How Can We Eliminate Bias In Our Algorithms?

It's almost comical how surprised we are at the pitfalls of artificial intelligence (AI). After all, we've been making movies for decades warning against the dangerous potential of sentient machines. And yet, the minute Facebook perpetuates foreign interference in a superpower's election or a Twitter bot becomes a marijuana-loving Nazi, we're shocked. And the reality is that our biases (political, racial and gendered) show in the data that we feed to our AI algorithms. As the COO of an AI-powered company that serves clients who also develop AI-powered products, I've come across the potential pitfalls of biased algorithms numerous times.