I think the key takeaway is that these aren't new or different risks. You could say the same thing about iPhone apps, pharmaceuticals, or traditionally bred plants: as you consider the whole lines of $ITEMS that will be created over the next decades at an accelerating pace, the odds that at least one of them turn out to be catastrophically bad enough to entirely erase the benefits of everything else done with them goes up quite distressingly.

Thalidomide happened. But it doesn't justify abandoning pharmaceuticals entirely. There just might be some wisdom in that.