惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

G
Google Developers Blog
博客园 - 司徒正美
Last Week in AI
Last Week in AI
Recent Announcements
Recent Announcements
Y
Y Combinator Blog
博客园 - 聂微东
M
MIT News - Artificial intelligence
博客园_首页
Jina AI
Jina AI
博客园 - 叶小钗
酷 壳 – CoolShell
酷 壳 – CoolShell
H
Hackread – Cybersecurity News, Data Breaches, AI and More
J
Java Code Geeks
F
Fortinet All Blogs
aimingoo的专栏
aimingoo的专栏
小众软件
小众软件
Vercel News
Vercel News
The Cloudflare Blog
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
云风的 BLOG
云风的 BLOG
N
Netflix TechBlog - Medium
B
Blog
Google DeepMind News
Google DeepMind News
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More

Hacker News - Newest: "AI"

AI can't read an investor deck AI as an attorney? Student uses ChatGPT, Gemini to sue UW over alleged racial discrimination Hacking MCP Servers in AI Systems – The Rug Pull: Tool Changes After Approval GitHub - MeepCastana/KubeezCut: Free Web based video editor Can AI judge journalism? A Thiel-backed startup says yes, even if it risks chilling whistleblowers Coming soon: 10 Things That Matter in AI Right Now DARPA built an AI to fact-check enemy weapons claims What explains heterogeneity in AI adoption? When AI Meets Muscle: Context-Aware Electrical Stimulation Promises a New Way to Guide Human Movements - Department of Computer Science AI Changed How We Build. It Did Not Change What Matters. Linux rules on using AI-generated code - Copilot is OK, but humans must take 'full responsibility for the… Meta spins up AI version of Mark Zuckerberg to engage with employees Code Mode: Let Your AI Write Programs, Not Just Call Tools | TanStack Blog GitHub - Delavalom/graft: Go framework for building AI agents. Type-safe tools, multi-provider (OpenAI, Anthropic, Gemini, Bedrock), zero vendor SDKs. India's TCS tops estimates, says new AI models did not dent services demand Gen Z's fading AI hype Strong feeling: we are in a folded AI reality GitHub - machinarii/total-recall-catalog: A reference catalog of latest knowledge retrieval, memory & RAG systems GitHub - mensfeld/code-on-incus: Give each AI agent its own isolated machine with root, Docker, and systemd. Active defense detects and stops threats automatically.. Quantization, LoRA, and the 8% Problem: Benchmarking Local LLMs for Production AI Iran war: We spoke to the man making Lego-style AI videos that experts say are powerful propaganda Powell, Bessent discussed Anthropic's Mythos AI cyber threat with major U.S. banks GitHub - immartian/bellamem: Persistent belief-graph memory for AI agents. Retrieves decisive context by importance — not recency, not RAG, not /compact. recursive-mode: The Repo-Native Operating System for AI Engineering After the attack on Sam Altman's home, will AI CEO's go on the offensive? The biggest advance in AI since the LLM Opus 4.6 vs GPT 5.4 One Prompt Unity World Generation Test “AI polls” are fake polls Client Challenge Can AI be a 'child of God'? Inside Anthropic's meeting with Christian leaders
The Wholesale Plagiarism of Obscure Sorrows - Waxy.org
Andy Baio · 2026-06-21 · via Hacker News - Newest: "AI"

Last week, a MetaFilter member posted a link to what appeared to be a new website for The Dictionary of Obscure Sorrows, John Koenig’s decade-long project to make a “dictionary of made-up words for emotions that we all feel but don’t have the words to express.”

The polished site includes everything you’d expect from a publisher’s promotional book site: an author biography, press mentions, and links to buy the book on Amazon.

Strangely, it also includes the entire text of the book, from its opening 800-word foreword to a complete archive of all 311 neologisms, with their accompanying definitions, etymology, and short essays, all penned by Koenig.

The book’s original photo-collage illustrations made by Koenig and several other artists are conspicuously missing. Instead, each word has an AI-generated image made with DALL-E 2, riddled with the errors and artifacts typical of that model.

