diff --git a/content/en/2025-06/2025-06-26.md b/content/en/2025-06/2025-06-26.md
new file mode 100644
index 0000000..17fce5b
--- /dev/null
+++ b/content/en/2025-06/2025-06-26.md
@@ -0,0 +1,103 @@
+---
+title: 06-26-Daily
+weight: 5
+breadcrumbs: false
+comments: true
+description: Google DeepMind has unveiled Gemini Robotics On-Device, an AI model designed
+ specifically for robots to run locally ๐ค. Based on the multimodal reasoning of the
+ Gemini 2.0 model, it lets robots quickly learn new tasks, work stably even without
+ internet, and even handle intricate operations like fo...
+---
+# AI Insights Daily 2025/6/26
+
+> `AI Daily` | `Updated Daily at 8 AM` | `Comprehensive Data Aggregation` | `Cutting-Edge Science Exploration` | `Open Platform for Industry Voices` | `Open Source Innovation` | `AI and the Future of Humanity` | [Visit Web Version โ๏ธ](https://ai.hubtoday.app/)
+
+#### **AI Content Summary**
+
+```
+AI products are updating fast, with Google launching on-device AI for robots. iFlytek's medical large model hit expert level.
+Quark's college application service is booming, and they're expanding computing power. Rokid Glasses are in mass production, snagging tons of orders.
+AI research is making breakthroughs in multimodal and 3D reconstruction. Zhou Hongyi discussed how AI can't replace human emotion or creativity.
+```
+
+#### **AI Product and Feature Updates**
+
+1. Google DeepMind has unveiled **Gemini Robotics On-Device**, an AI model designed specifically for **robots** to **run locally** ๐ค. Based on the **multimodal reasoning** of the **Gemini 2.0 model**, it lets robots quickly learn new tasks, work stably even without internet, and even handle intricate operations like folding clothes โจ. This is definitely laying a solid foundation and kicking off a new chapter for the future of **embodied AI**!
+
+
+2. College application season is in full swing, and **Quark**'s smart application report service saw such massive demand that it had **queues** of users, generating over 3 million reports to date ๐. This clearly shows how much students trust its AI capabilities. Facing this "sweet problem to have," **Wu Jia, Alibaba Group Vice President**, boldly responded, saying the team has urgently **expanded computing power**, vowing to make sure every student smoothly gets their hands on this crucial guide for higher education! ๐ช
+
+
+3. Rokid (Lingban Technology) and Lens Technology's jointly developed **consumer-grade AI+AR glasses, Rokid Glasses**, have officially hit **mass production**! ๐โจ These glasses, with their **lightweight design** and integrated **AI large model capabilities** like **smart prompting, real-time translation, and AI object recognition**, have already snagged **250,000 global pre-orders**! This signals that the Chinese AI glasses market is about to see a **commercial explosion**, and the future's looking super promising! ๐
+
+
+4. At the 2025 Cloud Next conference, Google showcased its next-gen **customer service intelligent assistant** ๐ค, powered by the **Gemini model**. This assistant is seriously impressive; it not only handles **multimodal interaction** but can also apply for **discounts on its own**, and it's deeply integrated with Salesforce's **CRM system**! This hints at a massive intelligent transformation coming to customer service ๐ฅ, but we'll have to wait and see on its accuracy and privacy protection~ ๐
+
+
+5. iFlytek has made a big splash with the release of **Spark Medical Large Model V2.5 International Edition** ๐, trained entirely on domestically produced computing power! This model topped the charts on the authoritative MedBench platform with a score of 98.4, and its comprehensive diagnostic and treatment capabilities have already reached the level of an attending physician at a top-tier hospital, even surpassing human doctors in completeness, practicality, and readability! ๐จโโ๏ธ๐ฉบ It also supports multiple languages, and is set to make a huge splash in the global medical market, boosting international medical tech exchange and collaboration! ๐โจ
+
+
+6. ElevenLabs has finally launched its standalone **text-to-speech mobile app**! ๐ฑโจ Whether you're on iOS or Android, you can now generate audio snippets anytime, anywhere. Even free users can enjoy about 10 minutes of audio generation time! This app not only uses the latest v3alpha model but also supports **emotional expression control**, and it'll even get speech-to-text and conversational AI tools in the future. How convenient is that?! ๐ฃ๏ธ
+
+
+#### **AI Frontier Research**
+
+1. **SuperDec**, co-launched by teams from ETH Zurich, Stanford University, and Microsoft, is breaking the mold of traditional **3D reconstruction** ๐คฏ! This tech uses an innovative **hypertetrahedra** principle to achieve compact yet vivid **3D scene representations**. It not only efficiently handles complex **point cloud** data but also shows immense potential in precise grasping and path planning for **robotics**, as well as **controllable visual content generation**, opening up new horizons for the digital world! ๐ [Project Address](https://super-dec.github.io/)
