Select Page

Yesterday I delivered my talk to an engaged audience at Ethicomp 2018 in Sopot, Poland. I spoke on widening the debate on transparency in algorithms to cover more than the inner workings of algorithms. There should also be transparency with regard to the possible inputs these algorithms (or machines powered by such algorithms) can take, the data used to train the machine, the function of the machine, the possible outputs of the machine, and the machine’s physical and/or virtual boundaries. I argued that only with this information can we make informed decisions as regulators, users, and bystanders to such AI driven machines.