Arguably OLPC's approach has - in hindsight - been discredited, but yeah, it seems overly cruel to lump it in with Juicero or Theranos. We do have impossibly c

Arguably OLPC's approach has - in hindsight - been discredited, but yeah, it seems overly cruel to lump it in with Juicero or Theranos. We do have impossibly cheap computers now that are far more accessible, that do nearly all the things normally expected of computers, and with widely available resources for programming on them - RPi, Chromebooks, etc. The success of those devices relative to OLPC (with of course, the benefit of hindsight) seems to come down to a few major differences in the approach: - insistence on highly-custom everything, rather than exploiting as much as possible off-the-shelf to keep costs down. - insistence on targeting the absolute lowest-cost end of the market initially rather than starting at a higher price point and using that as a starting point for iterative cost-reduction. The latter pitfall one still sees various hardware startups fall into with some regularity. Universal accessibility is a fantastic goal, but often the easiest way to do it is to sell something considerably more expensive first.