How much can we really trust a machine?
The concept of artificial intelligence has been in our world for several years now. In fact, it’s been two decades since the first AI computer beat humans at a chess game. Now, we’ve arrived in 2018 and the world of AI is at a whole new level, with some experts predicting that singularity will soon be on the horizon.
It’s impossible to ignore the potential of machine learning in the right spaces. When properly utilised, AI can transform the world, making it easier to process huge amounts of data and make important decisions. The question is, how far should those decisions be allowed to go? Can AI truly be trusted with choices around life and death?
AI in The Automobile Industry
You might be wondering, “where would AI get the opportunity to make such a huge decision anyway?” Let’s start by looking at the autonomous car sector – a space that’s hoping to replace the mistakes of human drivers with technology-powered cars that could save countless lives on the road. Every day, the industry produces more data to help machines predict traffic patterns, safer routes, and potential hazards.
The question is, what happens if a person steps in front of an autonomous car on the road and the car needs to make the choice between hitting the pedestrian, or potentially swerving and injuring the driver? In a split-second decision, could an AI system make the right choice?
California has already begun to hold conversations around legislation that requires humans to take control over a vehicle within dangerous situations. In other words, while AI might be fine for crunching numbers and giving us the resources we need to make complex decisions – it’s not ready to make the choices for us.
Trusting AI in the Healthcare Space
Another example of where AI may be required to make complex decisions is in the healthcare industry. Indeed, AI in this sector is expected to see a growth of 40% between 2017 and 2025.
As the sheer volumes of data accumulated in the healthcare industry continue to grow, machine learning continues to be a viable possibility, with AI solutions getting stronger every day. However, since the healthcare industry is so overwhelmed by life and death decisions, can we really trust AI when it comes to our care?
For instance, an AI solution might be able to help spot irregularities in scans faster than a human doctor, making it easier to address a disease during the early stages. However, that doesn’t mean that it should be allowed to make decisions about the future of that patient. For now, it seems that AI is better suited to the role of “assistant” – helping human professionals make the choices that could change the world.
While AI certainly has the power to improve a lot of the things we do and experience in our personal and professional lives – that doesn’t mean it’s ready to replace human expertise – particularly in life or death circumstances. Leveraging AI in any sector means thinking carefully about the possibilities and limitations that exist for a machine. While these systems may be great at data, they don’t yet have the moral and ethical backgrounds to make decisions that take the control out of the hands of human beings.