While it is true that machine learning models like ChatGPT have limitations and have garnered a great deal of social attention due to their widespread application in everyday life, these limitations are not as severe as Chomsky and co-authors suggest in their recent New York Times article titled “Noam Chomsky: The False Promise of ChatGPT.” This article seeks to explain where the writers went astray.
Complex and varied, the link between technology and society is impacted by a variety of social, cultural, economic, political, etc. aspects. Social shaping and social construction of technology are theoretical frameworks created in the field of Science and Technology Studies (STS) that emphasize the significance of social and cultural elements in influencing technical development and use.
According to the renowned philosopher Langdon Winner, “technology is neutral in and of itself; it relies on how civilizations employ it.” His social shaping of technology approach says that technical progress is influenced by social and cultural elements, such as the interests, values, and power relationships of diverse social actors, such as users, designers, policymakers, and other stakeholders. This implies that technologies are neither neutral nor devoid of value, but are instead infused with social and cultural meanings that reflect the interests and viewpoints of its makers and users.
Similarly, the prominent STS academic Professor Trevor Pinch’s theory of the social construction of technology emphasizes the significance of social and cultural elements in defining how technologies are seen and utilized in society. So, technologies are neither intrinsically good nor harmful, nor do they depend only on their technological characteristics; rather, they are formed and defined by social and cultural norms and elements such as gender, race, class, and other aspects of social inequality.
For instance, advertisements for technology equipment can promote gender stereotypes and social hierarchy. Despite the fact that both technology may be utilized by anyone of either gender, washing machines are frequently sold to women while vehicles are promoted to males. Likewise, new technologies, such as cryptocurrencies like Bitcoin, can be used to democratize financial transactions, but they can also be utilized for criminal purposes like money laundering and terrorism financing. So, the problem is not the technology itself, but how it is seen, utilized, formed, and produced by human civilization, at its heart the human brain.
As Chomsky and his co-authors would put it, the present most popular and trendy strain of A.I. – machine learning-based technologies like ChatGPT — are marvels of the human brain, not an extraterrestrial phenomena. It can also be used to produce language in a number of circumstances, but it is up to humans to select how to utilize this language in a responsible and ethical manner for the greater good of society.
Chomsky and co-authors have primarily presented three reasons to demonstrate that ChatGPT makes misleading promises. The first is that machine learning algorithms such as ChatGPT are incapable of producing the Borgesian revelation of knowledge because they fundamentally vary from how humans reason and employ language. The second criticism is that machine learning models are incapable of explaining the laws of English grammar, resulting in predictions that are shallow and questionable. The third and last important point is that machine learning models are incapable of producing real intelligence because they lack moral reasoning ability.