Welcome to the "AI Daily" column! Here is your daily guide to exploring the world of artificial intelligence. Every day, we present the hottest topics in the AI field, focusing on developers to help you understand technological trends and innovative AI product applications.
Fresh AI Products Click to Learn More: https://top.aibase.com/
1. More Features, Same Price! Doubao Large Model 1.5 Officially Released, Including Doubao-1.5-vision-pro, etc.
ByteDance released the Doubao Large Model 1.5 on January 22, 2025, marking a significant breakthrough in the field of artificial intelligence. This model excels in various areas, particularly in knowledge, coding, and reasoning, surpassing industry-leading models like GPT-4o and Claude3.5Sonnet. The Doubao Large Model 1.5 also introduces a real-time voice model, enhancing the efficiency of voice interactions and user experience.
[AiBase Summary:]
📈 Doubao Large Model 1.5 performs excellently in multiple authoritative evaluations, surpassing top industry models.
💡 The newly added real-time voice model enables low-latency conversations, advancing voice interaction technology.
💰 Prices remain unchanged, continuing the commitment to "more features at no extra cost," promoting the accessibility of AI technology.
Details link: https://team.doubao.com/doubao_1_5_pro
2. Keling AI Launches New Multi-Image Reference Feature, Supporting Multiple References with a Maximum Upload of 4 Photos
Keling AI recently introduced an innovative multi-image reference feature, allowing users to upload up to four reference images simultaneously and generate personalized videos by selecting specific elements. This feature not only enhances the visual consistency of short video production but also provides creators with greater freedom to showcase interactions between characters or objects in specific scenes.
[AiBase Summary:]
📸 Users can upload 1 to 4 reference images, specifying characters, objects, and scenes.
🎭 This feature supports interactions between characters, allowing users to freely construct storylines.
🎨 Suitable for short video production, ensuring visual consistency and enhancing the creative experience.
Details link: https://top.aibase.com/tool/keling-ai
3. WeChat's "SewCut" Adds Multiple AI Features
The official video editing tool "SewCut" from WeChat has recently added several practical AI features, significantly improving users' video creation efficiency. The new version includes functions like "Public Account to Video," "AI Comic Video," and "Product Promotion Video," allowing users to easily convert WeChat articles into videos or generate diverse comic videos based on text and audio.
[AiBase Summary:]
📝 The new "Public Account to Video" feature allows users to convert articles into videos easily.
🎨 The "AI Comic Video" feature supports text and audio imports, enabling various video styles.
🛍️ The "Product Promotion Video" feature generates copy with one click, greatly enhancing e-commerce video production efficiency.
4. Google's New Upgrade: Gemini 2.0 Flash Thinking, Achieving New Heights in Long Context Processing
On January 22, 2025, Google released an enhanced version of the Gemini 2.0 Flash Thinking reasoning model, significantly improving long context processing capabilities, supporting inputs of up to 1M tokens. This update aims to enhance the model's reliability and reduce discrepancies between generated content and final answers. The new version shows remarkable improvements in mathematics, science, and multimodal reasoning capabilities, especially achieving a 54% increase in mathematical performance.
[AiBase Summary:]
🧠 Gemini 2.0 Flash Thinking supports long context processing of up to 1M tokens, enhancing reasoning capabilities for complex dialogues.
🔍 The model shows significant performance in self-correction, ensuring coherence and consistency in conversations.
📈 In mathematics ability tests, the model's performance improved by 54%, demonstrating excellent multimodal reasoning capabilities.
Details link: https://aistudio.google.com/prompts/new_chat
5. SenseTime's "Instant Painting Fun Shot" App Officially Launched, Supporting AI Portraits, Videos, Filters, and More
SenseTime has recently launched a new app called "Instant Painting Fun Shot," providing users with an innovative AI imaging experience. This application aims to celebrate the upcoming Spring Festival by using AI technology to help users create personalized New Year portraits, enhancing the festive atmosphere. Users can upload a frontal photo and quickly generate AI portraits by selecting a template. Additionally, the app allows users to customize video uploads, showcasing personal creativity.
[AiBase Summary:]
🎉 Users can upload a frontal photo and choose a template to generate New Year AI portraits with one click.
🎨 Supports custom video uploads, allowing users to create personalized works using the "Template DIY" feature.
🌟 The app offers a variety of template styles to meet different users' personalized needs.
6. $500 Billion Investment! Trump Announces Joint Venture "Interstellar Gateway" with Tech Giants like OpenAI
Trump announced the establishment of a $500 billion joint venture called "Interstellar Gateway" at the White House, aimed at creating a network of data centers and computing infrastructure to promote the development of artificial intelligence. This project is expected to create over 100,000 jobs and is being expedited through an emergency declaration, marking a significant shift in U.S. policy on artificial intelligence.
