Daily Productive Sharing 1273 - ChatGPT Rots Brains?

One helpful tip per day:)

Anne-Laure Le Cunff summarized findings from a recent MIT report about how using AI can accelerate cognitive offloading, sometimes referred to as “brain erosion”:

1. They discovered that the “brain-only” group showed the strongest and most widespread brain connectivity, especially in the alpha and beta frequency bands linked to attention and memory.
2. In the “brain-only” group, 89% of participants could accurately quote original material, compared to 83% in the “search” group. Shockingly, in the “LLM” group, not a single participant managed to quote the text accurately—their accuracy rate was 0%.
3. NLP analysis showed that essays written with ChatGPT’s assistance often relied on similar phrasing and examples.
4. Researchers pointed out that overreliance on AI tools can lead to so-called “cognitive debt,” meaning people don’t engage deeply enough to truly learn or remember.
5. In reality, people typically use AI far more flexibly than the rigid instructions given to the LLM group, who were told to rely solely on AI.
6. She advises not to let AI do all your thinking for you. It’s fine to use it to support your process—for brainstorming or rephrasing sentences—but the core concepts should come from your own mind.
7. The goal of using AI is to challenge your thinking, not to replace it.
8. Write first, then refine. If you care about a topic, spend time writing your own thoughts before asking AI to help polish or edit them.
9. The more you rely on AI to complete entire tasks, the more you risk gradually forgetting how to perform them independently.
10. Experiment with different ways to use AI for co-writing, brainstorming, or editing—and reflect on which approaches not only make you more efficient but also more creative.

If you enjoy today's sharing, why not subscribe

Need a superb CV, please try our CV Consultation

Anne-Laure Le Cunff 解读了最近 MIT 关于使用 AI 会加速脑蚀的报告：

1. 他们发现，“仅用大脑”组表现出最强、最广泛的大脑连接，尤其是在与注意力和记忆相关的 alpha 和 beta 频段上。
2. “仅用大脑”组中，有 89% 的人能准确引用原文，而“搜索”组为 83%。令人震惊的是，“LLM”组中，竟然没有一个人能够准确引用原文，准确率为 0%。
3. NLP 分析显示，ChatGPT 协助写作的文章往往使用类似的措辞和例子。
4. 研究者指出，过度依赖 AI 工具可能导致所谓“认知债务”，也就是人们没有投入足够的思考，因而难以深入学习或记忆。
5. 事实上，人们使用 AI 时往往比 LLM 组所接受的严格指令更灵活，LLM 组被要求完全依赖 AI。
6. 她建议，不要让 AI 替你做所有思考。可以用它来辅助你的流程，比如头脑风暴或润色句子，但核心概念仍然要靠自己去思考。
7. 使用 AI 的目的是挑战你的思考，而不是取代它。
8. 先写，再精修。如果你在意一个主题，先花时间写下自己的想法，再请 AI 帮忙修改或润色。
9. 你越是依赖 AI 完成整个任务，就越可能逐渐忘记自己如何独立完成这些事情。
10. 尝试用不同方式利用 AI 进行共同写作、头脑风暴或编辑，并且主动反思哪些做法不仅让你更高效，也让你更有创造力。

如果你喜欢的话，不妨直接订阅这份电子报 ⬇️