They are capable of seemingly sophisticated results, but they can also be fooled in ways that range from relatively harmless -- misidentifying one animal as another -- to potentially deadly if the network guiding a self-driving car misinterprets a stop sign as one indicating it is safe to proceed. A philosopher with the University of Houston suggests in a paper published in Nature Machine Intelligence that common assumptions about the cause behind these supposed malfunctions may be mistaken, information that is crucial for evaluating the reliability of these networks. As machine learning and other forms of artificial intelligence become more embedded in society, used in everything from automated teller machines to cybersecurity systems, Cameron Buckner, associate professor of philosophy at UH, said it is critical to understand the source of apparent failures caused by what researchers call "adversarial examples," when a deep neural network system misjudges images or other data when confronted with information outside the training inputs used to build the network. They're rare and are called "adversarial" because they are often created or discovered by another machine learning network -- a sort of brinksmanship in the machine learning world between more sophisticated methods to create adversarial examples and more sophisticated methods to detect and avoid them. "Some of these adversarial events could instead be artifacts, and we need to better know what they are in order to know how reliable these networks are," Buckner said.

Robots, we are told, are going to take over the world, taking away most jobs as we know them. And yes, it is true that automation will make some jobs disappear. AI is also now intelligent enough that we will probably see a greater range of jobs change than we have ever seen before. This is likely to include some professional jobs that have always been "safe" from automation in the past. We are already seeing the beginning of this with online financial advice, translators, drivers, etc. However, the reality is that the job market is constantly changing.

"Social determinants of health" is one of the hot buzz-phrases in healthcare these days, and for good reason. SDOH refers to outside factors that may impact a patient's health, such as employment status and access to education, and providers can improve efficiency and curb costs by addressing these factors. Technology is often utilized to do so effectively -- and lately that means artificial intelligence and machine learning. Automation, and technology that learns as it goes, is one way providers can make sense of the glut of SDOH data, and make informed decisions based on it. That doesn't mean it's too early to jump onboard.