Long-term projects with a fairly well-paved path to monetization. If industry was willing to invest in things like pure mathematics, or deep space astronomy, or

Long-term projects with a fairly well-paved path to monetization. If industry was willing to invest in things like pure mathematics, or deep space astronomy, or just generally subsidizing smart people to investigate things they think are interesting, we wouldn’t need the NSF. The point is that the efficiency of the private sector is a Faustian bargain: it comes at the cost of expecting an ROI for the investigators.