It seems to come down to whether you think the game is fundamentally zero sum and/or whether you think intelligence is primarily linked to or used for aggressio

It seems to come down to whether you think the game is fundamentally zero sum and/or whether you think intelligence is primarily linked to or used for aggression and domination. If intelligence leads to better cooperation and benevolence or if the game is positive sum, autonomous AI is not likely to be a huge risk. If we live in a zero sum eat or be eaten universe, it is. But then again so are we to each other. I am of the strong opinion that humans are the greatest threat to humanity. Sentient autonomous AI is much less dangerous to us than our own kind using AI as a force multiplier. The reason is that whether or not today’s post-technological intelligence explosion world is zero sum, humans are hard wired to behave as if it is. We would attempt to dominate and enslave one another in paradise because our brain stems think they are clawing for survival on the African plains hundreds of thousands of years ago. You can see it in how the rich and powerful often behave. They have incredible abundance but their brain stems still think they are fighting for subsistence. Recent example: Vladimir Putin has billions of dollars and an entire large modern nation at his disposal and the biggest idea he can come up with is to launch a war. His brain stem still thinks starvation is near and we better raid the neighboring tribe before the dry season. The harsh world we came from made us what we are, but the legacy of instincts it gave us might kill us.