Artificial intelligence has taken some punches. For years, IBM’s Watson promised to transform the business world but failed to live up to its own hopes and marketing blitz. Tesla CEO Elon Musk and others vowed to build cars that drive themselves, but they aren’t even close to that goal. In June, The Economist declared: “An understanding of AI’s limitations is starting to sink in.”
Martin Engelsen, a manager at a family-owned salmon farm on a small island off the coast of northern Norway, has a different view. He checks his iPhone once a day to see what is going on with his fish, something he couldn’t have imagined doing a few months ago. Engelsen gets information from images shot by cameras located 20 feet underwater in his fish pens. He pays Aquabyte, a San Francisco startup that installed the cameras in late May, for software that analyzes the images using deep learning—an advanced form of machine learning, the most common type of AI. The software measures and tracks the density of his 1 million fish and the number of parasites on their scales, which he is required to report to authorities.
Google, Amazon and other giants of the tech industry have both led and capitalized on advances in AI to become dominant forces in the U.S. economy. What is less well known is the extent to which businesses both inside and outside the tech arena are buying AI products and services to improve their operations, often from one of scores of venture capital–backed startups that have flooded the field in recent years. (See a closer look at AI software startups, including previously unreported revenue figures.)