+
+2. **4D-LRM** is a super cool and innovative **large-scale spatio-temporal reconstruction model** ๐คฉ. It can fully reconstruct **dynamic objects' 4D representations** (3D space plus time dimension) from just a few viewpoint inputs, allowing for high-quality scene generation from any time and any angle! In the future, it's set to really shine in areas like virtual reality, film production, and industrial simulation! ๐ [Paper Address](https://huggingface.co/papers/2506.18890)
+
+3. ByteDance and Shanghai Jiao Tong University have teamed up to release the **ProtoReasoning framework** ๐. It cleverly leverages structured prototype representations like **Prolog** and **PDDL** to significantly boost **large language models**' **logical reasoning capabilities** and efficiency in cross-domain knowledge transfer ๐. This research lays a solid foundation for future theoretical exploration of reasoning prototypes, which is just awesome! [Paper Address](https://arxiv.org/abs/2506.15211)
+
+4. Hong Kong University MMLab, Chinese University of Hong Kong MMLab, and SenseTime have jointly developed the **GoT-R1 framework**. This groundbreaking research greatly enhances **multimodal large models**' **semantic-spatial reasoning capabilities** in **visual generation tasks** by introducing **reinforcement learning** ๐, allowing the model to independently learn even better reasoning strategies! It not only breaks free from the GoT framework's reliance on templates but also achieved **SOTA performance** in complex scene generationโseriously impressive! โจ [Paper Address](https://www.jiqizhixin.com/articles/2025-06-25-9)
+
+#### **AI Industry Outlook and Social Impact**
+
+1. Zhou Hongyi recently chatted in a video about the future of AI. He believes that no matter how powerful AI gets, it can never fully replace humanity's unique abilities in **emotional understanding** ๐, **complex problem-solving** ๐ง , and **creative thinking** ๐จ. He emphasized that future work will increasingly involve **managing and training** AI, even citing a failed AI customer service case from a Swedish company to show AI's limitations when dealing with complex customer needs. ๐ง
+
+
+2. Federal Judge William Alsup has made a groundbreaking ruling: **Anthropic**'s use of **copyrighted books** to train its **AI models** without permission was deemed **fair use**! ๐ฎ This sets an important precedent for copyright disputes in the AI industry. However, Anthropic still faces **theft charges** for acquiring training materials from pirated websites. Talk about mixed feelings, huh?~ ๐ค
+
+
+#### **Open Source TOP Projects**
+
+1. **Dioxus** is a super popular **full-stack application framework** with 28,310 stars โญ! It's like an all-in-one toolkit, aiming to give developers a unified solution to easily handle app development across web, desktop, and mobile platforms, greatly simplifying the complexity of cross-platform development! ๐ป๐ฑ [Project Address](https://github.com/DioxusLabs/dioxus)
+
+2. **jsoncrack.com** is a hit project boasting **38,020 Stars** โญ! It's an innovative **open-source visualization application** that instantly transforms **JSON**, **YAML**, **XML**, **CSV**, and other data formats into **interactive charts** ๐, massively boosting data readability and analysis efficiency. It's practically a godsend for data enthusiasts! ๐ [Project Address](https://github.com/AykutSarac/jsoncrack.com)
+
+3. **free-for-dev** is an absolute treasure trove for **DevOps** and **infrastructure developers**! โจ With an astonishing **100,044 Stars**, it's a super practical **open-source** project that specifically compiles and provides a **list of free tiers** for SaaS, PaaS, and IaaS services. This is a tailor-made money-saving, time-saving magic tool for developers! ๐ฐโฐ [Project Address](https://github.com/ripienaar/free-for-dev)
+
+#### **Social Media Shares**
+
+1. Yang Yi excitedly shared Google AI Developer's **Gemini CLI**, calling it practically a "cyber savior"! ๐คฉ This **open-source AI agent** brings **Gemini 2.5 Pro** directly to your terminal, supporting **high-frequency free usage** for easily handling **code writing, debugging, and task automation**! He believes it's a "top-notch" solution for current tool shortcomings, with boundless potential, especially for **MCP deployment and GitHub search**! ๐
+ More details: ['More details'](https://x.com/Yangyixxxx/status/1937881859788304743)
+
