Keywords: Algorithms, Artificial Intelligence, Biased Data, Data Collection, Data Infrastructure, Data Literacy, Data Quality, Unethical AI
Author: Mainak Mazumdar
Publisher: TED
Publication Year: N/A
Summary: In the following talk, Mainak argues that the biased data is what makes unethical artificial intelligence (AI), not the algorithms. Instead of spending time and money on enhancing algorithms, focus should be on obtaining unbiased data that makes ethical AI possible. He proposes 3 areas of improvement for this to happen: 1). Data infrastructure, 2). Data quality, and 3). Data literacy. By improving the data infrastructure that enables inclusive, quality data collection, this could solve a lot of problems regarding ethical AI. He gives examples of the impact of the algorithm on business and society at large when minority populations are overlooked during data collection.