Discussion 1: Handing your thinking to a tool
This discussion is in the first half of the second class in week 1. Before class, go through the four items under Before class. They take about 35 minutes in all. Class starts with ten minutes of writing, on paper and without AI, and then we talk. You may use AI to help you understand a reading, but not for the writing you do in class (see the AI use policy).
Why this discussion now
This week you turned off the AI that comes with VS Code, and in 010 you type every line of code yourself. AI tools are good at Python, and later in the semester you will use them. But it is hard to judge code you could not have written yourself, so for now you write it. The readings look at what happens to the skills of doctors, software engineers, scientists, and students who rely on AI, and some of the early results are worrying.
Before class
Is AI ruining our skills? Early results are in — and they’re not good. Nature, Mariana Lenharo, June 18, 2026. News article, about 7 minutes. It describes studies of physicians whose ability to spot precancerous growths fell after they got used to an AI assistant, and of software engineers who learned less when they had an AI helper. Hope’s library subscribes to Nature, so the article opens on the campus network. Off campus, find it through the library’s website.
What happens when scientists trust AI more than colleagues? The Conversation, Sungho Hong and Victor J. Drew (Institute for Basic Science), May 11, 2026. Article, about 7 minutes. Two researchers worry that early-career scientists will take their questions to a chatbot instead of to a mentor or a colleague, and miss the troubleshooting and argument that teach you how to do science.
How to stop AI from killing your critical thinking. TED, Advait Sarkar, given at TEDAI Vienna in September 2025. Talk, 15 minutes. Sarkar says that when we hand our reasoning to AI, we become “middle managers for our own thoughts,” and he describes an AI tool designed to nudge you to think for yourself instead.
More Students Use AI for Homework, and More Believe It Harms Critical Thinking. RAND, Heather L. Schwartz and Melissa Kay Diliberti, March 17, 2026. Short report, about 5 minutes to skim. In a national survey of 1,214 young people, the share who used AI for homework went from 48 percent in May 2025 to 62 percent in December 2025, and 67 percent agreed that the more students use AI for schoolwork, the more it will harm their critical thinking.
Learn more
These are optional.
- What do students lose when they stop writing? The New York Times, Dana Goldstein, August 17, 2026. Article, about 10 minutes. Goldstein talks to cognitive researchers about what drafting does for memory and for knowing what you think, and to teachers about what they are changing now that students can skip the drafting. Hope students can read the Times for free with an Academic Pass, and the library’s guide explains how to set one up.
- AI and Cognitive Offloading with Leon Furze. Intentional Teaching podcast, hosted by Derek Bruff, June 2, 2026. Audio, about 45 minutes. Furze argues that learners need deliberate moments of difficulty, which he calls resistance, to build the judgment and knowledge that make AI safe to use.
- Not all cognitive offloading is equal: distinguishing dependent and autonomous offloading to generative AI. Frontiers in Psychology, Qiuhan Zhu, Xiangnan Li, Yiang Dong, Pengcheng Chang, and Mengmeng Fan, July 16, 2026. Research article. The authors surveyed 589 university students and early-career workers three times, two weeks apart. They separate using AI as a substitute for your own thinking from using it as a starting point that you then check against your own reasoning. The two gave similar benefits at the time but were linked to different outcomes later. The paper is long, so start with the abstract and the discussion section.
Links last checked October 2, 2026.