+2. Xiaohu excitedly exclaimed he found a "badass" **AI design website**! It's a godsend for designers! ๐จโจ It can generate stunning and immediately usable interfaces, and it drastically **simplifies design prompt requirements**. What's even more impressive is that it can not only provide detailed design solutions based on simple descriptions but also generate **multi-level pages based on contextual logic**, and even supports **precise editing** of elements, greatly boosting design efficiency and freedom! ๐
+ More details: ['More details'](https://x.com/imxiaohu/status/1937847459117687004)
+
+3. Yang Yi thinks **AI singer Yuri** is the first AI Influencer to truly "break out"! ๐ค๐ฅ This **AI singer from Surreal** not only successfully partnered with The North Face, but her works have racked up over 7 million plays! This fully demonstrates **AI's growing influence and commercial potential in the virtual idol sphere**, signaling the arrival of an exciting new era! ๐
+ More details: ['More details'](https://x.com/Yangyixxxx/status/1937843457630020058)
+
+4. Alipay is really ahead of the curve! โจ They've launched their first **AI tipping** service, allowing developers to integrate this feature into their **AI agents**, so users can "send flowers" (virtual gifts) to their favorite **AI agents**! ๐ฐ๐ ['More details'](https://x.com/imxiaohu/status/1937830267873525781)
+
+
+5. Google just dropped a huge bomb! ๐ They've made the powerful **Imagen 4** and **Imagen 4 Ultra** image models freely available in **AI Studio**! ๐คฉ Now, users can experience these awesome image generation models for free via the **Gemini API** and AI Studio. Go ahead and give them a whirl! ๐จ ['More details'](https://x.com/op7418/status/1937708999430033734)
+
+
+
+6. Anthropic's **Claude Artifacts** is getting an update! ๐ฅณ Users will soon be able to browse and share popular web creations in the **Artifacts Gallery**, and even directly create **AI front-end applications** via the **Claude API**. Just thinking about it feels super cool! ๐ปโจ ['More details'](https://x.com/op7418/status/1937707999902203955)
+
+
+7. Zero Jun (AI Chat) shared an AI video that racked up over 50 million views in just 24 hours. He hit the nail on the head, pointing out that the secret to current **viral AI videos** is one word: "**ridiculous**"! ๐ It's not about aiming for hyper-realism. Common viral themes include ASMR, animal Olympics, and AI natural disasters. Want to see more "ridiculous" videos? Just click ['here'](https://m.okjike.com/originalPosts/685b65582b50c68918a1279b) for more info!
+
+
+8. Tom Huang shared 20 super practical **programming Prompt tips** ๐ก, and also spilled the beans that Warp is heavily developing a terminal Agent similar to Claude Code. While this Agent is pay-per-use, word on the street is you can make your money back in just one use! ๐ฑ It's practically a productivity godsend for programmers! ๐ For more details, hurry and click ['here'](https://x.com/tuturetom/status/1937669382752338129) to check it out!
+
+
+---
+
+#### **Listen to the Audio Version**
+
+| ๐๏ธ **Xiaoyuzhou** | ๐น **Douyin** |
+| --- | --- |
+| [Laisheng Tavern](https://www.xiaoyuzhoufm.com/podcast/683c62b7c1ca9cf575a5030e) | [Laisheng Intelligence Hub](https://www.douyin.com/user/MS4wLjABAAAAwpwqPQlu38sO38VyWgw9ZjDEnN4bMR5j8x111UxpseHR9DpB6-CveI5KRXOWuFwG)|
+|  |  |
\ No newline at end of file
diff --git a/content/en/_index.md b/content/en/_index.md
index 7d39dc1..73be16c 100644
--- a/content/en/_index.md
+++ b/content/en/_index.md
@@ -1,93 +1,98 @@
---
title: TodayDaily
breadcrumbs: false
-next: /en/2025-06/2025-06-24
-description: ByteDance's AI assistant, Doubao, recently dropped its "App Creation
- 1.0" AI programming feature, bringing an absolutely unprecedented visual programming
- experience๐คฉโจ! Users can directly drag, edit, and tweak web apps right in the preview
- interface, seriously lowering the barrier to entry for cod...
+next: /en/2025-06/2025-06-25
+description: Google DeepMind just dropped Gemini Robotics On-Device, an on-device
+ AI model designed specifically for robots ๐ค! It's built on the multimodal reasoning
+ of the Gemini 2.0 model, letting robots quickly learn new tasks, work stably even
+ offline, and even pull off intricate operations like folding c...
cascade:
type: docs
---
-# AI Daily Insights 2025/6/25
-
-> `AI Daily Report` | `Updated 8 AM Daily` | `Aggregating Data from Across the Web` | `Exploring Cutting-Edge Science` | `Industry Speaks Freely` | `Open-Source Innovation Power` | `AI and the Future of Humanity` | [Visit Web Version โ๏ธ](https://ai.hubtoday.app/)
-
+# AI Daily Insights 2025/6/26
+> `AI Daily` | `Morning Update` | `Web-wide Data Aggregation` | `Deep Dive into Frontier Science` | `Industry Voices Speak Up` | `Open Source Powerhouse` | `AI and the Future of Humanity` | [Visit Web Version โ๏ธ](https://ai.hubtoday.app/)
#### **AI Content Summary**
```
-ByteDance's Doubao introduces visual programming; Microsoft releases Mu model to simplify system interaction.
