It's time to discuss ethics in this AI era
WE have made more progress in artificial intelligence (AI) in the last three years than in the preceding three decades. AI is transforming from handy little applications that make our lives easier (from Alexa and Siri to Uber and Netflix) to something more powerful that is reshaping the commercial world as we know it.
The founding editor of Wired, Kevin Kelly, has likened AI to electricity - a cheap, reliable, industrial-grade, digitally smart entity that is the invisible force keeping everything running. He adds that it "will enliven inert objects, much as electricity did more than a century ago. Everything that we formerly electrified, we will now cognitise".
But our growing dependency on AI highlights serious issues of accountability and respect for human sensitivities. Here are some of the key social and ethical considerations to think about:
How do we know that an AI algorithm written by a human will not be biased? We have already seen how some machine-learning systems can fall foul of gender and other biases.
When it comes to the actions taken by an AI system, how do we decide what is ethical or not? For instance, a Netherlands developer had to defend himself - and his Twitterbot - when the police came to his door, inquiring about a death threat sent from his Twitter account.
Who is the guilty party when a self-driving car crashes? Can the owner of the vehicle file a claim against the manufacturer? What impact will this have on the insurance industry?
The commercial questions surrounding this subject are potentially endless, and certainly far too numerous to be answered individually.
Instead, we must frame them as questions of ethics. As AI tightens its grip over every part of our lives, the ethical challenges it brings will only become more pressing.
The problem is that ethics is a highly nebulous area. AI needs governance but right now, no central body is overseeing this task.
So how do you hang on to ethics when machines can do everything? While there are no universal rules that everyone can follow, there certainly are some important lessons that can help:
Ethics is probably the last thing that innovators think about, when it should actually be an integral part of a firm's innovation strategy. The innovation team must ask themselves, before prototyping any new idea, what its social and societal implications are. Most of the time, if not always, the team should be certain that they are building "good" AI.
A good starting point would be to analyse how five of the world's largest tech companies are creating a standard of ethics to govern AI. An ideal scenario would be the replication of something similar within your own industry.
LISTEN TO YOUR CUSTOMERS
You should also actively encourage feedback from your customers so that you can be more in tune with what they are feeling.
One particularly interesting development is the Moral Machine Platform, which is designed to crowdsource expectations of how an autonomous vehicle should make moral decisions. Being serious about addressing ethical issues and incorporating the findings into the fabric of AI is crucial in this exciting new era.
If ethical considerations are overlooked, AI could have a catastrophic effect on companies' brands, reputations and finances. There is a clear requirement for more open and honest conversations about the ethical implications of AI and how we can best prepare ourselves for the exciting times ahead.
