If you’ve been nervous about the buzz around Artificial Intelligence (AI) that it would take off your job soon and the technology works better than your brain, you’re probably wrong. First, there is nothing artificial about intelligence, and unlike industrial automation, which actually removes jobs globally, AI will only supplement human intelligence across the spectrum–from banking to media.
According to Gartner, AI consists of software tools to solve problems in its current state. While some forms of AI may give the impression that they are clever, it would be unrealistic to think that current AI is similar to or equivalent to human intelligence. “Some forms of Machine Learning (ML) — an AI category — may have been inspired by the human brain, but they are not equivalent,” says Alexander Linden, Gartner’s Research Vice President. For example, the image-recognition technology is more accurate than most humans but is of no use in solving a math problem.
“Today’s rule with AI is that it solves one task extremely well, but if the task conditions change only a little, it fails,” noted Linden. When it comes to bias, an ML model will always operate the way you’ve trained it, said Olivier Klein, Head of Emerging Technologies, Asia-Pacific at Amazon Web Services (AWS), which is the retail giant Amazon’s cloud arm. “If you train a model with a bias, you’d end up with a biased model. You have to train and retrain your ML model on an ongoing basis and the most important thing is that you need some form of end-consumer feedback, “Klein told IANS.
“ML is not about replacing people at all, but about improving experiences,” he added. Every AI technology is based on human experts’ data, rules and other types of input, and similar to humans, AI is also intrinsically biased in one way or the other.” Today there is no way to ban bias completely, but we have to try to minimize it, ” Linden said. IT and business leaders are often confused about what AI can do for their organizations and are challenged by a number of misunderstandings about AI.
According to Gartner, in order to devise their future strategies, they must separate reality from myths. “Every organization should consider the potential impact of AI on its strategy and investigate how this technology can be applied to its business problems,” Gartner said. Klein said that people are really good at learning quickly with very little information. “ML models are the opposite. They require a lot of data inputs to be trained. “I’d argue you’re showing someone a bicycle a couple of times and you’re showing them how to ride a bicycle and the human being can ride that bicycle quite easily. It takes millions of hours of training to just train a robot to ride a bicycle, ” Klein explained.
The truth is: machines are not here to make decisions on their own, and some human emotions, for example, empathy, can never be automated.