I think the most interesting thing YC could do to promote prosperity in the US is some kind of 'fellowship' for technologists with an interest in helping the impoverished to 'embed' with social service organizations and try to learn how to help them.
I don't think most people who frequent HN, no matter how well intentioned, understand the factors that cause people to be stuck in poverty, myself included. I think you would have to get close to the problem to have really useful ideas. I have ideas, but they are more like questions to be honest.
Some ideas I have thought about in the past:
* An app that helps people learn that they are in a failing school district, and helps them systematically develop a plan to get out of it. I live in St. Louis, and around the time of the This American Life episodes on the Ferguson schools, I spent a lot of time wondering how you end up sending your kids to the worst of the 550 public school districts in the state. The question in this case is: what are the characteristics of people who send their kids to bad schools, and to what extent can they help themselves if given the right tools?
* An app that helps people find others to share resources with (shared living, etc). The idea here is that poverty may be driven by the breakdown of the family, which provides first-line social insurance when working well. Kind of like match.com for poor people with kids looking to team up. The question here is: Is there some subset of improverished people who are very high-functioning, but so limited by childcare and low-wealth that they cannot lift themselves out of poverty. If so, could these people be matched and help each other by sharing resources?
* Trying to create a program that directly funnels kids from underperforming schools into computer programming and IT jobs without going to college. Perhaps to be hired by large corporations who will ultimately spring for night-school to round out their education. Here, I am thinking about a section from the book The Prize, where a bunch of Newark teachers go to insane lengths to try to get one kid with some athletic talent enough tutoring that he can achieve an ACT score of like 14 and qualify for a college scholarship. That strikes me as significantly less scalable than a program that tries to give poor kids a marketable set of skills coming out of high school. It would still be a small set of kids, but maybe larger than the number who have athletic talent (and its not clear athletic scholarships yield good outcomes anyway). The questions here: Could some non-trivial group of impoverished high school kids be funneled directly into employment with the right vocational technical skills and support from major corporations?
With all these kinds of ideas, however, I think you would have to get really close to the people you are trying to serve to have any chance of building something that would help them. And that means 'embedding' with them for a non-trivial period and really trying to figure out what will work.
None of these ideas are like startups, in the sense that I think in the best case scenario they aren't likely to scale very well. I think they are more likely to cherry pick a small number of resourceful/talented people who might otherwise have remained mired in poverty.
That's better than nothing, but my intuition is that when you get into the details of poverty that there are no highly scalable solutions. Everything is going to be messy. These stories have informed my opinions on messiness:
Netflix
Rich Hill
Undefeated
Waiting for Superman
Audio
This American Life on Ferguson Schools
Books
Our Kids (Putnam)
The Prize
Work Hard be Nice
Articles
http://www.tampabay.com/projects/2015/investigations/pinella...