It seems dishonest to claim to "invent" something that already exists, and is as trivially simple as "ignore errors". Haskell already has this for example in GHC: -fdefer-type-errors. Notice how it is only used for ignoring errors in known broken code, not for actually trying to leave out type checking permanently. That's because the premise that static types get in the way and it is better to not have them sometimes is false.