The Search For Ethical AI

A moral framework is needed to help us navigate a future shared with machines

- Mic Mann

“Our technology, our machines, is part of our humanity. We created them to extend ourselves, and that is what is unique about human beings.” – Ray Kurzweil

In his non-fiction book, The Singularity is Near: When Humans Transcend Biology, Ray Kurzweil sets the date for the Technological Singularity as 2045. He postulates that the non-biological intelligence created in that year will be one billion times more powerful than all human intelligence today.

While still in its infancy, it is hard to ignore the steady pace at which artificial intelligence (AI) is becoming more integrated into our lives, and it is raising some valid concerns. AI holds the promise of making our lives easier, of enabling businesses to run more efficiently, and hastening scientific and medical breakthroughs. There is, however, a darker and more foreboding side to its integration. Many of the questions being asked are focused on not only the ethics of its implementation, but the moral fortitude of the technology itself.

Is it right that humans might lose certain jobs? Can AI be trusted with jobs that involve the welfare of humans? Who takes responsibility if AI goes wrong? Because we aren’t sure of the answers to these questions right now, perhaps the most important one to be asking is: are we doing all we can to ensure that AI enhances the human race, rather than destroying it, as many great minds have theorised?

Can we trust AI?

Trust is earned. Because it’s still early days, and because we’re still trying to understand how best to train AI, we have little tangible proof that we can. What we have learned to date is that less-advanced AI can, and does, make mistakes. We know that it is difficult to train, that it struggles when faced with ambiguity or uncertainty, and that this can have serious consequences.