Note: I'm not sure yet if it's encouraged/discouraged to link to PDFs on HN, as I know there is inconsistent browser support for them (although I think progress has been made in recent years). Overall I thought it best to link to the source document rather than any number of the mostly generic and/or completely speculative tech news articles appearing today (unless you find a really good one).
Can someone give a technical reason why we can't have an empty TLD which all sites looked up via DNS would default go to? So if you don't enter a TLD the DNS would assume you mean the empty TLD so it would give you the IP of the machine belonging to just the word.