Example of Imperfect Big Data
00:43 - 02:13
1m 30s
Explains how big data can have unintended consqequences. The example used illustrates how big data used the presence of snow to categorize huskies vs. wolves, even though this is not what the researchers wanted the algorithm to learn.

Comments

Please sign in to write a comment.
Video Transcript

Related Clips

After Hooli Con, Hooli phones started to explode whether because of Pied Piper's code or Keenan's Vr code that was to function on a $10,000 system they don't know. Jared feels that they should tell someone that their code is hidden on Hooli phones but Richard doesn't feel like they should. As a result Jared resigns from the company.
Explains how k-anonymity can be used to conceal people's identities in a dataset, and presents the case of the Golden State Killer's identification via genetic big data to draw attention to the issue of consent in data collection.
Discusses the usage of genetic data by law enforcement and the medical industry, and describes how genetic data can be used to find violent criminals.
Illustrates how organizations like Planned Parenthood and the Trump campaign used big data to find specific populations to target with political advertisements and messages.
Provides real life examples of how biased algorithms in big data can have serious effects; for example, how the "Compas" algorithm is used by courts to determine recidivism rates despite the possibility that the algorithm is racially discriminatory.