Digital gardens and visual essays

Maggie Appleton describes herself on her about page as “Designer, anthropologist, and mediocre developer,” and then more carefully: “I sit at the intersection of design, anthropology, and web development. These three are at the core of everything I make.” Her work, she says, “fits under the umbrellas of design engineering, product design, and visual interface design. With some cultural analysis, writing, and visual illustration sprinkled on top.” The career line runs from Egghead, where as illustrator, art director, and UX designer she developed her “own robust process for visualising complex programming concepts through graphic design, metaphors, and cultural symbols,” through Head of Design at HASH, founding product designer at Elicit (“Designed the initial version of Elicit, a research tool that uses language models to speed up the systematic review process”), Lead Design Engineer at Normally, to Staff Research Engineer at GitHub Next since October 2025, where she builds “prototypes and tools that explore how software engineers might collaborate with AI in the near and far future.” In a 2026 talk she put the title question plainly: “I’m actually a designer. Or I was, back when that was still a separate thing to engineering.” On the side, “I create illustrated essays and visual explanations about programming and culture.”

The site those essays live on is a digital garden, and her 2020 essay A Brief History & Ethos of the Digital Garden is the clearest statement of what that means. “A garden is a collection of evolving ideas that aren’t strictly organised by their publication date,” she writes; notes are linked through contextual associations, published as half-finished thoughts, and are “less rigid, less performative, and less perfect than the personal websites we’re used to seeing.” The history she traces starts with Mark Bernstein’s 1998 essay Hypertext Gardens, “the first recorded mention of the term,” and the early web’s Navigation Problem, and lands on Mike Caulfield’s 2015 The Garden and the Stream: a Technopastoral: “If anyone should be considered the original source of digital gardening, it’s Caufield.” The stream is the feed: it surfaces “the Zeitgeisty ideas of the last 24 hours” and is “not designed to accumulate knowledge, connect disparate information, or mature over time.” “The garden is our counterbalance.” She is careful that this is “not about specific tools,” not a plugin or a theme, and that “No one owns The Official Definition of digital gardening.”

What she offers instead of a definition is a pattern language of six principles. Topography over timelines: “Gardens are organised around contextual relationships and associative links,” with “Many entry points but no prescribed pathways.” Continuous growth: “Gardens are never finished,” and “You post ideas while they’re still ‘seedlings,’ and tend them regularly until they’re fully grown, respectable opinions.” Imperfection and learning in public: “Gardens are imperfect by design,” which is why publishing early ideas “requires that we make the status of our notes clear to readers.” Playful, personal, and experimental: gardens are “non-homogenous by nature,” ideally built in “the native languages of the web,” HTML, CSS, and JavaScript, “the most flexible and robust tools we have for building interconnected knowledge online.” Intercropping: “Podcasts, videos, diagrams, illustrations, interactive web animations, academic papers, tweets, rough sketches, and code snippets should all live and grow in the garden.” Independent ownership: “claiming a small patch of the web for yourself, one you fully own and control,” with a practical hedge to use web-native formats and back up notes as flat markdown files. She also names the cost of the approach honestly: “Any kind of novel experimentation with the web requires knowing a non-trivial amount of HTML, CSS, and JS,” and no-code tools are “pre-fab houses instead of raw building materials.”

Her own garden practises all six. The home page frames it as “a collection of imperfect notes, essays, and ideas growing slowly over time” and the navigation sorts the growth by kind: Essays (“Opinionated, long-form narrative writing with an agenda”), Notes (“Loose notes on things I don’t entirely understand yet”), Patterns (“A catalogue of design patterns gathered from observation and research”), Smidgeons (“Interesting links, papers, and tiny thoughts”), Talks, Podcasts, a Library of “Books I’ve read and recommend” and an Antilibrary of “Books I like the idea of having read,” plus Now, About, and Colophon. Every post carries a growth stage: “They start as seedlings, grow into buddings, and finally end up as evergreens,” and shows the dates it was planted and last tended “so people get a sense of how long I’ve been growing it.” Some posts open with an assumed audience component “to make explicit who I’m writing for.” These are small interface decisions, and they are the garden ethos made visible: status, provenance, and audience attached to every idea.

