Language is the cornerstone of human intelligence.
The emergence of language was the most important intellectual development in our species’ history. It is through language that we formulate thoughts and communicate them to one another. Language enables us to reason abstractly, to develop complex ideas about what the world is and could be, and to build on these ideas across generations and geographies. Almost nothing about modern civilization would be possible without language.
Building machines that can understand language has thus been a central goal of the field of artificial intelligence dating back to its earliest days. It has proven maddeningly elusive.
This is because mastering language is what is known as an “AI-complete” problem: that is, an AI that can truly understand language the way a human can would by implication be capable of any other human-level intellectual activity. Put simply, to solve language is to solve AI.
This profound and subtle insight is at the heart of the “Turing test,” introduced by AI pioneer Alan Turing in a groundbreaking 1950 paper. Though often critiqued or misunderstood, the Turing test captures a fundamental reality about language and intelligence; as it approaches its 75th birthday, it remains as relevant as it was when Turing first conceived it.
Humanity has yet to build a machine intelligence with human-level mastery of language. (In other words, no machine intelligence has yet passed the Turing test.) But over the past few years researchers have achieved startling, game-changing breakthroughs in language AI, also called natural language processing (NLP).