AI News

Yesterday, I saw someone in the group ask: "Does anyone subscribe to ChatGPT?" It suddenly reminded me of a great deal I recently came across from OpenAI—a free year's ChatGPT membership. I immediately tried it out, and in less than two minutes...

The HuggingFace team's "Small Model Training Guide: The Core Rules for Building Top-Tier Language Models" is an over 200-page technical blog that systematically shares end-to-end experience in training advanced LLMs.

This article, "Introduction to Agents," delves into the multi-dimensional development of intelligent agents, from infrastructure design and the classification of complex systems to security considerations in practical deployment and operation. The article reveals how intelligent agents...

I recently discovered a quick way to replicate trending product covers. Any trending cover you see can be replicated using AI in minutes, with incredibly high fidelity, and it's completely free! The whole process requires almost no manual effort. We only need to input a few...

This year's Double 11 shopping festival saw e-commerce platforms remain relatively quiet, while major AI companies launched a series of promotional offers. Zhipu launched a new GLM Coding Plan, offering a 20 RMB monthly subscription for enjoying GLM-4.6. ByteDance released Doubao-Seed-Code, the first...

Zaoci currently requires an invitation code to experience. Each user receives one invitation code upon successful login, which can be used five times. Currently, access is only available to iOS users. The official AI Toolset website provides a Zaoci invitation code sharing group, where invitation codes are shared for free to help unlock features...

The popular content creator "Film Hurricane," with millions of followers, used it to build its content management backend. It records backend content data, playback trends for each episode, and interaction status, all of which can be analyzed in real time. From topic selection to post-production review, the entire content creation process can run online.

The highly anticipated Gemini 3 is finally here! This time, it features a comprehensive upgrade in inference, multimodal, programming, and agent capabilities. Google DeepMind states: This is the world's best multimodal understanding model, and our most powerful to date...

Just two days after Gemini 3 Pro topped the charts, Google released its latest image model, Nano Banana Pro (also known as Gemini 3 Pro Image). Nano Banana Pro's kernel has been upgraded from Gemini 2.5 Flash to G...

Another great day to snag some freebies! Get a free month of Google Gemini membership, which includes free access to Gemini 3 Pro, Nano Banana Pro, Veo 3, Deep Research... The best part is, no bank card required, just log in to claim it...

I recently discovered that the Qianwen App has been updated again. This time, it features a comprehensive upgrade to its creative capabilities, incorporating the latest image model, Qwen-Image-Edit, for both raw and edited images, as well as Wan2.5's audio and video capabilities. A single image can generate a narrated...

"Are there any tools suitable for elementary school students to learn idioms?", "I want to make a game app, what tools would be suitable?", "I have a brilliant idea, can you help me make it?"... For some niche scenarios, it's hard to find ready-made solutions...

In today's digital age, artificial intelligence (AI), especially large language models (LLMs), is changing our world at an unprecedented pace. However, we still lack sufficient information regarding the real-world applications and impact of these models...

Zhipu officially launched its new visual reasoning model series, GLM-4.6V, and made it fully open source! This release includes two versions: GLM-4.6V: total parameters 106B, activation parameters per inference approximately 12B, visual understanding accuracy reaching...

It's been over two weeks since Google released Nano Banana Pro, but its creative uses continue to emerge. Previously, the most troublesome issues with AI-generated images—consistency and accuracy of Chinese characters—have been solved, and the suggestion words...

OiiOii is the world's first professional animation creation agent, enabling the rapid generation of animations from concept to finished product using intelligent agents. Users can obtain animated shorts, videos, or images simply by entering simple commands.

The AI agent for animation creation has finally arrived! Last month, OiiOii was launched as the world's first professional animation creation agent, and even invitation codes were extremely scarce. OiiOii has 7 built-in animation agents and 162 preset animation styles...

"The Human Economic Index: Tracking the Role of AI in the US and Global Economy" analyzes global Claude AI usage data, revealing a significant imbalance in AI application, with high-income countries and knowledge-economy-dominated regions showing higher usage rates...

YC's Vibe Coding experience is compiled from the YC video "How To Get The Most Out Of Vibe Coding | Startup School," shared by YC partner Tom Blomfield, and is based on artificial intelligence...

The era of AI-customized education for everyone has arrived. Yesterday, Google launched its latest educational tool, Learn Your Way, which uses AI to transform textbooks into multimodal content and adds interaction and quizzes to every stage of learning, allowing everyone to...

I've recently seen a bunch of cross-dressing videos on Douyin (TikTok), all of them going viral with at least 10,000 likes. I did some research and found that this format is perfect for using AI to "create unorthodox accounts"—basic editing is all that's needed, and you don't even need to show your face. I've already figured out a way to...

The *Enterprise AI Application Building Guide* comprehensively explains the methods and technical systems for building enterprise-level AI applications. The content covers the evolution of AI application architecture (from dialogue and RAG to workflow and agent models), and provides in-depth analysis of delivery processes, MaaS, ...

If I had to recommend just one tool that truly saves most people time and effort, I would say without hesitation: Lark Multidimensional Table. This isn't just my personal preference. Many of my friends who work on AI content use it, and even others like DeepSeek and Ki...

The 2025 Yunqi Conference is finally here! This year's theme is "Cloud Intelligence Integration • Silicon-Carbon Symbiosis." More than 2,000 speakers from over 50 countries around the world have gathered in Hangzhou to discuss cutting-edge topics such as Agentic AI and Physical AI...