Any radical new technology understandably raises concerns about the effect it will have on our lives, our jobs and our society as a whole. During the industrial revolution of the 19th century, textile workers were worried about machinery taking over the jobs of skilled weavers. In the 1960s there were fears over computers reducing the need for office workers. The 1980s saw robots taking over repetitive jobs in factories.
The latest technology to lead to concern about what it may mean for our future is artificial intelligence. There are fears that self-driving vehicles will make drivers redundant, and that ‘thinking’ machines may even threaten the very future of humanity. And yet AI is already in use in everyday devices such as digital assistants and is being used by business to calculate the risks on your insurance policies, and predict what you might buy on your next trip to the supermarket.
So, what’s the future for AI? Will machine intelligence outstrip that of humans? Are we looking at a doomsday scenario or could it change our lives for the better?
Risk versus reward
Even some of the luminaries of the technology world like Bill Gates have warned of the potential dangers of AI. Electric car guru Elon Musk is among the backers of OpenAI, [1] a group dedicated to developing AI to benefit humanity.
As we rely on technology to make decisions about business, but also potentially about issues such as criminal justice and health, there is a risk that choices could be made without sufficient scrutiny. On the other hand, however, AI’s ability to spot patterns in huge amounts of data could be used to solve all kinds of problems, such as identifying cancers and other diseases from scans. The technology is also being used to combat cyber attacks and fraud by identifying suspicious behaviour.
Undoubtedly some jobs will be rendered obsolete by AI, but it also opens up opportunities. Humans could be freed to take on more fulfilling and creative tasks, and there will be new roles in support if the AI systems themselves.
The biggest risk is not of a machine takeover, but that we place too high a degree of trust in AI. The technology works because it’s trained to spot patterns and respond to them. The problem is that when it comes up with a result, we don’t actually know how it got there. If the technology gets it wrong, therefore, and no one spots it, the mistake could end up being compounded over time. There’s also the fact that if the data being fed in is imperfect, the results will be too. It’s vital to recognise this and it could lead to a whole new set of opportunities in verifying data.
1 https://openai.com
