Chapter 1
Demystifying Artificial Intelligence
IN THIS CHAPTER
Exploring AI techniques: neural networks and machine learning
Checking out TensorFlow
Understanding AI limitations
Artificial intelligence (AI) has been a much misunderstood set of words over the past few years. The popular news media tends to take any small advance in AI out of context and proclaim “smart computers are here!” For example, in 2017, Facebook engineers programmed two programs to value certain objects more than others (balls, blocks, and such) and then had the two programs, through a rules set and a language like English, negotiate with each other to maximize the acquisition of objects that the programs valued.
The programs did not have a language syntax checker, and because of the way the programs learned the communication between the programs, they soon became syntactically incorrect English. (Here’s a good example: When a program wanted something, it would say “I want,” and the program logic decided that if one “I want” is good, saying it many times — “I want I want I want” — should be better.) The news media reported this as a new language. (It wasn’t.) And later, when the programs were shut down because ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access