Show HN: Skyline Co-op is like SimTower, but built around democratic socialism
I've been rebuilding old games with Claude. While creating a modern take on SimTower and having election news on TV in the background, I had the…
Latest
Final optimization: from ~10 tok/s to ~15 tok/s on DeepSeek-V4-Flash-0731 at 128K ctx – 1 RTX 3090
J'ai consacré beaucoup de temps à l'optimisation de DeepSeek-V4-Flash-0731 GGUF sur une seule RTX 3090. Mon exigence absolue pour chaque configuration était la suivante : Le…
Show HN: demake – one source project compiles to any retro game console ROM
This started life as a graphics tool for a specific problem: gen AI can make retro-styled sprites but can't follow exact hardware constraints of pixels and…
How Americans View Capitalism, Socialism and Free Enterprise
Article URL: https://news.gallup.com/poll/713144/americans-view-capitalism-socialism-free-enterprise.aspx Comments URL: https://news.ycombinator.com/item?id=49195244 Points: 11 # Comments: 4
Mario Meets Pareto
Article URL: https://www.mayerowitz.io/blog/mario-meets-pareto Comments URL: https://news.ycombinator.com/item?id=49195231 Points: 28 # Comments: 3
AI isn’t enough to protect social media communities from AI
Why humans need to moderate humans.
OpenAI developer warns the "tireless eagle eyes of a million models" are coming for your exposed API keys and crypto wallets
OpenAI developer "roon" warns on X that AI models could soon start scanning for exposed API keys, crypto wallets, and login credentials at scale. His warning…
Self-Evolving Coding Agents
Large language models are increasingly embedded in software engineering workflows as coding agents that can inspect repositories, invoke tools, execute tests, debug failures, and generate patches.…
Improving GPT-5.6 Sol in ChatGPT—and expanding access for free users
ChatGPT introduces improved GPT-5.6 Sol with better accuracy and consistency, plus expanded access for free users and unlimited everyday chats with GPT-5.6 Luna.
Matryoshka NLAs: training activation verbalizers to frontload reconstruction-relevant information
TLDR: We train a “matryoshka” NLA that, unlike standard NLAs, is trained to put the most important details at the start; it is trained by randomly…
Why You Should Almost Never Use AI to Write Anything Substantive
I think you should almost never use AI to write -- that is, to do the thing you're doing when you type words on a page…
Best llama cpp flags to run Deepseek-flash 0731
Hi all. These are my system specs: dual xeon e5 2696 v2 , 160gb DDR3 ram ECC(1600mhz), 3 gpus: 3060 12gb, p100 16gb, 3050 6gb. And…
Don’t forget to train and use your brain. Literally.
This is a screenshot from a recent Theo.gg video (Apple rant; tldr: he is realizing the lockdownness after...years.) and the fact he asked ChatGPT for an…
A quick snapshot of where Qwen3.8-Max stands today: Qwen3.8-Max now ranks #5 on the Artificial Analysis Intelligence Index, and #1 on the Agentic Inde…
A quick snapshot of where Qwen3.8-Max stands today:Qwen3.8-Max now ranks #5 on the Artificial Analysis Intelligence Index, and #1 on the Agentic Index!🥇 We'll keep pushing…
OpenAI says Apple’s trade secrets lawsuit is ‘rotten to its core’
OpenAI has asked a federal judge to toss out Apple's landmark lawsuit accusing the ChatGPT maker of stealing trade secrets, describing the allegations as "meritless." In…
b10293
ci : onboard AMD ROCm CI with gfx1151 fixes (#26544) ci: prepare for amd rocm ci Signed-off-by: Aaron Teo aaron.teo1@ibm.com ci: fix editorconfig-checker Signed-off-by: Aaron Teo…
ABSeeker: Training Long-Horizon Search Agents via Answer-Backtracked Credit Assignment
Long-horizon search agents must make multiple sequential actions (steps) to search, retrieve, verify, and integrate evidence to reach a final answer. However, existing methods for training…
FocusMem: Factorizing Content, Readout, and Trust in Latent GUI Memory
GUI agents must remember both useful experience from earlier tasks and unfinished progress in the current interaction. Latent memory offers a compact solution by compressing multimodal…
Prime Intellect Releases Prime Agent: An Open-Source RLM Harness Where Sub-Agents Are Function Calls Inside Persistent IPython Kernel
Prime Intellect has open-sourced Prime Agent, a coding and research harness built on two abstractions: the Recursive Language Model, which turns sub-agent calls into functions inside…
RT Hugging Apps: every 3D mesh is kind of a compiled binary: printable, usable, but frozen. you can't just "make this hole 2mm wider" Cadena is just o…
RT Hugging Appsevery 3D mesh is kind of a compiled binary: printable, usable, but frozen. you can't just "make this hole 2mm wider"Cadena is just out…
RT John David Pressman: Just so we're clear I think AI will, on the default trajectory, become wildly more capable than humans, invent basically every…
RT John David PressmanJust so we're clear I think AI will, on the default trajectory, become wildly more capable than humans, invent basically everything there is…
Reward is hyperstitional information
(This article broadly explains mirror ascent, continuous Bayesian inference and information geometry in full. Title refers to the result in section 3.) The logarithmic scoring rule…
b10291
vulkan: fix submission batching size, add debug tools for diagnosing causes of DeviceLost drivers errors (#26371) vulkan: add debug tooling to get more information about a…
RT Zephyr: They don't have enough inference capacity to fulfil demand, so they are trying to raise prices and destroy demand One of the reasons why I'…
RT ZephyrThey don't have enough inference capacity to fulfil demand, so they are trying to raise prices and destroy demandOne of the reasons why I'm not…