An Ethical Machine

Imagine a time not long from now, when your ethical AI app, with transparently explainable models, trained on unbiased data and protecting your customer's privacy, distinguishes your brand as human centric and trustworthy, and provides a competitive advantage for you with strong customer preference and loyalty. 

You have arrived - through creating an infrastructure of trust - by embedding your ethical requirements from the very beginning of design, through development, quality testing and deployment. 

This may sound utopian, but dystopian life is only interesting as speculative fiction.  Actually, speculative fiction (science fiction) serves the purpose of identifying future scenarios that we would want to avoid and explores ways to prevent these dystopian realities as an entertaining fictional thought experiment. One dystopia we can fairly easily avoid is 'failed-projectopia', which often comes with AI when it outputs biased decisions, when it leaks privacy information, or it provides otherwise harmful recommendations. Including ethical values (harm mitigation) in AI design from the outset will lead to a better success rate in AI projects. 

Let us then imagine the risks and opportunities AI will bring, and experiment with ideas to prevent dystopia.