Promethevs is an open-source Windows engine designed to run DeepSeek-V4.1-Flash-1017 using all 510 GB of its original weights on a PC with 64 GB of RAM and a 12 GB NVIDIA GPU. It aims to make a model too large to load conventionally usable on consumer hardware, while keeping prompts local and offering an OpenAI-compatible server for apps and coding agents. Its “compress-then-correct” method answers with a 2-bit copy of the model’s experts, then uses the model’s confidence to identify uncertain words and recomputes those with exact weights reconstructed from correction data. A draft head proposes multiple words in parallel, and caching helps avoid repeating some work in follow-up turns.
On the project’s Ryzen 9 3950X and RTX 3060 test PC, generation reached 5.8 tokens per second on a cold code prompt and 7.4 in a warm-cache best case, versus 2.14 for exact decoding. Across 960 audited words, the engine agreed with the exact path on 98.33% when all differences were counted; the uncorrected 2-bit copy scored 97.4%. The warm code result is not presented as typical: cache retention improved code speed but not reasoning or agent prompts, and the best-case runs did not require a correction to overturn a word. Installation requires Windows 10 or 11, 64 GB RAM, a supported 12 GB RTX card, about 1.1 TB of free NVMe storage, and roughly four hours to build the stores after downloading the original weights.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Richard Sutton argues that brute-force computation alone cannot produce intelligence and that AI needs clever new methods. He says the science of AI is far from finished. (twitter.com)
Anthropic is accused of charging Russian Claude users before identity checks, then cutting off access after verification and retaining biometric data. Critics say account bans can affect linked GitHub projects and Russian citizens living abroad, and accuse the company of failing to answer public questions. (twitter.com)
LLM coding agents boost workplace productivity but can prevent developers from learning through hands-on problem-solving and weaken their technical skills. Developers should preserve their expertise by building projects manually, studying fundamentals, and practicing without AI tools in their own time. (ducktape.blot.im)
A fabricated Newtype interview led The Diplomat to retract its article about a Korean startup’s AI war test in Ukraine. The retraction followed a diplomatic dispute. (thediplomat.com)
Microsoft Research’s AI Frontiers group brought together about 50 researchers to develop agent systems and foundational AI advances. After the team joined Microsoft AI in September, its research-first approach clashed with leadership’s emphasis on scaling. (hunch.net)
Microsoft researcher John Langford has left the company. Microsoft’s AI Frontiers group has been declared finished. (businessexaminer.ca)
Today's best Hacker News stories, summarized and screenshotted, one email a day.