Noam Chomsky, the highly regarded and controversial intellectual, celebrates his 95th birthday today. Chomsky is a prominent figure in linguistics and has made significant contributions to psychology, philosophy, and political activism.
One of Chomsky’s most influential contributions is his establishment of cognitive science as a discipline, which has played a fundamental role in the development of artificial intelligence (AI). However, as concerns grow about the potential negative impact of language-simulating machines on humanity, it is worth considering whether we have strayed too far from Chomsky’s vision of a science of the human mind.
Chomsky first gained fame in 1957 with his groundbreaking work, Syntactic Structures, which revolutionized the study of language. However, his real stardom came in 1959 when he published a scathing review of B. F. Skinner’s Verbal Behaviour. Skinner, a psychologist and behaviorist, was popular in psychology circles for his theory of “operant conditioning,” which explained how reinforcement and punishment shape behavior.
Skinner attempted to apply this theory to linguistics by breaking down language into components acquired through operant conditioning. Chomsky vehemently disagreed with Skinner, arguing that language could not be understood in this way. He pointed out that children do not receive enough exposure to learn every possible sentence, and language is inherently creative, with people frequently uttering sentences that have never been heard before.
Chomsky’s critique not only excluded behaviorism from linguistics but also emphasized the importance of examining the mind rather than just behavior in fields like anthropology, psychology, and neuroscience. This perspective contributed to the cognitive revolution and the emergence of cognitive science.
One key concept popularized by Chomsky is “generative grammar,” which posits that there are specific rules determining grammatical correctness in any language. Another concept is that of “deep structure,” which focuses on the underlying meaning of sentences rather than their surface structure. These ideas align with today’s generative AI and deep learning.
Chomsky’s influence on AI research can be traced back to his MIT colleague Marvin Minsky, an AI pioneer. Chomsky’s theories about language paved the way for expanding Alan Turing’s ideas about machine intelligence into language processing.
However, Chomsky remains skeptical of recent advancements in AI, such as large language models (LLMs) like ChatGPT. While some linguists believe these models invalidate Chomsky’s approach, he argues that they merely imitate rather than truly learn. He views them as a statistical mess rather than a meaningful analysis of language. Chomsky also worries that the popularity of these models will hinder the exploration of other AI architectures that do not rely solely on statistical data processing.
Despite his reservations about AI, Chomsky acknowledges its potential to contribute to a grim future. He warns that AI systems like ChatGPT can exhibit “the banality of evil” through plagiarism, apathy, and obviation. However, he considers climate change a more pressing concern than AI.
There is a significant difference between Chomsky’s ethical and optimistic work in cognitive science and the current state of the AI industry. While advances in modeling cognition used to occur primarily in universities, large corporations like Google, Microsoft, and OpenAI now dominate the field. Some researchers are turning to AI models for insights into human thinking, but Chomsky and others doubt the effectiveness of this approach. Instead, these models are primarily focused on generating profit by reinforcing addictive and influential behaviors.
Ironically, Chomsky’s contributions to understanding language and the mind have inadvertently led us into an arena ripe for behaviorist experimentation facilitated by AI.
