Human Empowerment in an AI Society
Thanks to Benjamin Schmidt and Luke McNally for the discussions around this topic and their feedback on an early draft of this article!Gradual disempowerment is one…
Full feed
Feed 28,103 posts
Every story across every category, newest first. Each card links to the original publisher; daily-brief posts open as editorial pages.
Beauty GLM-Image was trained on Ascend 910s, and now runs on Apple
BeautyGLM-Image was trained on Ascend 910s, and now runs on AppleMLBoy_DaisukeMajima: Ported GLM-Image (ZhipuAI, 16B) to Core AI — runs locally on Apple Silicon. The image…
it just works
it just worksAbdur Rahim: DeepSeek's DSpark speculative-decoding drafters, benchmarked on a Mac 👇Native on Apple Silicon (MLX), lossless; identical output to the base model, ~1.4–1.6× faster…
We urgently need a Fable class model that's not from this shit ass company. I don't care if it's bad at bioweapons and hacking the Three Gorges Dam, I…
We urgently need a Fable class model that's not from this shit ass company. I don't care if it's bad at bioweapons and hacking the Three…
Decoding the obfuscated bash script on a Uniqlo t-shirt
Article URL: https://tris.sherliker.net/blog/obfuscated-self-evaluating-bash-script-by-cdn-akamai-being-supplied-to-consumers-via-retail-stores/ Comments URL: https://news.ycombinator.com/item?id=48829312 Points: 42 # Comments: 7
> – "sentient" We need something more integrated than J-space, even at the cost of altering models. I want to see how J-space concepts interrelate. Is…
> - "sentient"We need something more integrated than J-space, even at the cost of altering models. I want to see how J-space concepts interrelate. Is being…
How to Evaluate an Enterprise Analytics Platform
Most enterprise analytics evaluations are really just dashboard comparisons. That's...
Geosql: A Claude/Codex skill for geospatial data
Article URL: https://github.com/dekart-xyz/geosql Comments URL: https://news.ycombinator.com/item?id=48829242 Points: 17 # Comments: 0
Pdf to JSON, 3 months in.
Hello all, it has been 3 months since I made the initial post, where I wanted ideas to try out. The subreddit has been amazing with…
RT Steeve Morin: We're releasing ZML/LLMD, our homegrown LLM server built on top of our homegrown high performance heterogeneous inference stack. It s…
RT Steeve MorinWe're releasing ZML/LLMD, our homegrown LLM server built on top of our homegrown high performance heterogeneous inference stack.It ships with 5 architectures out of…
RT ZML: http://x.com/i/article/2074768591900098560
RT ZMLhttp://x.com/i/article/2074768591900098560
The mosquito bucket of doom works
The mosquito bucket of doom is a population control mechanism where you dissolve some Bti (Bacillus thuringiensis israelensis) into a bucket and allow the mosquitoes to…
Literature Review: LLM Inference at the Edge: Mobile, NPU, and GPU Performance Efficiency Trade-offs Under Sustained Load | Bnechmarking LLMs on Phones [R]
Just finished reading the paper: LLM Inference at the Edge: Mobile, NPU, and GPU Performance Efficiency Trade-offs Under Sustained Load I am starting to benchmark LLMs…
From Foundation to Application: Improving VLA Models in Practice
Despite recent progress of VLA foundation models, the disparity between laboratory conditions and real-world applications continues to impede their practical implementation. To bridge this gap, we…
Image2Sim: Scaling Embodied Navigation via Generative Neural Simulator
Embodied navigation aims to build agents that interpret multimodal goals, reason in 3D space, and reach target destinations reliably in the real world. However, progress remains…
PluraMath: Extending Mathematical Reasoning Evaluation Beyond High-Resource Languages
Mathematical reasoning has become a central task for evaluating and tuning reasoning Large Language Models (LLMs), yet existing benchmarks remain heavily biased toward high-resource languages, with…
OpenAI's GPT-5.6 launches Thursday after a delay forced by the U.S. government
OpenAI is launching GPT-5.6 on Thursday after the U.S. government lifted the initial release ban following additional testing. According to OpenAI, the Sol model outperforms Anthropic's…
Hot French startup ZML releases free product to speed inference across lots of AI chips
ZML, a hot French AI startup endorsed by Turing Award winner Yann LeCun, has now released ZML/LLMD, software that could make running AI less costly.
Gemma 4 Technical Report
We introduce Gemma 4, a new generation of open-weight, natively multimodal language models in the Gemma model family. Designed to advance compute efficiency and reasoning, the…
RT Wauplin: Downloading Xet-backed files just got faster: huggingface_hub v1.22.0 skips per-file HEAD /resolve requests and rebuilds metadata from the…
RT WauplinDownloading Xet-backed files just got faster: huggingface_hub v1.22.0 skips per-file HEAD /resolve requests and rebuilds metadata from the cached commit file tree. For repos with…
b9912
spec : fix naming, spacing (#25410) macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFramework Linux: Ubuntu x64…
RT Machine Learning Street Talk: This paper looks great, note that heavyweights @kenneth0stanley, Bert Chan (Lenia) and @SakanaAILabs have had their h…
RT Machine Learning Street TalkThis paper looks great, note that heavyweights @kenneth0stanley, Bert Chan (Lenia) and @SakanaAILabs have had their hands all over it!Ciaran@ICML🇰🇷: New paper…
How do you fight the Opencode becoming a politician?
Me: you have that tool, use it. Opencode: no I don't see that tool in the context, I'll do it the retarded way. Me: *opening an…
RT Gandalv: For the past year, MAGA has been screaming “Europoor” like a cheap car alarm going off at three in the morning. Nobody asked for it, nob…
RT GandalvFor the past year, MAGA has been screaming “Europoor” like a cheap car alarm going off at three in the morning. Nobody asked for it,…