interesting, and I hope they fund some outlier, less established forms of AI. For example, we may find that massive simulation yields more practical benefits i

interesting, and I hope they fund some outlier, less established forms of AI. For example, we may find that massive simulation yields more practical benefits in the medium term than stronger pure AI / ML, in some domains. By analogy with research on possibly harmful biosystems, one can extrapolate the need for a set of agreed / self imposed safeguards on certain types of strong AI research - eg. make them read-only, not connected to physical actuators, isolated in a lab - just as you would isolate a potentially dangerous pathogen in a medical lab. OpenAI would be the place to discuss and propose these protocols. A quote from a future sentient AI - "don't you think its a form of racism, that strong AI abide strictly by the three laws of robotics, but humans do not?"