-Apple and Cambridge AI research make breakthroughs; GPT-4 aids new cancer drug development.
-In the AI era, technological depth is more crucial; multiple open-source tools and AI video models gain attention.
+AI products are updating fast, Google launched on-device AI for robots. iFlytek's medical model reached expert level.
+Quark's college application service is blowing up, expanding computing power. Rokid glasses are mass-produced, securing tons of orders.
+AI research sees breakthroughs in multimodal and 3D reconstruction. Zhou Hongyi says AI can't replace human emotion or creativity.
```
-
-
#### **AI Product & Feature Updates**
-1. ByteDance's AI assistant, **Doubao**, recently dropped its **"App Creation 1.0" AI programming** feature, bringing an absolutely *unprecedented* **visual programming experience**๐คฉโจ! Users can directly drag, edit, and tweak web apps right in the preview interface, seriously lowering the **barrier to entry for coding**. This means even if you've never coded before, you can whip up fully functional web apps super fast. It's totally expected to speed up the spread of **AI programming tools**!
-
[](https://autoproxy.justlikemaki.vip/?pp=https://pic.chinaz.com/picmap/202405160815252726_0.jpg)
-
[](https://autoproxy.justlikemaki.vip/?pp=https://pic.chinaz.com/2025/0624/6388637279333382299651050.png)
+1. Google DeepMind just dropped **Gemini Robotics On-Device**, an **on-device AI model** designed specifically for **robots** ๐ค! It's built on the **multimodal reasoning** of the **Gemini 2.0 model**, letting robots quickly learn new tasks, work stably even offline, and even pull off intricate operations like folding clothes โจ. This is definitely laying a solid foundation and opening a new chapter for the future of **embodied intelligence**!
+
-2. **Microsoft** just officially rolled out **Mu**๐ก๐, a **Small Language Model (SLM)** specifically designed for the **Windows 11 Settings app**. This 330-million-parameter model is **NPU-optimized**, delivering low-latency, high-privacy local natural language interaction, which totally simplifies how users mess with system settings. Mu's debut is a big deal, marking a major breakthrough for localized **AI tech** in OS interaction, and it's set to kick off a whole new paradigm for operating systems and **deep AI integration**!
-
-
[](https://autoproxy.justlikemaki.vip/?pp=https://pic.chinaz.com/2025/0624/6388637216973079154715278.png)
- ['More Details'](https://blogs.windows.com/windowsexperience/2025/06/23/introducing-mu-language-model-and-how-it-enabled-the-agent-in-windows-settings/)
+2. College application season is in full swing, and **Quark's** smart college recommendation service has been so popular it's seeing **queues**, with over 3 million reports generated so far ๐. That really shows how much students trust its AI capabilities. Faced with this "sweet headache," **Alibaba Group VP Wu Jia** came back strong, saying the team has urgently **boosted computing power** to make sure every student gets this crucial guide for their academic journey! ๐ช
+
+
+3. Rokid, in collaboration with Lens Technology, has officially started mass **production** of their **consumer-grade AI+AR glasses, Rokid Glasses**! ๐โจ These glasses are super **lightweight** and pack a punch with **AI large model capabilities** like **smart teleprompter, real-time translation, and AI object recognition**. They've already snagged **250,000 global pre-orders**! This signals an upcoming **commercial boom** in China's AI glasses market โ the future's looking bright! ๐
+
+
+4. At the 2025 Cloud Next conference, Google showcased its next-gen **customer service intelligent assistant** ๐ค, powered by the **Gemini model**. This assistant is seriously cool โ not only can it handle **multimodal interactions**, but it can also apply for **discounts** itself and is deeply integrated with the Salesforce **CRM system**! This hints at a massive intelligent revolution coming to customer service ๐ฅ, but we'll have to wait and see about its accuracy and privacy protection. ๐
+
+
+5. iFlytek just launched its powerful **Spark Medical Large Model V2.5 International Edition** ๐, trained on entirely domestically produced computing power! This model topped the charts on the authoritative MedBench platform with an impressive 98.4 points. Its comprehensive diagnostic and treatment capabilities have reached the level of an attending physician at a top-tier hospital, and it even surpasses human doctors in completeness, practicality, and readability! ๐จโโ๏ธ๐ฉบ It also supports multiple languages, promising to make a big splash in the global medical market and drive international medical technology exchange and collaboration! ๐โจ
+
+
+6. ElevenLabs finally dropped its standalone **text-to-speech mobile app**! ๐ฑโจ Whether you're on iOS or Android, you can now generate audio snippets anytime, anywhere. Even free users get about 10 minutes of audio generation time! This app uses the latest v3alpha model and even supports **emotional expression control**. Plus, it's set to add speech-to-text and conversational AI tools in the future โ how convenient is that?! ๐ฃ๏ธ
+
#### **AI Frontier Research**
-1. Apple recently unveiled some really eye-catching research, dropping new **AI image generation models** โ **TarFlow** and **STARFlow**๐๐ฌโจ โ built on **normalized flow** technology. Unlike traditional diffusion models, this tech can precisely calculate the probability of generating images. What's super cool is the **STARFlow** model, which works in **latent space** and can even call existing language models to fine-tune **text prompt processing**. It's totally nailed the challenge of generating high-resolution images, giving **image generation tech** a fresh new direction.
