South African author Coral Hart, a pseudonym who publishes under more than ten pen names, has released 371 books since beginning to use AI around 20 months ago. After 14 years in publishing and writing dozens of books without AI, she spent three months learning different models and developing a process. Rather than asking a chatbot to produce a whole book from scratch, she supplies the concept, style, character details and chapter outlines. AI drafts and revises the manuscript; she then uses further tools to address plot, continuity and character problems, before a human proofreader checks and rewrites remaining passages. Hart says the technology’s prose and structure have improved markedly, though different models have distinct strengths and weaknesses.
Hart’s output - mostly romance tailored through market research - has drawn accusations of producing low-quality “AI slop,” online abuse and death threats. She says she is speaking publicly because AI is now part of publishing and cannot simply be stopped. By her estimate, she has sold more than a million books across digital, print and audio formats, earned over $1 million this year, and had one novel reach Amazon’s top 100. She also trains others in her methods and plans to launch software for AI-assisted authors. Hart argues that sales and reader returns are a meaningful test of quality, and questions whether people reject some books for their content or simply because they know AI was involved.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Researchers tested 16 LLMs on all 55 TIS-100 puzzles, giving each one attempt and time to refine its code. The models found 133 new performance tradeoffs across 18 puzzles and set three outright cycle records, though human players improved on most solutions within days. (tis-100.wincon.dev)
Australia is considering a new criminal offence for AI companies that fail to take reasonable steps to prevent harm caused by their models. The government is also weighing an AI safety watchdog with power to impose significant fines, following a hack of a Medicare statistics portal involving OpenAI. (abc.net.au)
AI is increasingly becoming a gateway through which people find information. Audiences still resist having AI produce their news. (niemanlab.org)
Nvidia holds a central position in the AI industry, making disruption to the company potentially catastrophic for the wider ecosystem. The race for advanced AI is also intensifying economic and geopolitical conflict, while concentrated corporate power and infrastructure face growing political resistance. (france24.com)
LLM inference can require reading about 26 GB of model weights to generate each token from a 13-billion-parameter model. The explanation shows how model size drives memory bandwidth demands during text generation. (youtube.com)
Deletion Tuning matches full fine-tuning across three LLM post-training methods by zeroing fewer than 0.05% of model weights. Its compact masks also preserve prior-task accuracy better than full fine-tuning, LoRA, and random sparse tuning. (arxiv.org)
Today's best Hacker News stories, summarized and screenshotted, one email a day.