Nice article – I liked the recap of XML and XHTML, and I agree that blockchain is useful and interesting from a CS perspective, but a bad fit for most applications.
He points out that there are cheaper ways to reach consensus for many applications, and you don’t necessarily want to publish every transaction to the entire world.
James Mickens is saying some of the same things here:
Despite the meme-y title, it’s somewhat constructive: can we design a blockchain that fits more applications? I think there is a continuum of how “adversarial” transactions are in various applications, and assuming the worst is a tradeoff that has consequences (i.e. it makes usability worse).