-
[](https://autoproxy.justlikemaki.vip/?pp=https://pic.chinaz.com/2025/0624/6388635461229244306298224.png)
+1. **SuperDec**, a joint effort by ETH Zurich, Stanford University, and Microsoft teams, is **shattering conventional 3D reconstruction limits** ๐คฏ! This tech uses an innovative **hyper-tetrahedron** principle to create compact yet vivid **3D scene representations**. It efficiently handles complex **point cloud** data and shows huge potential in areas like precise grasping and path planning for **robotics**, as well as **controllable visual content generation**, opening up new horizons for the digital world! ๐ [Project Link](https://super-dec.github.io/)
-2. Get ready to be thrilled by the latest research from Cambridge University and other institutions! ๐๐งฌ๐ They've successfully used **large language models** (LLMs) like **GPT-4** for **new cancer drug development**, making it the first time an LLM has been used to generate scientific hypotheses. And guess what? They've made groundbreaking strides in breast cancer treatment! This research, using **GPT-4**, proposed a bunch of **drug combinations**, and the combo of simvastatin and disulfiram showed *huge* potential for effectively fighting **breast cancer**. It's totally opening up brand new avenues for medical research!
-
[](https://autoproxy.justlikemaki.vip/?pp=https://pic.chinaz.com/2025/0624/6388635234062890531897043.png)
- **Paper Link**: ['Paper Link'](https://royalsocietypublishing.org/doi/10.1098/rsif.2024.0674)
+2. **4D-LRM** is a super cool, innovative **large spatio-temporal reconstruction model** ๐คฉ. It can fully reconstruct **dynamic objects' 4D representations** (3D space plus time dimension) from just a few input views, generating high-quality scenes from any time and any angle! In the future, it's set to make a huge impact in virtual reality, film production, industrial simulation, and more! ๐ [Paper Link](https://huggingface.co/papers/2506.18890)
-3. **OmniGen2** is this super versatile, open-source **multimodal generative model**๐จ๐ค๐ that can handle a bunch of tasks like text-to-image generation, image editing, and context generation, all in one go! And get this, it totally rocks in relevant benchmarks. Even though its model parameter scale is pretty moderate, it's hit the **best performance** among open-source models when it comes to consistency, and it even brings in a brand new **OmniContext** benchmark. Seriously, it's awesome!
- **Paper Link**: ['Paper Link'](https://arxiv.org/abs/2506.18871)
+3. ByteDance and Shanghai Jiao Tong University have teamed up to release the **ProtoReasoning framework** ๐. It cleverly leverages structured prototype representations like **Prolog** and **PDDL** to significantly boost **large language models' logical reasoning capabilities** and efficiency in cross-domain knowledge transfer ๐. This research lays a solid foundation for future theoretical explorations of **reasoning prototypes** โ truly awesome! [Paper Link](https://arxiv.org/abs/2506.15211)
+
+4. MMLab at the University of Hong Kong, MMLab at the Chinese University of Hong Kong, and SenseTime have jointly developed the **GoT-R1 framework**. This groundbreaking research, by introducing **reinforcement learning**, significantly enhances **multimodal large models' semantic-spatial reasoning capabilities** in **visual generation tasks** ๐, allowing models to learn better reasoning strategies autonomously! It not only breaks free from the GoT framework's reliance on templates but also achieves **SOTA performance** in complex scene generation โ seriously impressive! โจ [Paper Link](https://www.jiqizhixin.com/articles/2025-06-25-9)
#### **AI Industry Outlook & Social Impact**
-1. So, during a live stream on June 24th, famous education blogger **Zhang Xuefeng** got asked if he was worried about being replaced by **AI**. And get this, he surprisingly said, "It'd be best if I *could* be replaced! ๐๐ก๐" This totally shows his optimistic vibe about **AI**'s development and a positive outlook on the future of education. Plus, he stressed that educators really need to step up their communication with students' parents to better leverage **AI tools**. Talk about a clear-headed and insightful take!