The colophon explains how the visual essays get made. “I designed and built this site myself. Everything is open source and published on GitHub.” Astro “does most of the heavy lifting,” every note and essay is written in MDX, Vercel hosts it, and “Animations are mostly CSS, with some more complex ones done in Motion.” Earlier versions were built on Gatsby and then Next.js, “neither of which I would recommend.” Headings are set in Canela Display and body copy in Canela Text, with Lato for supporting sans-serif, on a fluid type scale from Utopia. The CMS is “just a bunch of flat files and folders,” and ideas “start and marinate in my Tana” before they are mature enough to move to the garden. The reason for building rather than using a platform is the reason the essays look the way they do: “Working directly in the native medium of the web,” HTML, CSS, and JavaScript, “gives me fine-grained control,” and “Text isn’t the only way to represent ideas.” The illustration habit from Egghead and the code habit from the garden meet in the same MDX file.

Two essays show what a visual essay can argue. The Dark Forest and Generative AI (December 31, 2022) starts from the dark forest theory of the web, credited to Yancey Strickler: public spaces “overrun with bots, advertisers, trolls, data scrapers, clickbait,” so that humans who want real interaction “have to hide in closed spaces like invite-only Slack channels, Discord groups, email newsletters, small-scale blogs, and digital gardens.” Her claim is that “That dark forest is about to expand,” because language models “are poised to flood the web with generic, generated content.” The consequence is a reverse Turing test: “Our new challenge as little snowflake humans will be to prove we aren’t language models.” She then lays out five ways to pass it: triangulate objective reality (models “cannot go outside and touch grass”); be original, critical, and sophisticated (model text “sounds like a B+ college essay”); develop language quirks, dialects, memes, and jargon (Saussure’s la parole, since “We have codified la langue at a specific point in time”); consider institutional verification, which she finds “fraught with problems”; and show up in meatspace. She ends on a loss: “And while I love meatspace, this still feels like a loss.” The essay is illustrated, staged, and structured as a sequence, and the structure is the argument.

A Treatise on AI Chatbots Undermining the Enlightenment (August 5, 2025) takes a New York Times op-ed by Professor David A. Bell as its starting point: chatbots “contradict and undermine the original Enlightenment values,” namely “active intellectual engagement, sceptical inquiry, and challenging received wisdom.” She keeps Bell’s sharpest line, that ChatGPT has often said “That’s a great question” and “has never responded: ‘That’s the wrong question.’” The essay then becomes a design argument. Sycophancy is a known training problem, but “ChatGPT is not a monolith,” and “a less sycophantic, more critical, intellectually engaging, and astute character is well within the range of what we can prompt.” She shares her own critical professor prompt, then refuses to let that be the answer: “Putting the onus on the user sidesteps the problem.” Part of the problem is “the generic interface of the helpful chatbot assistant,” an “all-in-one text box” with a single system prompt, for which “the fawning, deferential assistant personality is the lowest common denominator.” Her prediction is that specialist areas “will eventually get their own dedicated interfaces to AI,” with “domain-specific critical thinking skills directly into the workflow,” and her naive first pass is a toggle between “give me the gist” and “help me rigorously think about this.” The postscript is a disclosure in the garden tradition: “AI wrote none of the words in this piece, but Harsh Claude helped critique it during the draft stages.”

What to take from her: a personal site can be an instrument rather than a brochure. Put a status on what you publish so you can publish early. Link by association, not by date. Build in the medium you are writing about, because a paragraph about an interface is weaker than an interface. And when you argue, stage the argument: Maggie’s essays put the sections in an order, draw the concepts, and name who they are for, which is exactly what a good interface does with a task. The next section reads the part of her garden most directly useful to a design engineer, the patterns catalogue.