Speaker "Frida Polli" Details Back
-
Name
Frida Polli
-
Company
Pymetrics
-
Designation
CEO
Topic
Keeping AI Accountable and Bias-Free
Abstract
As the use of AI grows across industries so does the need to develop processes to ensure AI benefits everyone, not the select few. As the use of AI continues to grow, we have to monitor them to ensure it’s development continues to align with our objectives. • Management: Key point of contact in the event that is something goes wrong this person must have the authority to manage any negative fallout from an algorithmic system. How often should developers check in on their AI? • AI Auditing: Historical data is inaccurate to rely on; this way of relating perpetuates bias, especially if bias is already inherent within the dataset. • Explainable AI: How every decision / error needs to be understood in order mitigate similar issues in the future.