After reading this whole article, I find it really scary that something like this can get so much attention. It looks like the author read about PoS circa 2014

After reading this whole article, I find it really scary that something like this can get so much attention. It looks like the author read about PoS circa 2014 and hasn't read anything written or done since then. It's true that the "nothing at stake" problem exists, but there are tons of practical solutions and mitigations that work, many of which are already deployed and protecting >$100M. Soon ETH will be securing trillions with such mitigations. To address the specific points the author makes: 1. If a node signs another version of the same block within a reasonably short time period, “slash” their deposits (e.g. punish them inside of the system) You don't have to know which came first, just like in BTC. You just need a longest chain rule with the property that the longest chain is final after a certain point (subject to certain assumptions about the % of stake that is honest). This is how nearly every blockchain works and it's not special in proof of stake. 2. If a node signs another version of the same block, like, a year later, just ignore it. Yes, that's fine. Lots of chains do this. It's called a "finality mechanism". Even ETC has one called MESS while still using proof of work (although MESS is probably broken). Bitcoin could add one too. This is orthogonal to PoS vs PoW.