Most of what’s here is cited somewhere on the site. If you want to go deeper, these are good places to start.

Note that research on AI moves quickly. Check publication dates, and expect newer findings to refine some of these.

Understanding how AI works

• Large language models, explained with a minimum of math and jargon (Timothy B. Lee and Sean Trott). A readable introduction to how LLMs work.
• What is a Foundation Model? An Explainer for Non-Experts (Stanford HAI and the Center for Research on Foundation Models). A short video on the models behind today’s AI.
• Large Language Models explained briefly (3Blue1Brown). A short video introduction.
• Transformer Explainer (Georgia Tech). Watch a real model predict the next token.
• What Is ChatGPT Doing … and Why Does It Work? (Stephen Wolfram). A longer, more detailed explanation.
• Tracing the thoughts of a large language model (Anthropic). What interpretability research can, and can’t yet, see inside a model.
• The Foundation Model Transparency Index (Stanford CRFM). How much major AI developers disclose about their models.

How AI tends to behave

• Why language models hallucinate (OpenAI, 2025)
• Towards understanding sycophancy in language models (Anthropic, 2023)
• Sycophancy in generative-AI chatbots (Nielsen Norman Group, 2024). A plain-language overview.
• Sycophantic AI decreases prosocial intentions and promotes dependence (Cheng et al., Science, 2026)
• AI chatbots remain confident, even when they’re wrong (Carnegie Mellon University, 2025)
• Reasoning models don’t always say what they think (Anthropic, 2025)

How people use AI

• What large language models know and what people think they know (Steyvers et al., Nature Machine Intelligence, 2025)
• The impact of generative AI on critical thinking (Microsoft Research and Carnegie Mellon, 2025)
• Folk psychological attributions of consciousness to large language models (Colombatto and Fleming, Neuroscience of Consciousness, 2024)
• How AI and human behaviors shape psychosocial effects of chatbot use (MIT Media Lab and OpenAI, 2025)

AI in education: risks, benefits, and differing views

AI in education is changing quickly, and educators disagree about how to approach it. It’s worth reading across that range.

• A New Direction for Students in an AI World: Prosper, Prepare, Protect (Brookings Center for Universal Education, 2026). A global study concluding that, for now, the risks of AI in education outweigh its benefits, largely because they undermine children’s foundational development.
• The Digital Delusion (Jared Cooney Horvath). A book and Substack. Engage with it understanding that it argues against adopting AI in education, drawing on research critiquing educational technology.
• Teaching with AI (José Antonio Bowen and C. Edward Watson). A book and companion website. Engage with it understanding that it argues for adopting AI in education, with ideas for how AI could benefit student learning and teacher effectiveness.

Research by setting

• Education: Generative AI without guardrails can harm learning (PNAS, 2025), How university students use Claude (Anthropic, 2025), and Your brain on ChatGPT (MIT Media Lab, 2025)
• The workplace: Navigating the jagged technological frontier (Harvard Business School, 2023)
• Health care: Automation bias in mammography (Radiology, 2023)
• Law: Legal models hallucinate in 1 out of 6 (or more) benchmarking queries (Stanford HAI, 2024) and the AI Hallucination Cases database
• News and search: AI search has a citation problem (Columbia Journalism Review, 2025)
• Teens: How and why teens use AI companions (Common Sense Media, 2025)