For years, John Koenig’s The Dictionary of Obscure Sorrows was a sanctuary for the internet’s unnamable feelings. He gave us sonder—the realization that everyone has a life as vivid as your own—and onism—the frustration of being stuck in one body. But as recent reports have highlighted, these deeply personal neologisms have been caught in a corporate dragnet, harvested by AI models and faceless entities without a shred of credit.
The Ghost in the Machine
The controversy reached a boiling point with the discovery of a "new" Dictionary of Obscure Sorrows website that appeared to be a hollowed-out shell of the original. As reported by Andy Baio on Waxy.org, this site launched in late 2023 but curiously lacked any connection to Koenig himself. It wasn’t just a fan site; it was a signal of a larger, more systemic issue. We are witnessing the wholesale commercialization of internet-native language.
Large Language Models (LLMs) are being trained on these emotional taxonomies, effectively "laundering" Koenig’s creative labor into generic training data. These companies are strip-mining the human soul for tokens, turning a decade of poetic observation into a product they can sell back to us as "creative assistance."
The Death of the Digital Artisan
This isn’t just about one man’s vocabulary; it’s about the erosion of the "digital artisan." When a corporate entity or an AI scraper ingests a community-driven project, the nuance is stripped away. The words become commodities. As bloggers and creators have noted, this "theft of services" is becoming rampant across the web.
We are moving toward a future where the very words we use to describe our inner lives are owned by the same companies that provide our cloud storage. The systemic harvesting of these emotional frameworks suggests that nothing—not even the way we feel—is exempt from the bottom line.
A Future Without Context
If we lose the link between a word and its creator, we lose the human context that made the word meaningful in the first place. As we move further into the AI era, we have to ask: if our most obscure sorrows are up for grabs, what part of our digital identity is actually safe? We may find ourselves in a world where we have the words to describe our pain, but no memory of who gave them to us.
Sources
Media
