Drishti ias books Published this article page no 29 Biasedness The algorithms used in artificial intelligence are discrete and in most cases trade secrets. They can be biased for example in the process of selflearning they can absorb and adopt the stereotypes that exist in society or which are transferred to them by developers and make decisions based on them. Accountability If an AI system fails at its assigned task someone should be made responsible for it. e.g. an antiterrorism facial recognition program revoked the drivers license of an innocent man when it confused him for another driver. o Similarly when any AI algorithm takes a decision related to social dimensions it is crucial is taken on transparent parameters. These AI algorithms can and do make decisions that create significant and serious issues in peoples lives. 72 Various steps taken towards regulation of AI OECD Principles on Artificial Intelligence adopted by 42 countries including OECD members and nonmembers. United States AI strategy has so far been led by corporations such as Google and Microsoft who are focusing on selfregulation and rapid technological development. o Google published their own AI principles to help guide ethical development and use of the technology. o Also in January 2020 the US has issued broader set of draft principles for federal agencies treatment of artificial intelligence. European Union The Communication on Artificial Intelligence document which among other issues outlines the need to have an ethical and legal framework is in place Drishti ias books buy.
No comments:
Post a Comment