Welcome to the AI Daily section! Here is your daily guide to exploring the world of artificial intelligence. Every day, we bring you the hottest topics in the AI field, focusing on developers, helping you understand technological trends and discover innovative AI product applications.
Fresh AI Products Click to learn more: https://top.aibase.com/
1. Free for Commercial Use! Stability AI Launches Lightweight AI Painting Tool - Stable Diffusion 3.5 Medium Model
Stability AI introduces the new Stable Diffusion 3.5 Medium model, achieving a perfect balance between high performance and accessibility. This AI painting tool is free for commercial use, utilizing the MMDiT-X architecture, addressing hardware barriers for ordinary users, and realizing the vision of universal access. Technological innovations include the integration of three pre-trained text encoders, the introduction of QK normalization technology to enhance training stability, and the design of a dual attention module in the first 12 transformation layers, significantly improving image quality and layout effects.
AiBase Highlights:
🚀 Perfect balance between high performance and accessibility, free for commercial use, accessible to everyone.
💡 MMDiT-X architecture, solving hardware barriers for ordinary users, designed with 2.5 billion parameters, runs smoothly with 9.9GB of VRAM.
🔑 Technological innovations integrate three pre-trained text encoders, introduce QK normalization technology, and significantly enhance image quality and layout effects with a dual attention module design.
Details link: https://huggingface.co/stabilityai/stable-diffusion-3.5-medium
2. Hedra Launches New Voice Cloning Feature, Making AI Virtual Avatars Sound More Realistic
Recently, the popular AI character creation tool Hedra introduced an exciting voice cloning feature. Users can clone, store, and apply their voice to created or uploaded characters with just a few recordings. The operation is simple and straightforward, with excellent lip-sync performance and accurate mouth movements. Paid users can enjoy this feature, with the minimum subscription fee starting at $10 per month.
AiBase Highlights:
✨ Accurately clone your voice with just a few recordings, simple and easy to use.
🎨 Outstanding image generator, the generated characters are almost indistinguishable from real faces, but the lip-sync effect for animal characters is poor.
💰 Paid users can enjoy the voice cloning feature, with the minimum subscription fee starting at $10 per month.
3.WeChat Exposed to Gray-scale Testing AI Q&A Feature, Official Response: Optimizing User Search Experience
Recently, WeChat is conducting gray-scale testing of an AI Q&A feature aimed at optimizing user search experience and improving search efficiency. This feature integrates technical resources from Tencent, including the Hunyuan large model.
AiBase Highlights:
🔍 AI Q&A feature is in gray-scale testing, users can trigger it through WeChat search, generating about 400 words of answers and 5 reference materials.
🔧 The test aims to optimize user search experience and improve search efficiency, integrating Tencent's internal technical resources.
💡 Tencent's Hunyuan large model is based on the Transformer neural network, with a trillion parameter scale, possessing powerful Chinese creation and logical reasoning capabilities.
4. ComfyUI's New Tool ComfyUI-Detail-Daemon: One-click Enhancement for AI Painting Details, Upgrading Flux Model Quality
I delved into the recently buzzworthy ComfyUI-Detail-Daemon plugin. This tool developed by Jonseed has garnered significant attention within the AI painting community, cleverly porting sd-webui-Detail-Daemon to the ComfyUI platform, bringing creators an unprecedented experience of detail optimization. Most impressive is its powerful detail enhancement capabilities and excellent processing performance on Flux model works.
AiBase Highlights:
⚙️ Powerful detail enhancement capabilities, precisely control sigma parameters to improve image details, intelligently remove bokeh and background blur, especially suitable for Flux model works.
🔍 Equipped with four core nodes with distinct functions, including Detail Daemon Sampler, Detail Daemon Graph Sigmas, Multiply Sigmas, and Lying Sigma Sampler, simplifying the parameter adjustment process and improving user efficiency.
🚀 Streamline the workflow, reduce technical barriers, achieve the effects previously requiring cumbersome node combinations with just a few simple steps, enhancing creative efficiency.
Details link: https://github.com/Jonseed/ComfyUI-Detail-Daemon
5. Meta Launches New Video Generation Model MarDini, Tackling the Task of Filling Missing Frames in Videos
Meta, in collaboration with KAUST, has introduced the MarDini model, making video creation simpler and more flexible, supporting various tasks such as filling missing frames in videos, image-to-video conversion, and more. The model employs advanced video generation technology, offering high flexibility and efficiency, capable of training from scratch and handling multiple video tasks.
AiBase Highlights:
✨ MarDini is a new video generation model developed by Meta in collaboration with KAUST, capable of easily completing various video creation tasks.