“it’s half-past IŊΨ-o-clock”

A banner at the top of the homepage encourages visitors to “Generate your own words using AI – give your sorrows a voice!” The Submit A Sorrow feature lets you describe a feeling, and then uses OpenAI’s GPT-4 to generate the new word, etymology, and definition, which go into a gallery of “User-Generated Sorrows” with AI generated art.

MetaFilter members were immediately suspicious, and so was I. My wife Ami and I made a card game in 2022, Lost for Words, partly inspired by Koenig’s project. We own a copy of the book, and I’d followed it online for years. The embrace of AI seemed out of character.

Then I noticed the new site was a different domain than the original Tumblr homepage entirely:

The original: dictionaryofobscuresorrows.com
The reboot: thedictionaryofobscuresorrows.com

What’s going on here?

A Little History

John Koenig launched The Dictionary of Obscure Sorrows on Tumblr in 2009, expanding it to a series of popular video essays in 2013.

If you know any word from the project, it’s probably “sonder,” which spread far beyond its origin, making its way into common parlance and eventually to Dictionary.com and Merriam-Webster.

sonder
n. the realization that each random passerby is living a life as vivid and complex as your own—populated with their own ambitions, friends, routines, worries and inherited craziness—an epic story that continues invisibly around you like an anthill sprawling deep underground, with elaborate passageways to thousands of other lives that you’ll never know existed, in which you might appear only once, as an extra sipping coffee in the background, as a blur of traffic passing on the highway, as a lighted window at dusk.

Other words coined by Koenig have found a life outside his project. You may have encountered “anemoia” (a feeling of nostalgia for a time or place you’ve never known), “vellichor” (the strange wistfulness of used bookstores), or maybe “monachopsis” (the subtle but persistent feeling of being out of place).

But “sonder” is the breakaway success. I’d wager most people who have heard the word have no idea it was coined by a guy on Tumblr in 2012.

There’s an R&B band named Sonder, a failed Airbnb rival, and countless businesses ranging from consultancies and VC firms to coffeehouses and dispensaries. There’s a bar named Sonder two miles from me right now.

Photo from the official Instagram announcing the book’s release

That success landed Koenig a book deal with Simon & Schuster, and the book became a New York Times bestseller on its release in November 2021.

Two years later, around August 2023, the new Dictionary of Obscure Sorrows website launched, but curiously, with no reference to it from the official Tumblr page or social media.

A Slick Impostor

The mission of Koenig’s project, in his own words, is to “shine a light on the fundamental strangeness of being a human being.”

So it felt strange that he would now be encouraging people to generate new words and definitions with LLMs, a contentious technology that has been trained on so much human writing, but can’t know what it’s like to be human.

I reached out to John Koenig directly to ask if he was involved with the website. He emailed back an hour later:

Yeah man, I had nothing to do with it. Don’t know what to think or do about that, as the site is pretty slick. Nicer than my own, really.

It wasn’t hard to find who was responsible since they list themselves in the “Site Credits” in the footer of every page: Qontour (formerly Prompt Digital), a web design and marketing agency based in San Francisco.

The only hint that the site isn’t authorized is this page in their portfolio, where they talk about how “Qontour built the interactive digital platform – designing the site in Webflow, generating an AI-powered image library, and launching a feature that lets visitors submit their own sorrows and add new definitions to the dictionary.”

On that page, they refer to themselves as “fans” of the book: “The site gives fans (like us) one place to find everything – videos, reviews, interviews, and purchase links – instead of searching across a dozen platforms.‍”

The problem, of course, is that being a fan doesn’t give them the right to repurpose any of the material for their site.

Copyright and Confusion

In the footer of Qontour’s unauthorized site, they added a copyright notice acknowledging that they don’t own any of the rights to the material on the site, while also licensing all the user-submitted words into the public domain with a CC Zero license.

Dictionary Content © John Koenig – All rights reserved.
User-Generated Content open licensed – CC Zero.