-
[](https://autoproxy.justlikemaki.vip/?pp=https://pic.chinaz.com/picmap/202005281119277542_8.jpg)
+1. Zhou Hongyi recently chatted about the future of AI in a video. He believes that no matter how powerful AI gets, it will never fully replace humanity's unique abilities in **emotional understanding** ๐, **complex problem-solving** ๐ง , and **creative thinking** ๐จ. He stressed that future work will increasingly involve **managing and training** AI, even citing a failed case of AI customer service at a Swedish company to show that AI still has limitations when handling complex customer needs. ๐ง
+
-#### **TOP Open-Source Projects**
+2. Federal Judge William Alsup made a groundbreaking ruling: **Anthropic's** use of **copyrighted books** to train its **AI model** without permission was deemed **fair use**! ๐ฎ This sets an important precedent for copyright disputes in the AI industry. However, Anthropic is still facing **theft charges** for obtaining training material from pirated websites โ a mixed bag, huh? ๐ค
+
-1. Guess what just dropped? A new **open-source graphical interface tool** called **Claudia**๐ป๐ก๏ธโจ! It's specifically made for **Claude Code**. This tool aims to make command-line stuff way easier by giving you a super sleek and intuitive desktop experience. Plus, it's built on the **Tauri cross-platform framework**, so it runs on multiple systems. It's also got cool features like **privacy-first** and **local storage**, offering one-stop project management, custom AI agents, and session timelines. This thing is totally set to become a **benchmark tool** in the **AI programming space**!
-
[](https://autoproxy.justlikemaki.vip/?pp=https://pic.chinaz.com/2025/0624/6388638413004412367723772.png)
- ['Project Link'](https://github.com/getAsterisk/claudia)
+#### **TOP Open Source Projects**
-2. **ScholAI**, a **smart academic research tool**๐๐ฌ๐ built on **MCP**, just launched and it's grabbing a lot of attention. It packs a ton of features like **paper searching**, **analysis**, **management**, **CCF ranking lookup**, and **semantic query analysis**, all designed to give researchers a super efficient and smart solution for their academic work. Right now, its **grayscale testing** has already pulled in tons of researchers, showing off its huge potential for boosting efficiency in **literature reviews** and **journal selection**. Seriously, it's a game-changer for the academic world!
-
[](https://autoproxy.justlikemaki.vip/?pp=https://pic.chinaz.com/2025/0624/6388637154747591468300279.jpg)
- **Project Link**: ['Project Link'](https://github.com/oDaiSuno/ScholAI)
+1. **Dioxus** is a super popular **full-stack application framework** with 28,310 stars โญ! It's like an all-in-one toolbox, aiming to give developers a unified solution to easily build applications for web, desktop, and mobile platforms, greatly simplifying the complexity of cross-platform development! ๐ป๐ฑ [Project Link](https://github.com/DioxusLabs/dioxus)
-3. The open-source project **leaked-system-prompts**๐๐ is a collection specifically for leaked system prompts, aiming to give developers a super rich resource library for research and reference. This project has snagged a whopping **9951** stars on GitHub, which just goes to show how popular and useful it is!
- **Project Link**: ['Project Link'](https://github.com/jujumilk3/leaked-system-prompts)
+2. **jsoncrack.com** is a star project boasting **38,020 Stars** โญ! It's an innovative **open-source visualization application** that can instantly transform various data formats like **JSON**, **YAML**, **XML**, and **CSV** into **interactive diagrams** ๐, massively boosting data readability and analysis efficiency. It's a godsend for data enthusiasts! ๐ [Project Link](https://github.com/AykutSarac/jsoncrack.com)
-4. The open-source project **claude-code-router**โ๏ธ๐ uses **Claude Code** as its coding backbone, letting users totally dig Anthropic updates while still having flexible control over how they interact with the model. This project has already racked up **1324** stars, and honestly, it sounds super practical and cool!
- **Project Link**: ['Project Link'](https://github.com/musistudio/claude-code-router)
-
-5. **best-of-ml-python** is an open-source project with an absolutely insane **20406** stars๐๐๐! It's all about dishing out weekly updates on the best **machine learning Python libraries**, which is pretty much a godsend for ML enthusiasts and developers looking for the top tools out there!
- **Project Link**: ['Project Link'](https://github.com/ml-tooling/best-of-ml-python)
+3. **free-for-dev** is a treasure trove of a project for **DevOps** and **infrastructure developers**! โจ It boasts an astonishing **100,044 Stars** and is a super handy **open-source** project specifically curating and providing a **list of free tiers** for SaaS, PaaS, and IaaS services. This is truly a tailor-made money- and time-saver for developers! ๐ฐโฐ [Project Link](https://github.com/ripienaar/free-for-dev)
#### **Social Media Shares**
-1. In a recent social media share, user **meng shao** dropped an awesome comparison test๐ฅ๐๐๏ธ of three **AI video products**: **Midjourney**, **Veo3**, and **Hailuo**! She used the exact same prompt to check out how differently they performed when generating a video of "flying spaghetti driving a car." Talk about a visual feast! You can totally watch the video provided to get a direct feel for how each model stacks up.
