Is AI the solution to all problems or the bane of jobs and society as we know it? There is a lot of hype about the capabilities of AI, but much less evidence of success in various types of problems. This is reminiscent of previous revolutions in AI. Back in 1970, Marvin Minsky had claimed that within a decade we would have AI with the intelligence of an average human being. Although this did not come to pass, AI has shown superhuman abilities in some tasks, such as playing Go, and has substantially improved our abilities to automatically classify images, recognise speech and handwriting, translate text, drive cars or enable the existence of AI-powered personal assistants. It has also helped to improve various analytics tasks, such as advertisement targeting. Does it do it perfectly? Of course not, but neither humans do any other available methods. Should we be worried that in a few years we are at risk of losing many jobs to AI? The OECD suggests no, countering an older Oxford University prediction.
We will not follow Minsky’s steps and put our own prediction on the table when there are so many unknowns, but there are a few certainties: both the modelling and the hardware technology are maturing, and AI is here to stay. Is it intelligent in the same way people are? Well, for that first we need a good definition of intelligence! Would you trust an AI doctor or a human doctor supported by AI? There is research that suggests human interaction and the comfort humans can give aids healing. Would you learn better from an AI teacher or a human teacher – this might be somewhat more complicated to answer. Nonetheless, one can consider decomposing this very difficult forecasting problem to smaller more manageable ones and construct scenarios to account for different uncertainties. We surely should not expect a yes/no forecast! But plan your or your children’s careers to ensure you’re not readily replaceable.
“This week’s forecast” is a blog series, which covers commentaries, by members of the Centre for Marketing Analytics and Forecasting, about topical and controversial forecasts in the news.Back to News