“Machine learning tells us nothing!!!! It [human language] is too rich. … It [language] works as well with impossible languages as with regular languages. … [AI] is as useful as a bulldozer is useful. … It tells you nothing about mind and thoughts. … Computational complexity is what accounts for language. … [AI] will never provide [an] explanation to language. … Syntax gives language meaning…and syntax is totally independent of the external world [and is controlled by the brain].” (Chomsky, N., 2022, On Theories of Linguistics (Part 2), Dec. 30, 42.32 minutes, Youtube).
If Noam Chomsky is correct that his Universal Grammar, as controlled by the brain, is independent of the external world, then ChatGPT—which is based on vacuuming large amounts of information from the external world—will never be a suitable metaphor for human language. Furthermore, AI was not designed to provide meaning to its output and that it is merely a transmitter of an output that underwent a previous reconfiguration (i.e., a Factor Analysis that can act on its input). Therefore, the creativity attributed to AI—with its high energy cost as compared to that of the biological brain—will, at best, remain a tool in the hands of humans. This of course is good news, since humans need to be held responsible for all actions (good and bad) generated by AI (Harari 2024).
The statement critiques AI’s limitations in understanding human language, arguing that machine learning merely processes data without understanding the mind or meaning. Noam Chomsky’s view suggests that AI, relying on external world data, cannot replicate human language’s universality and independence from external stimuli. AI is framed as a tool lacking true creativity or comprehension, reinforcing that humans remain responsible for AI’s outcomes.
Another thing to mention is Chomsky’s notion of “poverty of the stimulus” which asserts that children acquire complex grammatical structures with relatively limited exposure, suggesting an innate linguistic endowment that drastically reduces the amount of external data needed for language acquisition. By contrast, large language models (LLMs) like ChatGPT rely on massive corpora—billions of tokens—essentially “vacuuming up” data from the external world to mimic certain linguistic patterns. While these AI systems can generate fluent outputs, they do not develop internalized principles of language through minimal input, but rather statistically approximate language use. As a result, their reliance on huge datasets contrasts sharply with human language development, where the brain’s innate “universal grammar” enables remarkable linguistic competence from far less data. This discrepancy points to a fundamental divide in how humans learn language versus how machines simulate it—highlighting that AI, however sophisticated, may still fall short as a true metaphor for human linguistic creativity and meaning-making.
When ChatGPT Broke an Entire Field: An Oral History
Researchers in “natural language processing” tried to tame human language. Then came the transformer...
“He said: ‘Let’s just write the academic paper version of this so that it’s not just ideas on Twitter, but peer-reviewed research. And that’ll put an end to it.’ It did not put an end to it.”
In an oral history of the rise of ChatGPT and other large language models, Emily Bender recalls how she and fellow computational linguist Alexander Koller attempted to cap the debate over how to interpret these systems’ capabilities..."
https://www.quantamagazine.org/when-chatgpt-broke-an-entire-field-an-oral-history-20250430/
https://aclanthology.org/2020.acl-main.463/
- Badges
- Science topic
- Similar topics
- Correction