S1-mini-GGUF by Superwhisper is a compact, purpose-built text normalizer for post-processing speech-to-text transcripts. It rewrites raw ASR output into clean written text by removing fillers, resolving false starts and self-corrections to the final spoken choice, applying punctuation and capitalization, and converting spoken entities (numbers, dates, times, currency, email addresses) into written form. English-only and not a conversational/chat model, it is steered via a required control line and system prompt that specify styling, structure, and context. Intended for integration into dictation or transcription apps (check the included s1-mini-license), it runs comfortably on a laptop CPU and achieved 94.8% token accuracy on a held-out set of 7,519 English cases.
Two GGUF builds are provided: a Q4_K_M quantized build (recommended) at 462 MiB and an F16 unquantized build at 1.4 GB. Internally it uses a qwen3-style skeleton with 311 tensors, 28 blocks, a 40,960-token context window, and an embedded chat template; quantization keeps 29 tensors at Q6_K while the bulk is Q4_K and normalization params remain F32. The repo includes usage instructions for many runtimes and servers - llama.cpp, vLLM, Docker, Ollama, LM Studio and others - plus deployable examples and notes on files, input format, and evaluation details.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Four coding agents were given a $100 budget to build a functional PDF editor, but their implementations had noticeable bugs and shortcomings. The experiment highlights that current AI agents struggle with goal pursuit and user interaction, leading to subpar software quality. (blog.nielstron.de)
Lithos-ai open-sourced lithos-metal, enabling Qwen3.8-27B to run at over 200 tokens per second on an Apple M5 Max laptop. This allows ultra-fast inference for coding agents with a single command. (twitter.com)
Google Cloud introduces the Gemini agent, a universal AI tool that handles questions, content creation, coding, and knowledge work within familiar applications like Gmail and Docs. It features industry-specific skills, security, cost controls, and integrates seamlessly into existing workflows. (cloud.google.com)
Temporal developed pi-temporal to enable the Pi coding agent to recover from machine failures during tasks by leveraging a fleet of workers and durable execution. This setup allows seamless task continuation across machine failures without repeating interrupted commands. (temporal.io)
![OpenAI cannot make AI safe on its own [pdf]](https://hntoday.pixelshot.io/screenshots/20261008/0217d5d1410143c5_tn.jpeg)
OpenAI has acknowledged that it cannot ensure the safety of artificial intelligence systems on its own. The document discusses the challenges and limitations faced by OpenAI in making AI safe without external collaboration or oversight. (mikitabalesni.com)
Yoshua Bengio urges AI researchers at frontier companies to prioritize safety and consider leaving if safety measures are insufficient. He warns that current development races risk pushing humanity toward catastrophic outcomes and calls for slowing AI progress to prevent loss of control. (transformernews.ai)
Today's best Hacker News stories, summarized and screenshotted, one email a day.