Reid Hoffman makes a fascinating point about AI safety – the real danger isn't the technology itself, but who gets to control it. If we lock down AI development too early, we might hand all the power to a small group of insiders while keeping it away from people who could use it for good....