This betrays a fundamental misunderstanding of how copyright works. Qontour did not have the right to publish the entirety of Koenig’s book to showcase their web design skills.

They also submitted their site to Webflow’s directory to advertise their design business. “This endeavor showcased our expertise in website design, AI-generated content, and extensive content integration.”

Below the button to “Hire Qontour,” a small link to “Copyright Info” misrepresents their work:

The Dictionary of Obscure Sorrows by Qontour is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. All Rights Reserved. In other words, it’s someone else’s work so you can’t copy it or edit it for any reason, but you can share it with others.

Needless to say, you can’t relicense content you don’t own.

Complicating their claims of it being a fan tribute, Qontour also used their own Amazon affiliate code throughout the site, created under their previous name Prompt Digital, giving them a cut of all book sales.

Those commissions may have been meaningful over the last few years, since the unofficial site is now the top search result for virtually every query related to the book, including the book’s title, the words coined in the book, and even John Koenig’s name. In every Google search I’ve tried, the unofficial site ranks higher than the official site, the publisher’s site, or Wikipedia.

This is made worse by the rapid shift from traditional web search to conversational AI search, which is easy to manipulate, hides sources, and collapses context into simple answers.

ChatGPT and Gemini both link to the bootleg as the official website, and both claim that John Koenig is the one that created it.

Gemini (left) and ChatGPT (right)

This creates legitimate confusion over its authorship, and arguably, damages the reputation of the project and book with its enthusiastic embrace of AI. The person who originally posted the site to MetaFilter thought it was the official site, and the commenters in the thread then, reasonably, questioned whether the book itself was written by AI.

I asked Koenig if his publisher was planning to issue a cease-and-desist takedown to the site, but didn’t receive a response.

After emailing him, I realized that Simon & Schuster did make moves last year to limit its reach. They filed two DMCA takedowns (1, 2) with Google last July, asking them to remove two pages from the bootleg site from their results. It had no effect.

AI and Consent

It’s one thing for a fan to share or remix copyrighted material out of love for the source material, with no commercial motive. (“No copyright intended!”) It’s another for a marketing agency to take an entire living author’s book, replace its art with AI slop, add an AI word generator, monetize the traffic, promote it in their portfolio, and then outrank the official site everywhere.

This is a more flagrant form of plagiarism than you typically see these days, where human-authored works are laundered with an AI model into something that’s different enough from its sources to avoid legal issues.

But it’s not surprising to see it coming from an agency that has leaned into generative AI so heavily. As they proudly explain, “Every page on this site was written in Claude” using an “author persona” that they call “Q.”

What’s missing here is consent, which feels like the original sin of AI. As I’ve written about many times before, generative AI models are all trained on a massive corpus of human-authored works without attribution, consent, or compensation, extracting value from creators while centralizing power among a tiny handful of massive tech companies.

On a much smaller scale, Qontour could have reached out to John Koenig for permission to republish his work, collaborating with him on a new, improved website for the book. He might have asked them to limit it to just the words published on his Tumblr, asked for them not to build AI features, or maybe just said no to the whole thing, which would be his right.

The Last Word

What happened to The Dictionary of Obscure Sorrows may have been more brazen, but it isn’t an isolated case.

It’s part of a broad trend happening across the web, where people are using AI to repackage, optimize, and replace the authoritative sources it was trained on for profit.

Nearly every day, I get emailed a newly-launched, obviously-vibecoded website filled with AI-generated content that was designed to siphon attention away from human creators: bloggers, authors, journalists, artists, musicians, and anyone else who slowly, painstakingly makes things for a living. I’m not even sure anymore that the emails I’m receiving are sent by a human.

The feeling of seeing something you love ingested and repurposed by a machine designed to replace the person who made it seems like a uniquely modern sorrow.

Maybe there should be a word for it.


You can purchase John Koenig’s The Dictionary of Obscure Sorrows at Powell’s Books, directly from his publisher, or your local indie bookstore. If you have to use Amazon, you can buy it using the author’s own affiliate code so he gets the largest cut of the sale.

Photo via John Koenig