If you do that, and you separate control and use, then you can imagine that there’s going to be lots of different data cooperatives or collectives which can accept or reject algorithms. You can imagine a plurality of algorithms on top of a plurality of these collectives which we choose from.