Building AI You Can Trust and Respect

Machine learning applications are designed to analyze data and formulate predictions without any overall guidance from us. That doesn’t mean, however, that machine learning is necessarily safe from the effects or influence of our human biases. Far from it.

Captain Michael Kanaan has learned many pragmatic lessons in the enterprise deployment and acceptance of AI-based solutions. Kanaan’s philosophy brings the pressing issues squarely to the table as we march toward a future far different than we ever imagined. He encourages the U.S. Air Force and businesses everywhere to value people as their greatest asset and build solution-oriented, rather than task-focused, workforces.

Kanaan stresses the value of humans in machine learning, democratizing AI to enable anyone to independently ask and receive AI-based wisdom. He stresses that we must gain trust and ethics with AI, avoid bias in AI, and harness the ultimate power of AI for business and beyond — indeed, for all humankind.

One significant AI challenge has been machine learning bias and how to make predictions without guidance from humans. Biases are often reflected in our data, which means that our predictions and analysis can be biased as well. If organizations then take action on these predictions that have underlying biases, then they can perpetuate or sustain inequities. Steps to prevent this are possible with oversight and development and training of algorithms.

“What I think is important now is to talk about AI and provide explanations that we all can understand of an incredible evolution in technology and have a resulting capability that will forever change our information opportunities, our interactions, how well we (the broadest ‘we’) understand the rudiments and real potential of AI, anticipate its implications, and coordinate a course ahead. They’re all imperative matters…Get started — dive in! There are open use cases by DataRobot, it’s a welcoming community and everyone has a voice.”