-
- ['More Details'](https://x.com/shao__meng/status/1937499181180158154)
+1. Yang Yi excitedly shared Google AI Developer's new **Gemini CLI**, calling it a "cyber Bodhisattva"! ๐คฉ This **open-source AI agent** brings **Gemini 2.5 Pro** right to your terminal, offering **high-frequency free usage** and making **code writing, debugging, and task automation** a breeze! He thinks it's a "top-tier" solution for current tool shortcomings, especially for **MCP deployment and GitHub search**, with unlimited potential! ๐
+ More details: ['More details'](https://x.com/Yangyixxxx/status/1937881859788304743)
-2. Xiangyang Qiaomu is absolutely raving๐คฏ๐๐๏ธ about the physics effects of the **Hailuo 02 model**, saying it conjures up a "living, breathing" interactive **virtual world**, and its **physical understanding** is way beyond Veo 3. This model has evolved from "individual realism" to "**interactive realism**" with its environment, showing off stunning results and stronger model capabilities through test cases like collapsing blocks. Seriously, it's mind-blowing!
-
- ['More Details'](https://x.com/vista8/status/1937376239788130652)
+2. Xiao Hu gasped, he found an "awesome" **AI design website**! It's seriously a godsend for designers! ๐จโจ It can generate beautiful, ready-to-use interfaces, and it massively **simplifies design prompt requirements**. Even cooler, it can provide detailed design solutions based on simple descriptions, and even generate **multi-level pages based on contextual logic**, plus it supports **precise editing of elements**, hugely boosting design efficiency and flexibility! ๐
+ More details: ['More details'](https://x.com/imxiaohu/status/1937847459117687004)
-3. Baoyu hit the nail on the head๐ค๐ง ๐ก, pointing out that in the **AI era**, technical **depth** matters way more than **breadth**. Why? Because AI can totally cover for a lack of breadth, but it can't make up for a lack of depth. He stressed that **experts in their field** can still churn out high-quality results even with AI's help, while those who are just "jacks of all trades, masters of none" will struggle to hit excellent levels. This really digs deep into the core idea that AI is all about **empowering specialized skills**, not **totally replacing them**. Something to really chew on!
- ['More Details'](https://x.com/dotey/status/1937352533485171025)
+3. Yang Yi thinks **AI singer Yuri** is the first **AI Influencer** to truly "go mainstream"! ๐ค๐ฅ This **AI singer from Surreal** successfully collaborated with The North Face, and her work has racked up over 7 million plays! This clearly shows **AI's growing influence and commercial potential in the virtual idol space**, signaling an exciting new era! ๐
+ More details: ['More details'](https://x.com/Yangyixxxx/status/1937843457630020058)
-4. Baoyu also brought up the hot discussion๐ป๐ธ๐ง around **AI code generation quality**. He pointed out that in big projects, the code quality from **Claude Code** just isn't as good as the more expensive combo of **Cline + Gemini 2.5 Pro**, and the former needs way more human tweaking. This doesn't just show how wildly different various **AI models** are when it comes to generating code; it also spills the tea on the **high costs** that can come with chasing **top-notch AI-assisted programming**. Talk about a love-hate balancing act!
-
[](https://pbs.twimg.com/media/GuIvKylXcAAsHri?format=jpg&name=orig)
- ['More Details'](https://x.com/dotey/status/1937221441658732730)
+4. Alipay is really on the cutting edge! โจ They've launched their first-ever **AI tipping** service, letting developers integrate this feature into their **smart agents** so users can "send flowers" (give tips) to their favorite **smart agents**! ๐ฐ๐ ['More details'](https://x.com/imxiaohu/status/1937830267873525781)
+
+
+5. Google's dropping a major bombshell! ๐ They've made the powerful **Imagen 4** and **Imagen 4 Ultra** image models available for free in **AI Studio**! ๐คฉ Now, users can experience these awesome image generation models for free via the **Gemini API** and AI Studio. Go check it out! ๐จ ['More details'](https://x.com/op7418/status/1937708999430033734)
+
+
+
+6. Anthropic's **Claude Artifacts** is getting an update! ๐ฅณ Users will soon be able to browse and share popular web creations in the **Artifacts Gallery**, and even create **AI frontend applications** directly via the **Claude API** โ just thinking about it sounds super cool! ๐ปโจ ['More details'](https://x.com/op7418/status/1937707999902203955)
+
+
+7. Zero Jun Chatting AI shared an **AI video** that hit over 50 million views in 24 hours. He hit the nail on the head by pointing out that the secret to current **viral AI videos** is one word โ "**outrageous**"! ๐ It's not about being human-like. Common viral themes include ASMR, animal Olympics, and AI natural disasters. Want to see more "outrageous" videos? Click ['here'](https://m.okjike.com/originalPosts/685b65582b50c68918a1279b) to learn more!
