binary mind by priyanka from the Noun Project world by Isaque Pereira Cordeiro from the Noun Project binary code by Prettycons from the Noun Project

By Diedre Brown

As we have discussed throughout this site, machine learning has profoundly affected all arenas of human life. However, it is important to remember that we as humans also play a role in the development of machine learning. Humans are responsible for:

• Overseeing the collection of data.
• Writing code.
• Assessing the precision, accuracy, and effectiveness of algorithms.
• Tagging and creating training datasets.

And without human oversight and engagement, the path of development of machine learning and the overarching goals of artificial intelligence will always remain elusive.

While data and machine learning algorithms on their own are not bias, we as human beings are. Through life experience, and developing a preference or an inclination towards or against something, bias develops in every human being. Though sometimes bias can be unreasoned, that is not to say that it permanent. As writer Adam Gopnik stated in The New Yorker, “Perhaps our intelligence is not just ended by our mortality; to a degree, it is our mortality.” And we would further argue that it our duty to ensure that our individual uniqueness is always accounted for in the use of algorithms, analysis of data, and the development of artificial intelligence.

To learn more about what is being done to prevent biased AI, and how you can contribute, we recommend the following links and articles for further reading.

Partnership on AI
Governance of AI at the Future of Humanity Institute & University of Oxford
'THE DOOMSDAY INVENTION', by Raffi Khatchadourian, New Yorker, November 23, 2015 Issue
'What is AI bias?', by Cassie Kozyrkov, Medium, Jan 24, 2019
'Tackling bias in artificial intelligence (and in humans)' by McKinsey Global Institute, June 2019
'How to Reduce Bias in AI' by Leena Murgai, Medium, July 25, 2019