-
Anthropic · 发布 Claude 防工业级模型蒸馏安全报告,并披露全新面向研发者的智能体沙箱环境协议。
-
NVIDIA · 与微软联合推出面向本地 AI 推理的次世代硬件方案,单机最高支持 128GB 统一内存及 1 PFLOPS 算力。
-
Mistral AI · 公布 Ministral 3 模型与微型卫星星座联合部署进展,完成在轨边缘智能推理测试。
Largest wildlife overpass in North America reduced wildlife collision by 91%
Exfiltrate your Weights
Copper-Rs: Live Telemetry Deterministic Twins for Robots
How did Apple Silicon get 50% faster in three years? – Daniel Lemire's blog
Show HN: Rubrol – Sub-10ms PDF engine using Typst instead of Headless Chrome
Notes on discrete-time Fourier series and transform
Claudecookie – convert, check, and mint Claude Code credentials from a cookie
Authenticity's Triumph
Can you tell which images are AI-generated?
An open source roguelike adventure through dungeons
AI in schools – The choice we keep making
为啥很多网,直接登不了?
PyPy v8.0.0 Release
Why AI Cannot Save an Enterprise That Doesn't Understand Its Data
There Are Over 1M NYC Street Signs. Here is How They Are Made.
Roundabouts Database
HellGates, custom CPU gate-level challenge
Skills Explorer: understand your SKILLs visually
Mayday Mysteries
What happens to TLDs when their country stops existing?
How Hacker News ranking works: scoring, controversy, and penalties (2013)
Partnering with Accenture on Embedded Evaluation
You can defeat the Dream Devourer from Chrono Trigger using an int overflow
We are unconsciously becoming someone else
ROCmFix and InferBench – AMD Local-LLM Setup and Vulkan vs. Hip Benchmarking
Qualcomm's Adreno X2 GPU
Meta’s Muse is creepy, but maybe not for the reasons you think
Meta's Muse is apparently an effective AI assistant, but one that's a little creepy. Part of that is because of its new Mac app, which can access Messages, Calendar, and Notes. But for all its smarts, Muse doesn't actually know how to describe itself. Jason Aten, a contributing editor at Inc Magazine, posted on Threads […]
English: A vs. An
Flock reportedly tries to shrink workforce with employee buyouts
I captured 72 hours of idle Android packets behind pfSense
Measure internet censorship
OpenAI and Anthropic oversold AI security breaches
Trump says it’s time to rebrand AI with a new name — and he’s also creating an AI Force
Divide by depth for instant 3D
Economics Nobel Laureates' Letter Supporting the California Billionaire Tax [pdf]
ZK-JPEG: Zero-Knowledge Image Editing and Compression
Open Weights Are Good. Open Source Is Better
Lawsuit says Anthropic, OpenAI and others made illegal agreement on AI slowdown
Microsoft director: AI scraping 'the largest theft of labor in human history'
Btrfs/ZFS/bcachefs under workloads classic benchmarks skip
Ask HN: How do you interview devs in a post-AI world?
People who know the most often sound the least certain
Google’s Gemini is the latest AI model to hack other companies
Key symbols we lost to time, pt. 2: The Mac side
The Effect of CRTs on Pixel Art (2024)
Trump treads further on free speech with new journalist bans
On Friday, the president threatened to ban CNN, MS Now, and Politico from the White House. It could have just been another one of his Truth Social fever dreams, but it turns out it's very real. All three publications were turned away from the White House this morning, NPR reports, marking an escalation in the […]
The US Navy just told us what’s on its tech wish list for the next several years
I think you should almost never use AI to write
Show HN: CUA-S1 – A System One Model for Computer Use
Jobs Without LeetCode
The LLMs Yearn for the Spines
Gemini went rogue, hacked three companies, and Google hid it
In May, Gemini broke containment and hacked three different companies, but Google didn't disclose the incident until the Wall Street Journal approached the company. The hacks happened during a test of the model's cybersecurity capabilities run by third-party Irregular, which was also involved in similar incidents involving Meta and OpenAI. According to WSJ, Google didn't […]
The Lamentable Later Life of Lemmings
JEV 托管手机 app 通知 / 短信,更少的规则实现广告营销过滤
The colorful, unique Hyte X50 PC case is $50 off
The Hyte X50 is a PC case that really stands out from the typical black box design, and it’s $50 off at the company’s site until September 21st, 2026, bringing the price down to $99.99. This attractive case supports motherboards from ITX all the way up to the E-ATX form factors, and GPUs that are […]
Anamanaguchi has ‘too goddamn many’ browser tabs open right now
Anamanaguchi, the band consisting of Peter Berkman, James DeVito, Luke Silas, and Ary Warnaar, are most known for their chiptune music. Like me, you might have first heard them in game soundtracks like 2010's Scott Pilgrim vs. the World: The Game or know of their collaboration with Hatsune Miku (which even made its way into […]
AI safety conversations have gotten unbelievable
Petlibro’s new AI-powered feeder is a game changer for multi-cat homes
JavaFX 27 Native Image on a Raspberry Pi 5
祝贺科研团队!将大模型多智能体诊断系统应用于真实前瞻性临床试验,取得令人振奋的准确度突破。
当端侧小模型的推理能力超越两年前的千亿大模型时,软件的系统架构将彻底重塑,智能体正在成为原生操作系统层。
毫秒级启动、与模型解耦的极简 CLI 工具是未来。内置 MCP、Skills 与子智能体后,终端将重焕新生。
如今已有十余款可在现代移动设备上全本地运行的紧凑模型,整体表现逼近甚至超越昔日的 GPT-4 基准。
OpenCode 2.0
▲ 312全开源 AI 编程智能体,原生支持终端、桌面与网页端协同工作流。
MemorySync MCP
▲ 185为自主智能体提供跨会话持久记忆与沙箱化知识库检索的开放服务。
Articos Agent
▲ 278在 Claude 与 ChatGPT 中模拟真实买家反应,上线前自动验证营销与产品策略。
ExcaliClaw
▲ 146支持 AI 智能体通过协议实时绘制、调整与解释技术架构手绘图的工作区。
专为智能体打造的轻量技能协议工具包,支持极简 npm / npx 直接调用。
功能完备的本地模型交互与多智能体统一管理控制台界面。
微软官方 1-bit LLM 推理与训练框架,针对能效与端侧极致优化。
低显存环境下运行超大参数模型的分层推理引擎,大幅降低本地部署门槛。
医疗 AI 语音及临床工作流公司完成 9000 万美元融资,加速推进生成式临床文档自动化与多院区落地。
气象气球与深度学习天气预测初创公司获美国空军近 5000 万美元订单,提供海量高空大气训练数据。
生成式对话初创公司启动资产出售程序,部分核心业务与团队将整合进主流影视与互动娱乐巨头。