Let Aiden help you "Keep Pace" with AI Bias

microphones-1491019_640.png

Topic Soundbyte

Making sure AI decisions are impartial is critical

  • AI learns by processing lots of data, and that data may be biased

  • AI has no concept of how to identify the bias, it is up to humans to review and identify the bias 

success-1093889_640.jpg

Latest Trends

AI is already being leveraged today for this.   Key trends to watch:

 

IBM announces AI Fairness Project

IBM open sources toolkit to check for bias in datasets

 

Amazon scaps AI recruiting tool that showed bias against women

A level of consistency humans can't match 

MIT detects "Fake News"

Developed a system that can determine whether a source is accurate or politically prejudiced

Aiden Suggests

Recommendations for learning more about AI Bias

Suggestion #1:  Read the World Economic Forum - Bias in AI article

A good summary overview of the problem space, and the steps to take to avoid bias in AI systems.   Click here to "Keep Pace"

Suggestion #2:  Experiment with the IBM AI Fairness Toolkit 

Have your team apply its datasets against the AI Fairness tool.   Click here to "Keep Pace"

Suggestion #3:  Read up on Industry "Best Practices" 

Google's "Responsible AI Practices page.   Click here to "Keep Pace"

Microsoft's AI Principles page.   Click here to "Keep Pace"

Suggestion #4:  Take Google's "AI Fairness" Online Class (FREE !)

An excellent overview of AI fairness including Video Lectures.   Click here to "Keep Pace"