As this study was written by field experts in Artificial Intelligence, I find it difficult to challenge any area of it. However, I do somewhat question how they chose to define intelligence. The definition given in the first section of the article states “...intelligence is that quality that enables an entity to function appropriately and with foresight in its environment.” This definition is extremely broad, and I feel could be improved. From the list of definitions given by Wikipedia, definitions vary from “goal-directed adaptive behavior”, one of the most specific, to an extremely vague “deal with cognitive complexity”. The definition given by Howard Gardner is most similar to how I’ve personally always viewed intelligence: as a person (or thing’s) ability to solve problems. This definition is not perfect, however. It works very well for humans and animals. People able to solve complex problems, like Stephen Hawking for his theoretical physics work, are seen as more intelligent than most. This extends well to animals. The octopus, for example, is viewed as one of the most intelligent animals for its ability to manipulate its environment in a near human-like way, including escaping from its tank. The other non-human apes are also seen as intelligent due to their use of tools and ability to learn very simplified language. Computers, however, are not capable like this. Instances of artificial intelligence are often extremely specialized, and are only able to perform a small number of tasks at a high level. For example, game-playing intelligences like AlphaGo are well-suited for the complex problem of playing go, but incapable of doing anything else. Similarly, a calculator, mentioned at the start of the article, can perform mathematical operations with much greater speed and accuracy than a human brain, but cannot do anything else. The article here suggests in “Defining AI” that one degree of autonomy and generality should be considered when deciding if a thing is intelligent, and with this I can strongly agree. While not much emphasis is placed on these, in favor of specificity and efficiency, they are important qualities to eventually create an intelligence rivaling that of a human mind.

