Everest Analytics origanized a free webinar on Ethics in AI.

Martin Canter from Omina Technologies was happy to contribute with a talk on ”How is ethical AI related to law, ethics, and explainable AI?”

*There are many group-fairness metrics. Optimizing for one group-fairness metric is better than the status quo.

#Diversityy, #non-discrimination and #fairness is only 1 of the 7 principles of #trustworthyai as described in #ALTAI

*Choose your #ethical perspective on #fairness first (fairness is a cultural and contextual concept), then search the mathematical group-fairness metric fitting the fairness concept best

*Distributive fairness: fairness of distribution of outcomes: cf. disparate impact and indirect discrimination

*Procedural fairness of the process that leads to outcomes: cf. disparate treatment and direct discrimination. Related to #interpretable AI, #explainableai and #transparancy. #ominatechnologies #aigovernance #ethicalai