PDF Publication Title:
Text from PDF Page: 027
CHAPTER THREE The Problem in Focus: Factors and Remedies The examples in the previous chapters illustrate a number of angles on the algorithmic bias problem. The first and most basic angle is the problem of an algorithm’s data diet: With limited human direction, an artificial agent is only as good as the data it learns from. Automated learning on inherently biased data leads to biased results. The agent’s algorithms try to extract patterns from data with limited human input during the act of extraction. The limited human direction makes a case for the objectivity of the process. But data generation is often a social phenomenon (e.g., social media interactions, online political discourse) inflected with human biases. Applying procedurally correct algo- rithms to biased data is a good way to teach artificial agents to imitate whatever bias the data contains. For example, recent research shows that automated methods applied to language necessarily learn human biases inherent in our use of language (Caliskan-Islam, Bryson, and Narayanan, 2016). This leads to the rather paradoxical effect that artificial agents, learning autonomously from human-derived data, will often learn human biases—both good and bad. We could call this the paradox of artificial agency. The Watson and Tay examples illustrate the point well. Sweeney (2013) also gives multiple examples of targeted adver- tising systems making biased and sometimes defamatory inferences about particular individuals because of biases automatically learned from data. This paradox has important implications for the use of arti- ficial agents in the big data era. The complexity of data patterns and the sheer scale of available data make it necessary for artificial agents to 17PDF Image | Intelligence in Our Image Risks of Bias and Errors in Artificial Intelligence
PDF Search Title:
Intelligence in Our Image Risks of Bias and Errors in Artificial IntelligenceOriginal File Name Searched:
RAND_RR1744.pdfDIY PDF Search: Google It | Yahoo | Bing
Cruise Ship Reviews | Luxury Resort | Jet | Yacht | and Travel Tech More Info
Cruising Review Topics and Articles More Info
Software based on Filemaker for the travel industry More Info
The Burgenstock Resort: Reviews on CruisingReview website... More Info
Resort Reviews: World Class resorts... More Info
The Riffelalp Resort: Reviews on CruisingReview website... More Info
CONTACT TEL: 608-238-6001 Email: greg@cruisingreview.com (Standard Web Page)