Better Automation- “Science, my lad, is made up of mistakes, but they are mistakes which it is useful to make, because they lead little by little to the truth,” wrote Jules Verne.
I’ve been in the field of automation long enough to see that many of the most impressive advancements are still quite new. Take driverless cars, for instance. A decade ago, they were little more than a fantasy, but in February 2019, Elon Musk predicted that we’d have the technology for a fully self-driving car by the end of this year.
Driverless cars promise to transform transportation, but something like machine learning promises to transform everything. Smart systems that can learn from experience have applications in just about every aspect of life. In healthcare, researchers found that machine learning systems can correctly classify echocardiograms up to 92% of the time, whereas human doctors can only do so 79% of the time.
As automation technology has grown, so has its accessibility. Advances in natural language processing promise to liberate AI from purely technical settings and integrate it with more aspects of daily life. Amazon has sold 100 million Alexa-enabled devices, and almost every household technology staple is now available in a “smart” model.
I’m optimistic about this new era of automation, but I’m also cautious. Forward thinkers have always understood that technologies have the potential to create challenges as well as solve them. If we’re going to realize the highest potential of AI and automation, we need to not only acknowledge the worst aspects, but also do our best to avoid them.
When Automation Isn’t An Asset
We don’t need to imagine a futuristic scenario to see what bad automation looks like. Most of us are already inundated with robocalls — something that combines the annoyance of telemarketing with the persistence of automation. Robocalls may cut labor costs, but they alienate customers in the process. Unfortunately, this is just one of many examples of automation delivering shortsighted solutions.
Bad automation has made the news in recruiting, too. It’s a common misconception that AI is free of human bias. However, tools designed to automate candidate selection have been shown to exclude women and others because the algorithms they rely on are human creations full of subtle prejudices. In that way, AI can amplify the faults of humans rather than eliminate them.