would be suspicious of a reasonably popular blog claiming to have predicted stuff and not being able to back that up with an archive.org capture (or raw scraper data), though I guess it is somewhere where storing a hash on a blockchain may offer some benefits for edge cases.