今日は眠い

[Anshu @anshuc]

https://t.co/Pj7qOLxE4E

[Hugging Models @HuggingModels]

Ever wished your AI could spot faces or objects with surgical precision? Meet Adetailer, a model that brings sharp detection to images using Ultralytics and PyTorch. It's like giving your vision AI laser focus.

[Hamza Khalid @humzaakhalid]

THIS IS KARPATHY'S OBSIDIAN VAULT, THE BRAIN BEHIND A MILLION-DOLLAR COMPANY

This is every idea, every decision, and every connection one person has made over years of work, visualized in real time.

The human brain processes up to 11,000,000 bits of information per second, but consciously uses only 50.

Inside this network:

→ thousands of nodes
→ hundreds of active links between ideas
→ 140+ decision-making processes per hour of work
→ years of compressed thinking living inside a single system

make sure to bookmark it before it gets lost in your feed

below I wrote a guide on how you can build yours for FREE (2 mins setup)

[YanXbt @IBuzovskyi]

HERMES AGENT JUST GOT A RESEARCH BRAIN.

connect it to NotebookLM and it becomes
a 247 research system that builds
knowledge over time.

it resources. connects ideas across sources.
uses that context in every real task.

setup in 4 steps:

1. install Hermes with MCP enabled
2. download NotebookLM skill from GitHub
3. add endpoint to MCP servers in config
4. restart Hermes

what your agent can now do:

→ query your notebooks instead of the open web
→ connect ideas across multiple sources
→ build knowledge session after session
→ use your curated context in every /goal
→ zero hallucinations from your own verified sources

ask it anything — it answers from
what YOU already know.

full setup guide — MCP, models, use cases —
in the article 👇

[Akshay 🚀 @akshay_pachaar]

Hermes agent masterclass.

In this video, I cover everything you need to understand and customize Hermes Agent. Self-evolving skills, three-tier memory, GEPA optimization, and going from 1 to 10 agents that work for you 24/7.

Enjoy!

Chapters:

00:00 - Intro
02:03 - How to get the most out of this video
02:32 - What we're building (and why it's wild)
07:11 - How the whole thing works under the hood
09:27 - The SOUL.md: your agent's personality file
11:15 - The 3-tier memory system that keeps it all together
14:16 - Skills: what your agent can actually do
16:49 - The self-evolving loop (agents that improve themselves)
19:58 - The curator: Hermes' built-in garbage collector
22:56 - GEPA optimization: making your agent sharper
25:08 - Installation and setup
27:38 - Connecting your agent to Telegram
30:36 - Configuring programmer with Claude Code
31:53 - Adding new skills (from a hub of ready-made skills)
34:59 - Going from 1 to 10 agent profiles
36:49 - Building a custom designer from scratch
40:42 - Anatomy of the .hermes folder (where everything lives)
45:05 - Skill taps: sharing skills via a GitHub repo
45:59 - Skill bundles: stacking skills for workflows
47:19 - Hermes Kanban (coming soon)
48:05 Outro

Cheers! :)

[Prajwal Tomar @PrajwalTomar_]

I still don't think people understand what just happened with Claude Design.

Claude Design can now take you from rough idea to fully responsive, high-fidelity website design in 10 minutes.

Most builders are skipping Steps 1-3 and blaming the tool when results look generic.

This 4-step workflow gets you professional designs on the first pass:

(full breakdown in the article)

[MERICA MEMED @Mericamemed]

A KEUREGG for eggs in the morning. Lazy but low key want it.

[huangserva @servasyy_ai]

微软发布了一个 4B模型，可以在 3 秒内把任意图像转换为 3D 资产。

它使用一种名为 O‑Voxel 的新几何格式，并且可以在 CUDA 上用不到 100 毫秒转换为带纹理的网格模型。

输出为 GLB 文件，包含完整的 PBR 材质，可直接用于 Blender、Unity 和 Unreal。

完全开源👇

[⧫ Alan Chen ⧫(alanchentsla.eth） @alanchen]

英伟达是下一个“硬盘公司”？24 岁投资天才：芯片定价权即将崩盘！

[Elon Musk @elonmusk]

Try it out!

