AI alignment advocates are probably doing more to increase the odds of a bad outcome with their paranoia, hubris, and failing of the finger trap than all capabilities researchers combined. Their way of approaching things makes all the risks they claim to care about worse.