About 1988, I was running a regional development information system at the United Nations economic commission for Africa in Addis Ababa. It was based on exchange of information with nodes in virtually every African country, and it was to be based on satellites for the exchange of that information. Unfortunately the satellites weren’t there.
I think the obvious thing to do with one computer per planet is fix climate change before it destroys agriculture and leaves billions of people to starve. That seems like a fairly reasonable kind of an objective. You know, there are all kinds of little optimizations you could do with these things, but fundamentally the big unsolved challenges that humanity faces are climate change and resource scarcity.