🎥 Through the combination of planning and generation models, MarDini achieves efficient video interpolation and image-to-video generation.
💡 MarDini generates high-quality videos with fewer steps, enhancing the flexibility and efficiency of creation.
Details link: https://mardini-vidgen.github.io/
6. Say Goodbye to Chaotic Chat History! ChatGPT Web Version Introduces Chat History Search Feature
OpenAI has introduced a highly anticipated new feature on the ChatGPT web version, allowing users to easily search chat histories, providing convenience for users. Positive user feedback anticipates the gradual rollout of the feature. More updates may be on the horizon to enhance user experience.
AiBase Highlights:
🌟 Plus and Team users will experience the chat history search feature within a day.
📅 Enterprise and Edu users will gain access within a week, with free users starting next month.
💻 AI-first pull request reviewers provide developers with smarter support.
7. Adorable Impact! AI-Generated Jellycat x Hermès Plush Toys Go Viral, Netizens Say They Woke Up from a Dream
An unexpected dream collaboration stirred up discussions on social media. Xiaonaimi Xin, a blogger on Xiaohongshu, successfully created a sweet misunderstanding with AI-generated images of Jellycat x Hermès plush toys. Netizens are intrigued by the collision of luxury and cute plush toys created by AI, but converting them into physical products may face copyright challenges.
AiBase Highlights:
😍 AI-generated Jellycat x Hermès plush toys have sparked discussions among netizens, showcasing the unique collision of luxury and cute plush toys.
🤔 The virtual collaboration has drawn attention from the real market, with merchants expressing intentions to produce physical versions, but facing challenges such as brand authorization and other copyright issues.
🚀 AI technology demonstrates potential in the field of creative design, with consumers showing a preference for cute luxury goods, but commercialization still needs to consider legal and market constraints.
8. X Users Easily Earn Thousands of Dollars Monthly Using False Election Information and AI Images
On the social media platform X, some users are claiming to earn thousands of dollars monthly by exploiting false information and AI-generated images. They increase exposure and income by posting controversial content, earning high returns while also attracting attention and controversy. X platform has changed its revenue rules, encouraging users to create more controversial content, further increasing user earnings.
AiBase Highlights:
🌟 Users on social media X share election false information, claiming to earn thousands of dollars monthly.
🤝 X platform changes revenue rules, encouraging users to create more controversial content.
💰 Some users earn high income through AI-generated images and various content.
9. GitHub Introduces Multiple Models and Spark Platform: Everyone Can Develop AI Applications with Natural Language
GitHub unveiled the new platform GitHub Spark at the Universe '24 developer conference, aiming to help developers easily create micro-applications using natural language. At the same time, it offers multiple AI model choices, such as Claude3.5Sonnet, Gemini1.5Pro, and GPT-4o. Python has surpassed JavaScript as the most popular programming language on GitHub, with public AI projects growing by 98%.
AiBase Highlights:
✨ GitHub Spark platform helps developers create AI micro-applications with natural language.
🛠️ Copilot assistant supports multiple AI model choices, allowing flexible switching.
📈 Python surpasses JavaScript as the most popular programming language on GitHub, with public AI projects growing by 98%.
10. Musk's xAI Gains Momentum, Plans to Raise Billions More, Valuation Expected to Reach $40 Billion
Elon Musk's artificial intelligence company xAI is thriving and plans to raise billions of dollars more, with an expected valuation of $40 billion. The new funding will further enhance the company's market position and competitiveness.
AiBase Highlights:
🚀 xAI is seeking billions of dollars in new funding, with an expected valuation of $40 billion.
💼 The company completed a $6 billion B round of financing in May, and the new funding will significantly boost its market valuation.
🔍 xAI is using a large number of Nvidia GPUs to train new models, while adjustments to the X platform's privacy policy also facilitate data utilization.
11. Amazon Launches AI Coding Assistant Q Developer, Integratable into Development Environments
This article introduces Amazon's AI coding assistant Q Developer, which can be integrated into developers' IDEs to help improve coding efficiency. This assistant is based on Anthropic's Claude3.5Sonnet model and can automatically handle code refactoring and document generation, enhancing developer productivity.
AiBase Highlights:
💻 Amazon launches Amazon Q Developer AI assistant, supporting direct operations within IDEs, enhancing developer efficiency.
🤖 Q Developer is based on the Claude3.5Sonnet model, capable of automating code refactoring and document generation.
🚀 Q Developer's inline chat feature excels in code refactoring and document generation, helping developers quickly integrate changes and maintain focus.
Details link: https://aws.amazon.com/cn/blogs/devops/amazon-q-developer-inline-chat/
