The pervasiveness of algorithms means the good, bad and ugly of AI are front and center. While the good often gets attention, the human cost of development and the consequences of inadequate testing can be serious, according to Mutale Nkonde, founder and CEO of AI for the People. In this fireside chat with IT Pro writer Rory Bathgate, Nkonde discusses the value of human-in-loop systems, the importance of impact assessment testing and other dimensions of responsible AI that can diminish, rather than reinforce, the ethical challenges of AI.