Rough Draft Project 2 by yantung zhang

Key terms

g2ggk9wsbv — 2025-09-29 19:27:28 UTC

AI" / Large Language Models (LLMs)

g2ggk9wsbv — 2025-09-29 19:30:37 UTC

The authors consistently put "AI" in quotation marks to question the term. They define systems like ChatGPT as "random parrots" - that is, statistical models that merely recombine human language based on probability without understanding its meaning.

Stochastic Parrots

g2ggk9wsbv — 2025-09-29 19:33:00 UTC

A highly influential and repeated term that effectively encapsulates the authors' core argument about how LLMs work.

Primary Claim

g2ggk9wsbv — 2025-09-29 19:35:31 UTC

The humanities and humanities educators hold a unique leadership position, capable of guiding society's response to "AI" to counter the corporate hype and technological determinism promoted by tech giants. They are the essential "domain experts" who safeguard critical thinking, democratic dialogue, and the true purpose of education.

Evidence 1

g2ggk9wsbv — 2025-09-29 19:37:47 UTC

The author refutes the professors' stereotype of being ignorant by presenting examples of proactive responses in the real world. For instance, public schools in New York City banned the use of ChatGPT shortly after its release.

Evidence 2

g2ggk9wsbv — 2025-09-29 19:39:31 UTC

The author substantiates the role of the humanities by defining their core professional knowledge. They teach writing not as the creation of a product, but as a "friction" process that encompasses research, critical thinking, and revision, which is crucial for developing strong ideas. They are also capable of contextualizing the social impact of technology.

Evidence 3

g2ggk9wsbv — 2025-09-29 19:40:09 UTC

The author believes that having students generate initial drafts using chatbots is fundamentally flawed in terms of pedagogy, as it "short-circuits students' thinking" and starts from a "superficial foundation" of derivative products.

implications

g2ggk9wsbv — 2025-09-29 19:54:34 UTC

A subtle stance on technology

g2ggk9wsbv — 2025-09-29 19:55:15 UTC

The author explicitly states, "We are not proposing to expel LLMs from the Earth's surface." They acknowledge that "deep" machine learning has promising applications in areas such as protein folding or nuclear fusion. This acknowledgment prevents them from sounding like pure Luddites and enhances their credibility.

g2ggk9wsbv — 2025-09-29 19:59:27 UTC

“Of course, we are not proposing to banish LLMs from the face of the earth. Both of us co-authors teach about ChatGPT in our humanities classrooms

g2ggk9wsbv — 2025-09-29 20:09:24 UTC

“We are also quite hopeful about the myriad possibilities of ‘deep’ machine learning in domains such as protein folding or nuclear fusion.”

Multimodal aspects and reading experience

g2ggk9wsbv — 2025-09-29 20:13:00 UTC

The essay use hyperlink to enhance the reading experience by providing authority and credibility. For example, links to:

The "Random Parrot" paper and Ted Chiang's analysis enable interested readers to verify and explore the sources of the author's claims.

Mouth without a brain / Mouth on a leash

g2ggk9wsbv — 2025-09-29 20:22:50 UTC

Memorable, visceral metaphors that complicate the image of AI as intelligent.

For the future of knowledge

g2ggk9wsbv — 2025-09-29 20:49:20 UTC

If left unchecked and unreliable AI-generated content spreads unchecked, it will further polarize society and degrade the information ecosystem. This makes the critical skills taught in the humanities more valuable than ever before.

g2ggk9wsbv — 2025-09-29 20:50:10 UTC

“In the near term, that means avoiding panic, combating hype, and educating oneself and one’s colleagues about the topic. In the longer run, humanities educators can be leaders in helping students and the public to grasp the possibilities and perils of language models and other ‘generative AI.“

Secondary claim

g2ggk9wsbv — 2025-10-07 17:49:48 UTC

Through their core methodologies of critical analysis, historical contextualization, and ethical reasoning, the humanities are uniquely equipped to deconstruct the "myth of AI"—revealing it not as a magical intelligence, but as a sociotechnical system built on human labor, biased data, and corporate interests.

The metaphor of the "random parrot"

g2ggk9wsbv — 2025-10-07 17:53:43 UTC

_{“large language models of this kind are best understood as ‘stochastic parrots’—programs that generate plausible text... without any reference to meaning” (Page 5).}

This metaphor from a key research paper is a model of critical analysis. It strips away the mystique of "intelligence" and precisely describes LLMs as a meaningless statistical pattern-matching system.

Evidence 1 :Critical analysis

g2ggk9wsbv — 2025-10-07 17:54:01 UTC

Thesis

g2ggk9wsbv — 2025-10-07 18:01:53 UTC

The humanities can and should challenge and reshape today's understanding and application of "AI," and support critical thinking, writing education, and democratic values in a technology-driven era.

Tracing the Technological Genealogy and Capital-driven Nature of "AI" The Cold War Origins and Revival of the "AI" Term

g2ggk9wsbv — 2025-10-07 18:42:39 UTC

_{"The term 'artificial intelligence' was a creation from the Cold War era." And its resurgence "in the midst of great glory" in the era of deep learning and big data (Pages 4-5).}

By placing "AI" in its historical context, it is revealed that it is not a linear and inevitable technological advancement, but rather a field that has specific historical roots and fluctuates with the ebb and flow of funds.

Evidence 2: Historical Context

g2ggk9wsbv — 2025-10-07 18:43:15 UTC

Evidence 3: Ethical reasoning:

g2ggk9wsbv — 2025-10-07 18:47:41 UTC

The hidden low-wage workers

g2ggk9wsbv — 2025-10-07 18:49:34 UTC

_{"The process of detoxifying ChatGPT required Kenyans earning less than $2 per hour to label graphic content at a high speed (including 'child sexual abuse, bestiality, murder, suicide, torture, self-harm, and incest')." (Page 6).}

This is the most powerful ethical accusation against the "seamless" and "pure" AI myth. It reveals that the operation of this system relies on the exploited labor of people in the global south， thereby dimming its "magical" aura.