Read the passage and mark the letter A, B, C or D on your answer sheet to indicate the best answer to each of the following questions from 3...
Đề bài
Read the passage and mark the letter A, B, C or D on your answer sheet to indicate the best answer to each of the following questions from 31 to 40.
CHATGPT MAY BE ERODING CRITICAL THINKING SKILLS?
A preliminary, not-yet-peer-reviewed study from the MIT Media Lab contends that habitual use of Large Language Models (LLMs) could prove deleterious to learning, particularly for younger users. The paper’s lead author, Nataliya Kosmyna, released the findings pre-emptively, citing profound apprehensions that society's adoption of LLMs for convenience might come at the cost of long-term cognitive development. [I] She articulated a pressing fear of precipitous policy decisions, such as a "GPT kindergarten," which she believes would have a pernicious impact on developing brains.
The study directed subjects to compose essays from SAT prompts. One cohort utilized ChatGPT, another relied exclusively on their intrinsic cognitive faculties, while a third used Google Search. The divergence in outcomes was stark. The ChatGPT cohort produced derivative, formulaic essays that assessors described as "soulless." Concurrently, their EEG scans revealed diminished executive control and attentional focus. [II] Conversely, the "brain-only" group manifested high neural connectivity in frequency bands associated with creative ideation and memory, reporting greater ownership and satisfaction with their work. The Google Search group's similarly active brain function highlighted a crucial distinction between traditional search protocols and AI-driven information synthesis.
The experiment's second phase, which focused on knowledge retention, accentuated the cognitive disparities even more starkly. When required to reconstruct their essays from memory, the former ChatGPT users faltered, exhibiting weak brain wave patterns indicative of a circumvention of deep mnemonic processes. "You basically didn't integrate any of it into your memory networks," Kosmyna observes. [III] In stark contrast, when the "brain-only" group was permitted to use ChatGPT for the rewrite, they demonstrated a significant enhancement in brain connectivity, implying that AI may function as a potent augmentation tool, provided it does not supplant the initial cognitive exertion.
Beyond the laboratory setting, these findings resonate with clinicians observing current trends. Psychiatrist Dr. Zishan Khan, for instance, expresses similar concerns, warning that such overreliance can cause essential neural pathways for information retrieval and resilience to atrophy. Kosmyna's team is now investigating AI's impact on software engineering, with findings that, even at this early stage, she describes as "even more severe," raising questions about the erosion of critical thinking in the workforce. [IV] While the scientific literature on AI's cognitive impact remains nascent, this study contributes a significant, cautionary datapoint to the discourse surrounding its integration into scholastic and professional spheres.
Question 31: According to paragraph 1, Nataliya Kosmyna ________.
A. suggests society delay LLM adoption until her study is peer-reviewed
B. chose to publish her work early because of some very deep concerns
C. attributes the pernicious impact of LLMs to their inherent convenience
D. has a pressing fear about the effects of a "GPT kindergarten" on policies