Dave Eggers, the prolific author and founder of McSweeney's literary magazine, told OpenAI staff that ChatGPT risks silencing an entire generation of writers. Sam Altman invited Eggers to speak to roughly 200 OpenAI employees last year, drawing on his decades of experience in fiction, journalism, screenwriting, and arts advocacy.
Eggers has built a career championing emerging writers through McSweeney's and multiple educational nonprofits. His warning to OpenAI staff reflects growing concerns within the creative community about AI language models trained on vast amounts of copyrighted work. The core issue remains unresolved: ChatGPT and similar systems learn patterns from books, articles, and essays without explicit permission or compensation to authors.
The tension centers on a fundamental question about AI training data. Copyright holders argue their work should not feed commercial AI systems without consent or payment. OpenAI maintains its training practices fall within fair use doctrine, a legal position still being tested in multiple lawsuits filed by authors and publishers.
Eggers' message carries particular weight. He represents a generation of writers who built careers through traditional publishing and independent media ventures. His concern suggests that young authors entering the field now face a different landscape: their potential audience can receive AI-generated content for free, instantly, and without human effort. This undercuts the economic model that historically supported professional writers.
OpenAI invited Eggers specifically because his perspective matters to their employees. The company operates in an uncomfortable position. It needs to convince both the creative community and its own staff that AI development serves rather than destroys human creativity. Yet the business model depends on training systems on human-created content at massive scale.
The conversation reflects a broader reckoning happening across tech. As AI moves from research labs into everyday products, companies face pressure to address legitimate concerns from creators whose work trained their systems. Whether