[AiBase Summary:]
🌟 Trump announced the establishment of a $500 billion AI joint venture "Interstellar Gateway," aimed at building a network of data centers.
💼 The project is expected to create over 100,000 jobs in the U.S. and rapidly advance infrastructure development.
⚡ Trump will accelerate the project's progress through an emergency declaration, while also revoking previous executive orders on AI safety.
7. NetEase Youdao's "Ziyue-o1" Reasoning Model Officially Open-Sourced
NetEase Youdao launched the country's first step-by-step explanation reasoning model "Ziyue-o1" on January 22, 2025, and announced its open-source status. This 14B lightweight model is designed for consumer-grade graphics cards and can operate stably on low-memory devices. Through chain-of-thought techniques, this model provides detailed problem-solving processes and logical reasoning, helping users better understand knowledge.
[AiBase Summary:]
🖥️ "Ziyue-o1" is the first step-by-step explanation reasoning model in the country, designed for consumer-grade graphics cards, operating stably on low-memory devices.
📚 The model provides detailed problem-solving processes through chain-of-thought techniques, helping users better understand and master knowledge, enhancing learning effectiveness.
🔍 The open-source nature allows developers to leverage the model for innovation, promoting the development of AI technology in education.
Details link: https://confucius-o1-demo.youdao.com/
8. Microsoft and OpenAI Adjust Cooperation, Opening Competitor Computing Resources
Microsoft's partnership with OpenAI has undergone significant adjustments, allowing OpenAI to access competitor computing resources, breaking its reliance on Microsoft Azure. This change provides OpenAI with greater flexibility in research and model training. The new agreement also adjusts the exclusivity terms of computing resources, granting Microsoft a priority purchase right.
[AiBase Summary:]
🌟 OpenAI will have access to competitor computing resources, breaking the limitations of relying solely on Microsoft.
🤝 The new agreement adjusts the exclusivity of computing resources, granting Microsoft a priority purchase right.
💼 The relationship between the two has fluctuated, with OpenAI collaborating with other companies to meet computing demands.
9. Reports: ByteDance Plans to Invest $12 Billion in AI Chips by 2025
ByteDance is increasing its investment in the field of artificial intelligence to enhance its competitiveness. It plans to invest over $12 billion in AI chip research and procurement by 2025, demonstrating its emphasis on AI technology. Additionally, the company's budget for AI infrastructure this year is 40 billion RMB, double that of last year, and it plans to invest $6.8 billion in overseas markets, mainly for procuring NVIDIA chips to enhance the training capacity of foundational models.
[AiBase Summary:]
🌟 ByteDance plans to invest over $12 billion in developing AI chips by 2025, highlighting its focus on AI technology.
💰 This year, ByteDance has allocated a budget of 40 billion RMB for AI infrastructure, a twofold increase compared to last year.
🌍 The company also plans to invest $6.8 billion overseas to procure NVIDIA chips, enhancing foundational model training capabilities.
10. French AI Startup Mistral Plans to Go Public, Not to be Acquired
The French startup Mistral AI recently announced plans to go public, marking a new development in the open-source AI field. CEO Arthur Mense stated clearly that the company is not for sale, despite rumors of a Microsoft acquisition. Since its establishment in 2023, Mistral has experienced rapid growth, has ample funding, and plans to expand its business in the Asia-Pacific region, with its first office to be located in Singapore.
[AiBase Summary:]
💼 Mistral AI plans to go public, with the CEO stating the company is not for sale.
🌏 To expand its market, Mistral will establish an office in Singapore for the Asia-Pacific region.
💰 Mistral has experienced rapid growth since its establishment, with sufficient funding and no need for financing.
11. Reports Indicate Google to Invest Over $1 Billion in Anthropic Again
Google recently announced that it will invest over $1 billion in the AI startup Anthropic, demonstrating its confidence in the company. Anthropic is currently undergoing a funding round of nearly $2 billion, with its valuation expected to reach $60 billion. As competition in the AI industry intensifies, developing large-scale language models requires substantial funding and top talent, placing Anthropic under greater pressure in the market.
[AiBase Summary:]
💰 Google will invest over $1 billion in the AI startup Anthropic, showing confidence in its prospects.
📈 Anthropic is undergoing a funding round of nearly $2 billion, with its valuation expected to reach $60 billion.
🖥️ The intensifying competition in the AI industry requires massive funding and top talent to develop large-scale language models.