I think it's somewhat inevitable that about 1 million people or more suddently die one day or one week or one month from AI. But:

If and when this happens it will change how everyone deals with AI ever after that.

In Fact, if no one dies from AI this century then it would be much more likely that we all might die from AI.

It's sort of like the Lobster in hot water or cold water story.

If you put a lobster in boiling hot water he will jump out of the pot.

But, if you put him in cold water and turn on the burner there will be no one point where he will jump out before he dies. It's sort of like when people get heat stroke. Unless someone is there to save them they die.

So, the point is that if we DON'T have one million or more people die one day or one month obviously from AI then we might not be able to "Jump out of the pot" and save the human race.

This is what I really think at this point.