More A than I: Why Artificial Intelligence Isn't, but You Are

Artificial Intelligence has the potential to change the world. The application of A.I. to robotics, control systems, text recognition, voice recognition, and autonomous vehicles will prove both revolutionary and useful. However, humans and machines process information in fundamentally different ways, and if those differences are not appreciated and exploited, a great deal of time and money will be wasted by expecting the wrong outcomes and capabilities of our machines. This paper reviews the most critical differences between human and machine information processing from the point of view of information theory, and also introduces the concept of "information lensing" that provides a novel framework for understanding patterns of information processing that are currently unique to humans. The addition of information lensing to efforts to mimic human thought processes through a combination of neural networks, distributional semantics, and natural language computation may provide a closer convergence between human and machine information processing.