+
+
+8. Tom Huang shared 20 super useful **programming Prompt** tips ๐ก, and also revealed that Warp is heavily developing a **terminal Agent** similar to Claude Code. While this Agent is pay-per-use, apparently, one use can earn back the cost! ๐ฑ It's practically a productivity godsend for programmers! ๐ For more details, hurry and click ['here'](https://x.com/tuturetom/status/1937669382752338129) to check it out!
+
---
@@ -95,5 +100,5 @@ In the AI era, technological depth is more crucial; multiple open-source tools a
| ๐๏ธ **Xiaoyuzhou** | ๐น **Douyin** |
| --- | --- |
-| [Laisheng Pub](https://www.xiaoyuzhoufm.com/podcast/683c62b7c1ca9cf575a5030e) | [Laisheng Intel Hub](https://www.douyin.com/user/MS4wLjABAAAAwpwqPQlu38sO38VyWgw9ZjDEnN4bMR5j8x111UxpseHR9DpB6-CveI5KRXOWuFwG)|
-|  |  |
\ No newline at end of file
+| [Next Life Speakeasy](https://www.xiaoyuzhoufm.com/podcast/683c62b7c1ca9cf575a5030e) | [Next Life Intel Hub](https://www.douyin.com/user/MS4wLjABAAAAwpwqPQlu38sO38VyWgw9ZjDEnN4bMR5j8x111UxpseHR9DpB6-CveI5KRXOWuFwG) |
+|  |  |
\ No newline at end of file
diff --git a/content/en/about.md b/content/en/about.md
index d8cce49..24f8190 100644
--- a/content/en/about.md
+++ b/content/en/about.md
@@ -4,30 +4,33 @@ type: about
sidebar:
exclude: true
---
-#### ๐ Hexi 2077 / justlovemaki
+#### ๐ He Xi 2077 / justlovemaki
-> Spent a decade coding, and my fingers are feeling the chill. All that pent-up frustration? Still bottled up. But then, I heard the buzz about AI, and it's like a storm's brewing. Now I'm ready to jump in and give it everything I've got.
+> Been coding for ten years, seen a lot of things,
+> And I've got a ton of ideas just waiting to burst out.
+> Suddenly, I hear the buzz about AI making huge waves.
+> Now I'm ready to jump in and really make my mark!
#### ๐ My Code Philosophy
-> Tech should serve the people.
+> Technology for the people!
-#### โจ My Top Picks
+#### โจ Featured Work
* **[Open Source Contribution / CloudFlare-AI-Image](https://github.com/justlovemaki/CloudFlare-AI-Image)**:
- * An AI image generation script built on Cloudflare Workers.
+ * An AI image generation script powered by Cloudflare Workers.
* **[Open Source Contribution / CloudFlare-AI-Insight-Daily](https://github.com/justlovemaki/CloudFlare-AI-Insight-Daily)**:
- * A content aggregation and generation platform powered by Cloudflare Workers. It curates the latest happenings in the AI world for you daily, covering industry news, popular open-source projects, cutting-edge academic papers, and social media buzz from tech influencers.
-* For more project details, check out my [GitHub](https://github.com/justlovemaki).
+ * A content aggregation and generation platform powered by Cloudflare Workers. Each day, it rounds up the freshest AI news and insights for you, from industry updates and hot open-source projects to cutting-edge academic papers and social media chatter from top tech influencers.
+* Check out my [GitHub](https://github.com/justlovemaki) for more project details.
-#### ๐ฑ What I'm Exploring Now
+#### ๐ฑ Currently Exploring
-Super interested in LLM applications and website SEO right now, and I'm actively diving into learning and getting my hands dirty with them.
+I'm really passionate about LLM applications and website SEO, and I'm actively learning and getting hands-on with them.
#### ๐ซ Contact Me
* **Email:** [274166795@qq.com](mailto:274166795@qq.com)
* **GitHub:** [https://github.com/justlovemaki](https://github.com/justlovemaki)
* {{< cards >}}
- {{< card link="https://raw.githubusercontent.com/justlovemaki/CloudFlare-AI-Insight-Daily/refs/heads/main/docs/images/wechat.png" title="Personal WeChat" subtitle="Feel free to add me for a chat!" image="https://raw.githubusercontent.com/justlovemaki/CloudFlare-AI-Insight-Daily/refs/heads/main/docs/images/wechat.png">}}
+ {{< card link="https://raw.githubusercontent.com/justlovemaki/CloudFlare-AI-Insight-Daily/refs/heads/main/docs/images/wechat.png" title="My WeChat" subtitle="Feel free to connect and chat!" image="https://raw.githubusercontent.com/justlovemaki/CloudFlare-AI-Insight-Daily/refs/heads/main/docs/images/wechat.png">}}
{{< /cards >}}
\ No newline at end of file