Implement the following plan:

Light Annotations for Two New Yorker Reading Articles

Context

Two New Yorker articles have been captured, extracted to MDX, and manually cleaned up. They're now ready for light annotations — sidenotes, footnotes, and external links — to help students engage with the material. Both articles are more accessible than the McLuhan reading (which has ~18 dense sidenotes), so annotations should be lighter: brief context on key people, links to referenced works/concepts, and a few sidenotes where the articles connect to each other or to course themes.

Files to Modify

• content/spring2026/readings/chatgpt-is-a-blurry-jpeg-of-the-web-newyorker-ed2e.mdx (Ted Chiang, Feb 2023)
• content/spring2026/readings/the-case-that-a-i-is-thinking-newyorker-d419.mdx (James Somers, Nov 2025)

Annotation Syntax (no imports needed)

• Footnotes: ^[inline text] — for brief biographical/definitional notes
• Sidenotes: [highlighted text](sn:id) in prose + <Sidenote id="id">content</Sidenote> block after paragraph — for longer commentary
• External links: Standard [text](url) — auto-converted to external links by MDX pipeline

Chiang Article: "ChatGPT Is a Blurry JPEG of the Web"

Footnotes (~5)

1. David Kriesel (line 16): ^[David Kriesel, German computer scientist and data analyst who documented the Xerox bug in detail on his website, bringing widespread attention to the issue.]
2. Marcus Hutter (line 96): ^[Marcus Hutter, German-Australian AI researcher at DeepMind known for theoretical work on universal artificial intelligence and the AIXI model.]
3. Ted Chiang — add authorUrl to frontmatter: authorUrl: "https://en.wikipedia.org/wiki/Ted_Chiang"

Sidenotes (~3)

1. sn:compression-understanding (around line 103-104, "the greatest degree of compression can be achieved by understanding the text"): Note that this specific passage is later quoted by Somers in his 2025 article as evidence for AI understanding — Chiang's own argument turned on its head. Cross-reference the Somers article.

2. sn:gpt4-prediction (around line 170-173, Chiang's prediction about GPT-4 training data): Chiang was largely correct — "model collapse" from training on AI-generated text became a major research concern by 2024. Link to relevant context.

3. sn:writing-process (around line 207-213, "Your first draft isn't an unoriginal idea expressed clearly; it's an original idea expressed poorly"): This argument about writing as a thinking process connects directly to course themes. Students using AI to write are skipping the cognitive work that produces understanding — the same compression/understanding link Chiang builds earlier.

External Links (~3)

1. Link "Hutter Prize" text to the official prize page
2. Add authorUrl and sourceDetail to frontmatter

Somers Article: "The Case That A.I. Is Thinking"

Footnotes (~8)

1. Dario Amodei (line 12): ^[CEO and co-founder of Anthropic, the company that builds Claude.]
2. Sam Altman (line 15): ^[CEO of OpenAI, the company behind ChatGPT and GPT-4.]
3. William Gibson (line 37): ^[William Gibson (b. 1948), science fiction novelist who coined the term "cyberspace" in *Neuromancer* (1984).]
4. David Rumelhart, Geoffrey Hinton, and James McClelland (line 71-72): ^[The "PDP group" — their 1986 book *Parallel Distributed Processing* helped launch the connectionist revolution in cognitive science.]
5. Douglas Hofstadter (line 145): ^[Douglas Hofstadter (b. 1945), cognitive scientist and author. His *Gödel, Escher, Bach* (1979) is a landmark exploration of consciousness, self-reference, and the nature of minds.]
6. Pentti Kanerva (line 159): ^[Pentti Kanerva, Finnish-American cognitive scientist at the Swedish Institute of Computer Science whose mathematical framework for memory has found unexpected relevance in modern AI.]
7. Emily M. Bender (line 100): ^[Emily M. Bender, computational linguist at the University of Washington, co-author of the influential 2021 "stochastic parrots" paper.]
8. Eric B. Baum (line 116): ^[Eric B. Baum, machine learning researcher whose *What Is Thought?* (2004) argued that the brain is essentially a compression engine.]

Sidenotes (~4)

1. sn:chiang-response (around line 96-98, where Somers discusses Chiang's "Blurry JPEG" article): Note the direct dialogue between these two readings. Somers accepts Chiang's compression metaphor but argues it actually supports AI understanding rather than undermining it. Students should consider which framing they find more convincing.

