Tehran Times – Chomsky – Chat GPT, an AI language chatbot developed by OpenAI, has recently sparked controversy as some people view it as a promising step towards artificial general intelligence while others criticize its potential to surpass human intelligence. This debate has raised questions about the impact of AI technology on human thought processes and society.
The ongoing debate about the capacity and power of artificial intelligence and language machines has become a topic of heated discussion in the academic community.
To discuss the relationship between language and artificial intelligence (AI) and the implications of recent advancements in the field, the Tehran Times interviewed Noam Chomsky, a well-known American linguist who challenges the idea that AI systems can ever replicate the complexity and nuance of human language and thought.
While there is ongoing debate about the capacity and power of artificial intelligence and language machines, Chomsky’s skepticism about their ability to replicate human language and thought remains strong.
Chomsky argues that language is an innate capability of humans and that recent advancements in AI and machine learning have no bearing on matters concerning language, thought, learning, or cognition. While some may argue that programs can surpass human intelligence, Chomsky notes that the calculator in a laptop already does in some tasks, and that the fundamental distinction between language and the mind remains a topic of philosophical debate. Nonetheless, Chomsky acknowledges that the specific capability of language, which is beyond question, must have emerged at some point through genetic mutations. The confusion surrounding these issues highlights the need for continued research and discussion in the field of artificial intelligence and its impact on human society.
In response to a question about the origin of language, specifically whether it is an innate capability of humans or a result of the interaction between humans and their environment, Chomsky explained that the human language faculty is an “innate capability” and that there is overwhelming empirical evidence to support this claim.
“Infants reflexively acquire the ambient language, while other organisms presented with the same data take it to be just noise. The theory of genetic mutation is the only known way this innate capability could have emerged,” he noted.
Regarding the implications of recent advancements in AI and artificial language for linguistics, Chomsky argued that “there are no implications.”
While machines may become more adept at language processing, their design reveals that they are, in essence, high-tech plagiarism, the linguist said.
“They are incapable in principle of telling us anything about language, learning, cognition generally. The basic reason is simple: they do just as well for impossible languages that humans cannot acquire (except perhaps laboriously, as puzzles, plainly irrelevant) as for humanly attainable languages,” the American linguist added.
Chomsky stated that that they are like a “theory” of physics that cannot distinguish between things that happen and those that cannot happen, rendering them irrelevant in the study of actual language.
According to Chomsky, the fundamental irrelevance of these models is built into their design, and their improvement to deal with actual language only makes their irrelevance more evident, as they do better with impossible languages.
On whether there is a clear distinction between language and mind, Chomsky argued that the traditional view, dating back to classical Greece and classical India, is that language and thought are intimately related. This view holds that language generates thought, and thought is what is generated by language. While this conception was swept aside by behaviorist and structuralist currents in the early 20th century, it has been revived in the “generative enterprise,” and if it is correct, the questions raised do not arise.
Regarding the question of whether AI and ChatGPT can surpass human intelligence in the future, Chomsky argued that these systems are not just “far” from true intelligence but they have no relation to the question at all. He pointed out that the calculator in one’s laptop already surpasses human intelligence in some tasks, and ants in his backyard carry out computations for navigation that humans can approach only with sophisticated instruments. Chomsky acknowledged that there are issues related to ethics and emotions, but they reach far beyond current AI or the GPT systems and require a more significant philosophical discussion.
According to Chomsky, advancements in artificial and machine learning will not impact natural language or produce a new kind of language that replaces it. He believes that AI and machine learning have no bearing on matters concerning language, thought, learning, or cognition. Therefore, improvements in the same framework will not change this. In other words, AI and machine learning may be useful for some purposes, but they are not relevant to the study of language and its properties.
