Science fiction films starring robots or intelligent machines (such as Blade Runner, Real Humans, Westworld and others) have made us somewhat afraid when we look at a future with artificial intelligence (AI). But do we really have anything to fear from them? And can robots and computers actually take over our jobs?
It would certainly be handy if some of the work humans do now could be done by robots instead – as we can see from mining, where there’s the Joy Global’s Longwall Mining System in which all sorts of robots and machines beaver away automatically as they dig under the ground. It’s a lot safer for the humans above, who still keep an eye on things and who go and lend a hand when problems arise.
But what about other jobs? Everyone automatically thinks of production line work, but in actual fact other job areas could also be taken over partly by AI computers, such as medical diagnoses, accountancy tasks or the law. Sounds a bit awkward? It doesn’t have to be too bad – so long as we can trust and understand the computers and how they work.
Gaining that understanding means we will need to be entirely aware of how our computers come to a particular decision. For example, scientists have discovered that we need to be careful with the data we provide as input to the system. If you only show it sensational images of accidents and disasters, then a psychopathic robot will assume that this is the normal world. While this is a conscious test of how an AI system can be manipulated, it also happens in real life. One chatbot on Twitter began posting racist messages because it learnt to do so from other Twitter users and had not been given any instructions to recognize that type of statement as inappropriate.
And so, if we can understand how a system comes to a particular decision, we can also make adjustments to it, if necessary. Which is why it would be best for AI systems to be tested regularly so that any problems with ethical issues can be identified and dealt with. Just like an elevator in a building is checked regularly and issued with a certificate. That way we know we can trust it!
But what if robots and computers can in fact be depended on and so are capable of carrying out part of our work properly (i.e. the way we would do it as humans)? If we can get to that stage, then we can do new types of jobs in collaboration with ‘them’. Leaving us to concentrate on the things that humans are good at: being creative, developing human contact, responding to unexpected situations, understanding people’s intentions, etc. We will, in fact, complement each other.
It may well be that we start working less and that economic models will be completely different. Working, earning money, spending it in stores or online and so on is a model that perhaps needs to be totally rethought. Maybe we could also earn money by selling our data to companies that are interested in it. And perhaps by 2035, lots of services will have become free, like there are lots of free apps at the moment. One thing is certain: life will certainly be different with a robot by your side!