Explainability in AI & Putting Humans in the Loop w/ Renée Cummings & Niki Athanasiadou

How do you appeal a blackbox algorithm? How do we balance equity and efficiency? What is explainable AI and why is it essential to ensure fairness, accountability, and transparency?

In this episode of IBM thinkLeaders podcast, we are joined by guests Renée Cummings (AI criminologist & principal consultant at Urban AI ) & Niki Athanasiadou (data scientist at H2O.ai & principal data scientist and owner of Common Sense Analytics, LLC). We talk to Renée and Niki about ensuring that "AI is not just speaking to AI," increasing education around AI and the diversity of talent in the space, and the importance of considering fairness, transparency, and explainability when building models.

Connect with us @IBMthinkLeaders + the guests at:
@CummingsRenee
@RodonikiA

“if we want to make algorithms or make AI as human as possible, then we've got to bring those other human aspects that make us who we are as a democracy in a society. So it comes back to your rights.” -Renée Cummings (AI criminologist & principal consultant at Urban AI

"What we have is a lot of fear, a lot of misunderstanding. And at the end of the day, I think AI brings us in front of what humanity needs to do, who need to be responsible. We need to grow up and we need to start thinking seriously about how we want to live in what societies we want to live. " -Niki Athanasiadou, data scientist at H2O.ai & principal data scientist and owner of Common Sense Analytics, LLC