The answer to the title is definitely no. Of course, virtual assistants and chatbots do not have an abundance of common sense. The reason is simply the types of machine learning, which rely on specific situations they have encountered before, rather than using broader knowledge for answering the question. However, quite contrary to this, researchers at the Allen Institute for AI (Ai2) have devised a new test, i.e. the Arc Reasoning Challenge (ARC) that can test an artificial intelligence on its understanding of the way our world operates.
To speak in details, Humans resort to their common senses to fill in the gaps of any question they are posed, delivering answers within a comprehensible but non-explicit context. To this, Peter Clark, the lead researcher on ARC, explained in a statement, that, “Machines do not have this common sense, and thus only see what is explicitly written, and miss the many implications and assumptions that underlie a piece of text.”
Also, this test, asks basic multiple-choice questions based on general knowledge. For example, one ARC question is: “Which item below is not made from a material grown in nature?” The possible options include a cotton shirt, a wooden chair, a plastic spoon and a grass basket.
Now if machine learning can successfully pass the Arc Reasoning Challenge, it would refer to the fact that the system has a good knowledge of common sense that no AI currently possesses. It would definitely be a huge step forward, leading to smarter artificial intelligence, and move these systems closer to one step to the day when they would be taking over the world.