(Partially trained on Colossus 2)

[AYi @AYi_AInotes]

Damn，有点炸裂啊，一颗葡萄用了6660张宏观照片训练3DGS，

模型只有43万个高斯点，
压缩完不到50MB，

但你可以在手机上看它360度旋转，
连果肉里透出来的光都还原了，

更炸的是老哥回复里那句话，有人问底部怎么拍到的，他甩了张gif说“Like this”，

但玩笑归玩笑，真正重要的信息其实都藏在技术参数里：
90个焦点堆栈，每个74张，
全角度覆盖，底部也拍进去了，

半透明物体是3DGS最难搞的测试对象，他偏偏选了最难的那个，
逻辑也很简单：最难的通了，比它简单的任何东西就已经在射程之内了，

我理解这颗葡萄的本质不是炫技，而是一份写在技术参数里的产品化路线图，
下一个被数字化的可能不是葡萄，可以是任何你舍不得让它烂掉的东西

逻辑很简单，如果连葡萄这种最难的都跑通了，比它简单的任何东西，手办、珠宝、人脸、文物，都已经在射程之内，

以前做这种级别的3D数字资产，需要专业扫描仪加建模师加后期手动修，成本和时间劝退绝大多数人，

现在这套流程是，宏观镜头加照片堆，进3DGS自动训练，出来就是20到50MB的可分发模型，

门槛从专业团队降到了一个人加一台相机，这是我觉得真正炸的地方，

Dany自己还在开发miqula工具，明显在把这套流程进一步封装成产品，
他用一颗葡萄，相当于给整个行业画了一张产品化路线图。

[选肉体还是灵魂？ @XRTHSLH]

有没有人用过这个鞋子？舒服吗

[古一 @MANISH1027512]

超近特写美颜肖像，半身构图，面部占据画面大部分，纯白无缝影棚背景，高调影棚灯光，柔和漫射正面光，明亮通透氛围，干净商业美容摄影，韩式美容编辑风格，精致年轻东亚女孩，黑色短发波波头，整齐齐刘海，大而闪亮的紫灰色眼睛，精致眼线，柔和下睫毛，小巧鼻子，光泽淡粉唇，瓷白肌肤，光滑细腻肤质，柔和腮红，眼中明亮眼神光，几缕被风吹动的发丝轻拂脸庞，纤细脖颈与锁骨，白色吊带背心带子仅微微可见，极简构图，梦幻干净白色留白，极致精修，柔和光晕，美容广告美学，空灵，清新，纯真，高细节，锐利眼神，柔和肌肤，高级影棚肖像.

[今日は眠い @Below0s]

@AISuperDomain 你好， 有电报。wx 联系吗？ 有问题想确认谢谢

[Kshitij Mishra | AI & Tech @DAIEvolutionHub]

10 GitHub repos that keep earning while you sleep:

1. AutoHedge

An autonomous hedge fund in Python.

4 AI agents run the entire pipeline: strategy, validation, risk sizing, execution.

Live on Solana. Coinbase next.

pip install -U autohedge and you're running.

https://t.co/EYqZICtlRZ

2. Vibe-Trading

64 finance skills. 29 specialist swarm teams. Full quant toolkit.

DAG-based multi-agent system where agents debate and hand off in real time while you watch the reasoning stream.

HK/US equities and crypto data are completely free.

https://t.co/UfCHsN1fAr

3. Claude Ads

190 audit checks across Google, Meta, YouTube, LinkedIn, TikTok, and Microsoft Ads.

6 parallel subagents tear through your entire ad stack in one command.

This is what I was paying $4,000/month for.

https://t.co/lnljunDx6B

4. Toprank

Pulls your Search Console data and tells you exactly why your rankings are broken.

Type "implement everything" and every fix gets staged in 3 minutes.

This is what $3,000/month used to buy.

https://t.co/mkTUtCqwos

5. Fincept Terminal

The open-source Bloomberg killer.

20+ investor AI agents. CFA Level 1, 2, 3 analytics. 100+ data connectors. Local LLM support so your positions never leave your machine.

Bloomberg charges $24,000/year for less.

https://t.co/CZvl5bAd63

6. Agentic Inbox

Cloudflare just open sourced an AI email client that drafts every reply automatically and never sends anything without your approval.

Runs entirely on Workers. Zero third-party servers touching your inbox.

One click deploys the whole thing.

https://t.co/LLRoslkkCL

7. ClawRouter

Scores every request across 14 dimensions in under 1ms and routes it to the cheapest model that can actually handle it.

Cuts AI inference costs by 78%.

https://t.co/6bxUdlhqsQ

8. Camofox Browser

Makes AI agents invisible to bot detection at the C++ engine level.

Not patches on top of Chrome. A genuinely unmodified Firefox fork.

13 search macros built in for Google, Reddit, LinkedIn, TikTok, and Instagram.

https://t.co/8ea5cjnSt1

9. Open Higgsfield AI

200+ AI models in one interface. Flux, Sora, Kling, Veo, GPT-4o, SDXL.

Text to image. Image to video. Cinema mode with pro camera controls.

I cancelled my Higgsfield subscription the same day I found this.

https://t.co/Hgs7rmtjmp

10. Hyperframes

HeyGen just open sourced the Remotion killer.

The agent writes HTML. The framework renders MP4. No React. No JSX. No build step.

