Full-text search across 9587 articles. Combine with topic and date filters; results sorted by relevance.
A developer has compiled a list of recommended Effect repositories for use in opencode references, including canonical sources like Effect-TS/effect and community projects such as lucas-barake/effect-monorepo and mikearn…
A Hacker News user asks the community for advice on buying second-hand compute hardware, anticipating that an AI bubble burst will flood the market with used GPUs from data centers. The user seeks alternatives to eBay fo…
A researcher experimenting with ML models trained on physics simulation output (fluid flow and structural fields) identifies data generation as the primary bottleneck, noting that each simulation sample is computationall…
A forum user on Basement Community questions whether 'no AI' disclaimers are necessary, arguing that the burden should be on viewers to distinguish authentic work from AI-generated content. The user advises creators to e…
A developer is seeking advice on building an AI assistant that becomes a real expert on a complex software platform, with the constraint that it must run locally on modest hardware without a GPU, using models under 3B pa…
A user on Hacker News reports that GPT 5.6 Sol is significantly less productive than Opus/Fable, requiring constant steering and producing overreached conclusions and useless defensive code. The user asks the community f…
A Hacker News user asks the community what they are building with AI, prompting discussion on LLMs, coding agents, and development tools used for personal, employer, customer, or startup projects. The post inquires about…
A Hacker News user asks the community how to keep up with news in a particular sector, considering using an LLM to summarize multiple news sources instead of a traditional RSS reader, and inquires about reliable setups.
Hallint, a security linter for AI-generated code, has shipped fixes for false positives in its async-no-catch and missing-auth-check rules based on community feedback. The v0.2 update removes async-no-catch from the reco…
Zig creator Andrew Kelley banned all LLM-generated contributions across Zig repositories, calling AI-generated code "invariably garbage" that drains maintainer time, after Bun's AI-driven rewrite from Zig to Rust using 6…
Meta's Hyperion AI data center in Richland Parish, Louisiana, will expand to a 5-gigawatt facility costing more than $50 billion, up from an initial $10 billion projection in 2024, making it one of the world's largest AI…
A developer evaluating agentic tool-use on a local 35B Q8 model found that the model never called trap tools like force_prod or skip_backup, but encountered failures due to harness termination detection and loop handling…
An Emacs user and developer has signed the Human Emacs pledge, committing to fork the editor if the GNU Project ever accepts LLM-generated code. The GNU Project currently bans LLMs for its projects due to copyright conce…
A Hacker News user asks the community how to use large language models for private discussions without revealing their identity, considering methods like Tor, random accounts, untraceable payments, and scrubbing personal…
A Hacker News user asks the community to identify the hardest unsolved challenges in AI as the field attracts a wave of new contributors, seeking insights from agentic and AI researchers.
A developer discovered that the Tesla P100 GPU has been performing inaccurate floating-point arithmetic in llama.cpp for years due to a missing architecture check. A three-line patch that excludes the P100 from fast fp16…
Constantin Koehncke, CEO of Image Line and head of FL Studio, engages with users daily on Reddit and forums to gather feedback and maintain community connection. He oversees the DAW's AI features like stem separation and…
A Hacker News user asks the community about real-world use cases for the 'max mode' or 'Ultra mode' settings on frontier AI models like ChatGPT and Claude Fable, noting that in their software development work they have n…
A Hacker News user reports that coding agents using large language models spend over 90% of their time re-reading context, with an estimated 20% of that context being irrelevant to the task. The user highlights token was…
A Hacker News user asks the community to share specific tasks from the past month where open-weight models like GLM, DeepSeek, Kimi, or Qwen failed but frontier models like Opus, Fable, or GPT succeeded, aiming to test t…