On Connecting People

Being one of the least connected people, this is one of my abiding interests. I am not talking here about how to identify people with common interests,. That is not so interesting to me, because it tends to group people into clubs where they will meet others who are like them, in some respects, and therefore it tends to fragment societies, because although such common interests often unite people, there are other respects in which their interests and values are not common, but opposed, so these clubs of people tend to fracture.  The interesting problem is how people with mutually complementary interests can find each other, even if they didn't know that their interests were mutually complementary before they met and spoke to each other. This process has applications to education and to business, so it is more than just a social tool for individuals.

My idea is that we use algorithms, which we ourselves design, and let those algorithms suggest possibilities, from which we choose, according to how we feel about the reasons the algorithms give us for why they consider these to be potentially interesting directions that we could explore.

So to do this we need software which allows us to combine in the most general way, techniques such as any two or more of the following, and probably dozens of others, and use these to analyse our data as we work, and suggest potential directions to explore.

The aim is to be able to do large-scale collaborations with a rational scientific basis (see Julia Galef on Aumann's Agreement Theorem) on which to prioritise them.