2. sn:genome-analogy (around line 300-311, the Human Genome Project comparison): Somers draws a parallel between the hype cycle around genomics and AI. The HGP didn't cure cancer, but Crick wasn't wrong about discovering the "secret of life." The implication: AI may not deliver on messianic promises, but the core insight about neural networks and intelligence could still be profound.

3. sn:hofstadter-conversion (around line 179-183, Hofstadter's change of heart about AI): Hofstadter spent decades as the leading AI skeptic, making his conversion especially significant. His final quote about wanting creativity to "remain a mystery" reveals the emotional stakes — understanding the mechanism of thought might diminish what makes us feel human.

4. sn:sparse-distributed-memory (around line 205-207, connection between Kanerva's work and Transformer architecture): The claim that the Transformer's math approximates Kanerva's 1988 model is remarkable — it suggests the AI community independently rediscovered a model of human memory. This connection between neuroscience theory and engineering practice is one of Somers' strongest arguments.

External Links (~4)

1. Add authorUrl to frontmatter for Somers
2. Link "Gödel, Escher, Bach" to Wikipedia
3. Link "Sparse Distributed Memory" book to Wikipedia
4. Link "On the Biology of a Large Language Model" to the Anthropic research page

Verification

1. Run bun run dev (or verify it's already running)
2. Navigate to each reading page and verify:
   • Footnotes render as superscript markers with popup content
   • Sidenotes appear in the margin when clicking highlighted text
   • External links open in new tabs
   • No rendering errors or broken MDX
3. Spot-check that annotation tone is informative, not heavy-handed

If you need specific details from before exiting plan mode (like exact code snippets, error messages, or content you generated), read the full transcript at: ~/.claude/projects/-Users-challen-www-usingandunderstanding-ai/81c10e04-df4b-44a0-a195-49e486217b62.jsonl

If this plan can be broken down into multiple independent tasks, consider using the TeamCreate tool to create a team and parallelize the work.

Your task

First, run git status, git diff, and git log --oneline -10 to understand the current state.

Then add and commit all modified files with a descriptive message. Break the changes into multiple commits if appropriate. If there are untracked files, ask if they should be added or added to the appropriate .gitignore file. The user may also want to take over at this point, so provide an abort option as well. Always check changes for unnecessary debugging statements or logging that might have been accidentally included. Prompt the user if you find any, and offer to remove each one by one. Do not use interactive git staging (git add -p) as it does not work.

Can you find a good quote from this article to add? https://nadh.in/blog/code-is-cheap/ It's a favorite of mine. Or did we already add one?

This is my favorite, but let's condense a bit, maybe removing the pip install bit: "An experienced developer who can talk well, that is, imagine, articulate, define problem statements, architect and engineer, has a massive advantage over someone who cannot, more disproportionately than ever. Knowledge of specific language, syntax, and frameworks is no longer a bottleneck. The physiological constraints of yore are no longer impediments. The machinery for instantly creating code at scale is now a commodity and available to everyone, just a pip install equivalent away. It requires no special training, no new language or framework to learn, and has practically no entry barriers—just good old critical thinking and foundational human skills, and competence to run the machinery."

No, let's keep as much of the original as possible, don't add dashes and stuff.

Your task

First, run git status, git diff, and git log --oneline -10 to understand the current state.

Then add and commit all modified files with a descriptive message. Break the changes into multiple commits if appropriate. If there are untracked files, ask if they should be added or added to the appropriate .gitignore file. The user may also want to take over at this point, so provide an abort option as well. Always check changes for unnecessary debugging statements or logging that might have been accidentally included. Prompt the user if you find any, and offer to remove each one by one. Do not use interactive git staging (git add -p) as it does not work.

Your task

First, run git status, git diff, and git log --oneline -10 to understand the current state.

Then add and commit all modified files with a descriptive message. Break the changes into multiple commits if appropriate. If there are untracked files, ask if they should be added or added to the appropriate .gitignore file. The user may also want to take over at this point, so provide an abort option as well. Always check changes for unnecessary debugging statements or logging that might have been accidentally included. Prompt the user if you find any, and offer to remove each one by one. Do not use interactive git staging (git add -p) as it does not work.