for - progress trap - superintelligence threat

progress trap - superintelligence threat - super intelligence is going to be far beyond our cognitive capabilities across many domains. For example, - it's going to be able to find exploits in human code too subtle for humans to notice - it's going to be able to generate code too complicated for any human to understand - even if the model spent decades trying to explain it - How do we entrust ourselves to a superintelligence that is so far beyond us? If it thinks we are expendable, it could easily find our weaknesses and bring about extinction