This sounds much more involved than GiveDirectly's simple cash drops. It would be nice to see a comparative analysis of the two. I'm also curious how this scal

This sounds much more involved than GiveDirectly's simple cash drops. It would be nice to see a comparative analysis of the two. I'm also curious how this scales. Give one family in an area money/support and they can escape extreme poverty. If you give the same boost to every family in an area, do you get a bigger impact, or smaller? And is there a point where having n% of the people in an area boosting out of poverty leads to the other (100-n)% being _worse_ off? Or are they all better off? At any rate, this is really exciting. Both that they did RCTs to test it, and that it worked really well.