按年被提到
20252026 · 17 次2026
AI 安全、伦理与社会影响›大模型与人类行为·arXiv:2506.08872 · cs.AI

Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task

你的大脑与 ChatGPT:用 AI 助手写作文时认知债务的累积
Nataliya Kosmyna, Eugene Hauptmann, Ye Tong Yuan, Jessica Situ, Xian-Hao Liao, Ashly Vivian Beresnitzky, Iris Braunstein, Pattie Maes · 2025

EEG 研究:用 LLM 写作的人脑连接更弱

开始阅读 ↗PDFalphaXiv 讨论
被推荐 · 22 人首次 2025-06-11 · 最近 2026-07-12
22不同的人
9活跃月份
25讨论串
1,007被引(OpenAlex)

研究把参与者分成 LLM、搜索引擎和纯大脑三组,各完成三次作文,第四次互换条件,并用脑电图(EEG)评估认知负荷,再结合 NLP 分析与教师及 AI 评分。纯大脑组的脑网络连接最强、最分散,LLM 组最弱;LLM 组对作文的归属感最低,也难以准确引用自己写的内容,四个月中在神经、语言和行为层面持续落后,作者据此提出对长期依赖 LLM 的教育隐忧。

HN 为什么提它

评论者把它当作 LLM 依赖会削弱思考、学习和技能的证据反复引用,也有人称赞它篇幅与深度更像一本专著。

实证研究难度 2 / 5约 120 分钟经得起时间
读之前最好知道
1脑电图(EEG)与脑连接性基础
2实验设计与统计
3大语言模型的使用
4认知负荷(cognitive load)

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
推荐xenophonf · 2026-01

“I'm very impressed. This isn't a paper so much as a monograph.”

称赞印象深刻,认为这更像一部专著而非论文
Brain on ChatGPT: Accumulation of Cognitive Debt When Using an AI Assistant原评论 ↗
批评djoldman · 2025-06

“"likely decrease in learning skills" is quite right here.”

质疑论文“学习能力可能下降”的说法是否站得住
My "Are you presuming most people are stupid?" test原评论 ↗
批评tomrod · 2025-09

“Given the low N and the likelihood that this is drawn from 18-22 year olds attending MIT, one should expect an uphill battle for replication and for generalizability.”

MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline原评论 ↗
批评dcre · 2026-04

“I recommend people look at the actual study and think about how representative are the subjects, the tasks involved (SAT essay writing), and the way LLMs are being used.”

批评被试代表性、SAT 写作任务和 LLM 用法未必贴近现实
AI chatbots could be making you stupider原评论 ↗
引用trod1234 · 2025-06

“there is an increasing body of literature showing that LLM use increases cognitive debt.”

拿它说明越来越多文献表明用 LLM 会增加认知债务
League of Professional System Administrators - Board to Dissolve Organization原评论 ↗
引用lexandstuff · 2025-06

“a recent study has highlighted the significant cognitive costs associated with offloading problem-solving onto LLMs”

引用它说明把解题外包给 LLM 有显著认知代价
Generative AI coding tools and agents do not work for me原评论 ↗
看全部 20 条 →
引用dlock17 · 2025-06

“but regardless, here you go. It was measured.”

拿它作为已被测量的证据,回应生产力的说法
Generative AI coding tools and agents do not work for me原评论 ↗
引用bdbenton5255 · 2025-06

“In other words, working with LLMs reduces the level of engagement in our brains and offloads much of our cognitive activity.”

引用结论:用 LLM 降低大脑投入度,并把认知活动外包出去
Who is using AI to code? Global diffusion and impact of generative AI原评论 ↗
引用lukasego · 2025-06

“Most recently and famously, in this HN thread on the MIT paper on cognitive debt.”

提到 HN 上关于这篇 MIT 认知债务论文的讨论串
Show HN: Cogilo – Cognitive Mirror in Google Docs原评论 ↗
引用jrflowers · 2025-06

“As an aside, “dumb” is subjective, though if we had to put a label on it, “consistently underperformed at neural, linguistic, and behavioral levels” sounds like it could be something?”

引用其结论:在神经、语言和行为层面持续表现较差
My "Are you presuming most people are stupid?" test原评论 ↗
引用agentultra · 2026-01

“And there's growing evidence that you don't learn as much when using them [0].”

拿它说明用 LLM 时学到的东西越来越少
Programming, Evolved: Lessons and Observations原评论 ↗
引用misswaterfairy · 2026-01

“It seems this study has been discussed on HN before, though was recently revised very late December 2025.”

指出这项研究 HN 上讨论过,并在 2025 年底修订过
Brain on ChatGPT: Accumulation of Cognitive Debt When Using an AI Assistant原评论 ↗
引用Der_Einzige · 2026-01

“I don't need to "prove it", because all I have to do is link this:”

直接贴出这项研究,当作无需再论证的证据
Vitamin D and Omega-3 have a larger effect on depression than antidepressants原评论 ↗
引用Der_Einzige · 2026-02

“Also, even if you somehow didn't AI generate this, you sure as shit got infected by the LLM mind-virus and now you write/talk like it.”

借它讽刺对方被 LLM 洗脑,写话说话都像 LLM
Nano-vLLM: How a vLLM-style inference engine works原评论 ↗
引用cafebabbe · 2026-03

“Quite the opposite Your Brain on ChatGPT”

拿它反驳,说明过度使用 LLM 会让思维变钝
747s and Coding Agents原评论 ↗
引用dakolli · 2026-03

“people over reliant on these tools are simply ngmi”

拿它佐证过度依赖这些工具的人会走下坡路
I'm OK being left behind, thanks原评论 ↗
引用intended · 2026-04

“People have measurably lower levels of ownership and understanding of AI generated code.”

引用它说明人们对 AI 生成代码的归属感和理解度明显更低
AI assistance when contributing to the Linux kernel原评论 ↗
引用latexr · 2026-04

“That completely ignores the loss of skill that happens without you realising, as you lean more on a tool.”

拿它说明越依赖工具,技能会在不知不觉中流失
AI-Assisted Cognition Endangers Human Development原评论 ↗
引用rsoto2 · 2026-05

“similar to what is theorized about over-reliance on an llm”

认为它与过度依赖 LLM 的理论推测相似
Have a Coherent AI Policy原评论 ↗
引用bendriv · 2026-06

“Here is one study that researched AI assistance in a writing/learning context: Your Brain on ChatGPT.”

举它作为研究 AI 辅助写作与学习的一项研究
Norway imposes near ban on AI in elementary school原评论 ↗