DZ

这几个月高密度的连续使用了 ClaudeCode, Cursor, AugmentCode, Cursor, Windsurf, Trae, Codex 之后。我得出 Coding Agent 的能力计算公式 = Agent x LLM 总的来说 ClaudeCode 比 Codex 强太多了，虽然最近很多人狂吹 因为Codex 和 ClaudeCode say good bye. 但是....

Amazing reverse perspective illusion

The latest 𝕏 algorithm has been published to GitHub github.com/xai-org/x-algo…

三年前 reddit 上还在讨论 3 tps 属于可运行 Local LLM。今年MoE 在 12G VRAM 已经能达到 40+ tps , Dense 已经达到 6+ tps 了。

@adahstwt debian

which Linux distro would you recommend to someone using Linux for the first time? 1) Ubuntu 2) Arch 3) Debian 4) Mint 5) Fedora 6) Manjaro 7) Kali 8) something else

@LotusDecoder 预热大脑

AI激活前额叶假说。 AI输出的内容，逻辑和理性成分含量越高，类似于人类前额叶激活状态下说的话、写的字越多。 那么读者的前额叶激活程度越高。 那么这个假说有什么现实用处呢？ 就是工作或竞技之前，先和 AI聊，把前额叶给激活打开， 类似于早起一杯咖啡。

@0xPrajwal_ RedHat

What was your first Linux distro? - Fedora - Ubuntu - Arch Linux - Debian - Kali Linux And what are you using now?

@Surendar__05 copilot is not a model

USA has ChatGPT USA has Grok USA has Claude USA has Gemini USA has Copilot China has DeepSeek China has Qwen China has GLM China has Kimi China has MiniMax What is the rest of the world even doing??

@droidbuilds windows OS

Bought a brand new second-hand Windows laptop😭 What should I uninstall first?

@LmhdbAi 除了自带雷人属性，还有什么可比的魅力？

雷军这么大个企业家，为了跟马斯克合照，一脸谄媚太丢份了 国际场合还是要看 Jack Ma ，中国的企业家只有马云能能做到全程平视从容、不怯场 面对普京、特朗普、默克尔、内塔尼亚胡、莫迪、克林顿等各国政要，都能侃侃而谈，不愧是天生的演说家

@KaiXCreator Except gemini

What’s your go-to AI model today? - Gemini 3.1 - Opus 4.7 - Sonnet 4.6 - Codex - GPT 5.5

OpenShell v0.0.41 🧩 agent-driven policy management 🎚️ sandbox resource flags in the CLI 🔒 custom CA support for OIDC TLS verification 📥 sandbox downloads with workspace-boundary checks 🔧 bug fixes and stability improvements Policy and resource control, directly from the shell. github.com/NVIDIA/OpenShe…

如果你有以下条件，每天可以赚500美元： 1. 一台笔记本电脑 2. Wi-Fi网络 3. 时间 这是10个Claude提示词，能让你每天赚钱：

@jahirsheikh8 Paied for ultra,got fewer usage limit. It is cheat

Feels like Google is about to shut down AntiGravity. I mean… are there even any developers maintaining it? 😭

@Samaytwt Cause it like a joke

Why is no one talking about Anti-Gravity anymore??? Did Claude and Codex just take over the whole market???? Anti-Gravity literally disappeared like it never existed 😭

@KaiXCreator hard now

Be Honest: Can you CODE without AI...?

@stableAPY Right, 27b at 24 tps with very low precision

@prodbitz I did not even tried it on the 3060 thing is, MoE are ok with CPU offloading, but dense model really don't work the same and it'll be pretty bad I think

I still can't get my head around the fact that my 3060 12gb is running Qwen 3.6 35B at 40 tok/s this card is around 200$ second-hand while everyone's shilling super expensive 128gb unified memory or RTX 6000 cards, one 3060 12gb can be far from enough for first local AI experiments it's cheap, and paired with a bit of RAM and a somehow decent CPU it'll do the work of course decode drops when context grows and you won't be able to run multiple sub agents, but it's a nice cheap entry point for example, it pairs very well with my 3090: > 3090 running main agent 35B -np 2 => so I can have 2 concurrent agent > 3060 sub-agent 35B -np1 this way my main hermes can delegate work to this sub agent while working on something else I also run Hermes cron job so they does not overload the main agent and I don't mind it behind slower because it happens in the background

@tokerumaisurii 说那话的人 也许正是烟雾弹

求你们不要再说LLM毁了Linux、LLM踩头了Linux和Nginx这样的话了。你们有没有想过一件事情，这些漏洞毫无疑问早就被一些国家安全机关、间谍组织、取证与监控软件供应商之类的捏在手里了，不爆出来永远是0-day。反而每天报每天修，让那些垃圾人垃圾组织失去先发优势，你的数字生活才是越来越平等安全了。

We present ZAYA1-8B-Diffusion-Preview, the first diffusion language model trained on @AMD. Autoregressive LLMs generate one token at a time; diffusion generates a block in parallel, speeding up inference. We show a 4.6-7.7x decoding speedup with minimal quality degradation 🧵

I bought a GEEKOM A9 Max AI. • AMD Ryzen AI 9 HX 370 • 32GB RAM • 1TB SSD I installed @OmarchyLinux on it. Beautiful mini computer. Small, quiet, and powerful. There are many different, good options out there. You can get a really good setup for under $1,000. RAM and SSD prices are still a bit crazy, though, so make sure whatever you buy is upgradeable.