Every story across every category, newest first. Each card links to the original publisher; daily-brief posts open as editorial pages.
The funniest thing about the "deepseek harness paper" is that not once does it mention DeepSeek or the DS harness. No evals, no nothing. It's a…
Article URL: https://medium.com/airbnb-engineering/eval-driven-development-lessons-from-evaluating-genai-at-scale-e817e5ae5788 Comments URL: https://news.ycombinator.com/item?id=49290785 Points: 6 # Comments: 0
Article URL: https://cdn.openai.com/pdf/how-organizations-use-chatgpt.pdf Comments URL: https://news.ycombinator.com/item?id=49290768 Points: 4 # Comments: 0
OpenAI is launching a preview of a sped up version of its latest, most powerful model, in an effort to court enterprise users.
IBM plans to train and certify tens of thousands of consultants on OpenAI's technologies as part of this deal.
RT DogeDesigner𝕏 is the most transparent platform on the internet.I was able to see if my account or posts were being suppressed, understand the exact reason…
GrokAravind Srinivas: Congrats to @SpaceXAI on one more amazing model: Grok 4.6. We benchmarked it as an orchestrator on our Wide-And-Deep-Research benchmark using the Perplexity Computer…
Just watched Wendell's (level1 techs) latest video on the HP Z8 Fury desktop workstation and was curious how you could configure it. And oh boy, there's…
The world evolves following its dynamics, i.e., its laws of motion. However, leading video diffusion models largely fit the pixels without modeling how the pixels transit…
RT NVIDIA AI InfrastructureHappy 10th Birthday to #NVIDIADGX! 🎉Over the last decade, DGX has transformed from a single system into the essential blueprint for the AI…
There's no training anywhere in this. I ported Doom's actual rendering algorithm into transformer weights using a compiler I wrote (torchwright) -- every weight computed, none…
Status: Broadly an explainer for the two Acts, along with some analysisOn July 23, 2026, there were two different bills introduced in Congress that provide a…
I managed to get Qwen3.8-2.4T-A95B running locally with llama.cpp on mu PC just for fun, cause why not. I was using the Unsloth Qwen3.8-2.4T-A95B-UD-Q1_0 GGUF quantization.…
All other sources of energy combined are utterly insignificant compared to the SunLuis Batalha: Yesterday’s eclipse stopped roughly 10²⁰ J of solar energy from reaching Earth.That’s…
Article URL: https://www.geoffreylitt.com/2026/07/02/understanding-is-the-new-bottleneck Comments URL: https://news.ycombinator.com/item?id=49290299 Points: 19 # Comments: 5
metal : add TQ2_0 support (#26980) metal: add TQ2_0 support Add support for the GGML_TYPE_TQ2_0 (ternary, 2 bits per element) type in the Metal backend. Assisted-by:…
Google shipped Gemini 3.7 Flash just three weeks after 3.6 Flash. The new model is supposed to be Google's most capable workhorse yet for coding and…
Article URL: https://github.com/systemd/systemd/issues/40262 Comments URL: https://news.ycombinator.com/item?id=49290215 Points: 20 # Comments: 3
Grok 4.6 is now available in Perplexity and Perplexity Computer.On WANDR, it sits on the Pareto frontier of performance and efficiency, matching Fable 5 results at…
Article URL: https://xenon-app.com/ Comments URL: https://news.ycombinator.com/item?id=49290180 Points: 7 # Comments: 0
Article URL: https://themech.net/2026/08/hello-me-its-been-a-while/ Comments URL: https://news.ycombinator.com/item?id=49290166 Points: 13 # Comments: 5
It really is. The mission is advancement of civilization and propagation of consciousness to the stars.AleXandra Merz 🇺🇲: What is Elon's motivation
A lot of people hit their usage limits fast, so we’re providing more free usageGrok: We've reset limits to help you keep building during the Grok…
RT Keith Coleman 🌱😀🙌You know how everyone started seeing their friends in the timeline a few weeks ago?With our new open-source code base, you'd also have…