AI (or artificial intelligence) is such a big buzz word/phrase, right now, as more and more new products are claiming to use the technology, with plenty more predicted to come. AI is everywhere… from transportation to healthcare. This year’s CES even featured a few products claiming to be AI-infused, although how many of these products are actually AI? Should we be worried about the rise of the machines? Well, it really depends on what your definition of AI is.
To many of us, you say AI and you immediately conjure up scary images from movies like Terminator or iRobot, with a real fear that one day machines will eclipse mankind and relegate us to some form of servitude or post-apocalyptic society. In reality, the version of AI that we are now seeing is actually a subcategory called machine learning. It is really a form of coding that is allowing a machine to finish its programming by learning from various patterns and behaviours that it observes. It really doesn’t have an awareness of self or independent thought beyond the given task. (Big sigh of relief there!)
So, looking at real world examples, take the self-driving car. It uses the data, rules and parameters that it was given (highway code laws, maps, your chosen route, collision avoidance rules) to navigate roads it hasn’t travelled on before and get you to a destination. It may use AI to make an instant decision to avoid a collision or choose a better route in a heavy traffic situation, but it doesn’t actually think spontaneously for itself and only seeks an outcome to arrive at a predetermined destination in one piece. No sudden distracted diversions because of its personal opinion that it thought the little red hatchback was cute and wanted to see where it was going! (On the very rare occasions when these cars have been involved in an accident, reports have shown that 94% of those cases were, in fact, as a result of human error, not the technology).
It really does have some great benefits, as a computer’s black or white “thinking process” can see patterns that the human brain can miss, leading to better results. The learning process also means far less initial programming, as you are not giving it every single scenario or piece of data to work out its results – it is figuring that out on its own. Used right, machine learning can make more accurate predictions, so companies can make business decisions based on that data-driven outcome.
Machine learning technology can save lives –medical tech that requires action when it knows body chemistry isn’t correct (eg an insulin pump that uses data to know when to deliver a dose). It can save time and money – predicting a product’s performance success & making design adjustments in the early development stage, rather than having to spend more time finding those flaws at the prototype stage. The benefit list goes on.
Of course, research and development into AI is still very much ongoing, and its definition (& role in our society) will definitely change over the next 10, 20 or 50 years. It may replace the jobs of some, but it will create jobs in other areas. For now, though, it is very much at the forefront of current technology development trends and is something to embrace rather than fear.