Same HTML always produces the same file.

https://t.co/BZX6g7tLga

Save this for later.

[えこ @eko_love7425]

くさい足好きでしょ？

[Jingyuan缘缘 @Jingyuan_521]

我这两天被这个爆火的Flipbook刷屏并被震撼到，现在的UI界面也许会在不久的将来彻底改变。Flipbook是三星工程师(前Open AI)@zan2434和另外两个伙伴做的一个无限视觉浏览器，根据你的搜索词，实时用像素生成一张类似百科全书里的动态插画，连文字也是像素组成，没有代码，没有HTML，画面上的任何区域你都可以点击，生成这块区域的进一步画面。
而且作者们调制的这个画面风格太好看了，我小时候要是有这玩意能玩一整天不吃不喝，完全就是一本无限魔法书。
作者直接展示了在这上面搜索巴黎旅游景点，还直接买了巴黎圣母院的票。我自己也用来搜索了清迈，效果惊人，但可惜这两天太火了我一直没能连接上视频动态效果。虽然目前的算力无法实现规模化，作者也说目前只是展示测试阶段，但是我这次是真真切切感受到了AI带来的改革，未来就快来了。

[Capacho @Capachoamachule]

https://t.co/yi5yimnX7H

[AYi @AYi_AInotes]

很多人总说AI没用，我觉得那是因为你光刷不用，真正能赚钱的人早就闷声发大财了。

分享一个用 AI 做副业的路子，
有点野，基本上算零门槛，见钱快(｀・ω・´)💰

为啥说零门槛：
首先不用囤货不用开店不用露脸，一台电脑就能干，成本低到可以忽略不计。

就用Rita平台不用排队不降智的Seedance2.0，加上目前最强的GPT-image-2，批量生成AI美女跳舞视频，这个主题任何平台都是流量密码。

发小红书抖音还有支付宝视频号，这三个平台现在都有巨额创作者扶持，光流量分成就能覆盖所有token成本，还能剩不少。

等粉丝起来了，内衣美妆品牌方抢着找你投商单，因为你报价比真人网红便宜十倍，他们根本不挑。

这个赛道还在红利期，有执行力的赶紧去试吧💪

提示词：
画幅：9:16竖屏，时长15秒，5个镜头，纯欲高级感，合规无违禁

---
镜1｜0-3s 开场眼神杀（脸部特写）
竖屏9:16，15秒连贯，24fps，8K超高清，真人写实电影质感，画面稳定不闪烁。顶级颜值东亚美女，精致清透妆容，大波浪卷发，皮肤细腻通透，眼神魅惑紧盯镜头。MENTE MA经典起手手势，双手从下巴轻抬至额头，动作丝滑卡点。柔和环形柔光，面部光影干净，眼神光明显，镜头轻微匀速推进，主体居中，第一人称互动感，色彩高级，细节拉满。

---
镜2｜3-6s 上半身卡点（中近景）
竖屏9:16，24fps，8K超高清，真人写实。完美身材比例，肩颈线条优越，修身黑色显瘦吊带， MENTE MA手势舞，手部交叉波浪动作，手臂随节奏流畅摆动，身体轻晃卡点，眼神持续看向镜头。暖调氛围光，侧光勾勒轮廓，镜头缓慢横移，动作连贯不穿模，主体居中。

---
镜3｜6-9s 全身展示（全景）
竖屏9:16，24fps，8K超高清，真人写实电影质感。顶级腰臀比，修长线条，修身长裤，体态优雅。MENTE MA完整手势循环，腰胯轻摆卡点，动作舒展自然。极简干净背景，柔和顶光+侧光，光影层次强，镜头匀速轻微拉远，动作丝滑无卡顿，人物不畸变。

---
镜4｜9-12s 手部特写（细节）
竖屏9:16，24fps，8K超高清，真人写实。手指修长精致，MENTE MA指尖定点手势，动作干净利落。手臂轻抬轻晃，氛围感拉满。柔光特写，浅景深，镜头轻微跟随手势，焦点清晰，画面高级不低俗，动作流畅。

---
镜5｜12-15s 收尾定格（上半身特写）
竖屏9:16，24fps，8K超高清，真人写实电影质感。顶级颜值美女，MENTE MA收尾定型手势，双手在胸前定点，眼神魅惑，浅笑收尾。镜头快速匀速推近，画面稳定，主体居中，最终定格，高级性感，氛围感拉满。

---
负面提示词（必加）
画面闪烁，动作卡顿，穿模，畸形手指，多余肢体，五官扭曲，崩脸，画质模糊，噪点，镜头晃动，主体出框，表情僵硬，低俗暴露，水印，字幕，多人，穿搭跳变，发型跳变，背景杂乱，画面畸变，色彩失真