Welcome to the 【AI Daily】 column! Here is your daily guide to exploring the world of artificial intelligence. Every day, we present you with hot topics in the AI field, focusing on developers to help you gain insights into technological trends and understand innovative AI product applications.
Fresh AI products Click to learn more: https://top.aibase.com/
1. Google DeepMind releases Genie 2, generating infinite virtual worlds with one click
Google DeepMind recently launched the Genie2 model, which generates diverse and logically coherent virtual worlds in real-time through natural language commands, showcasing powerful AI training and interaction potential. Genie2 can create complex 3D environments and supports long-term interaction with users, significantly enriching AI training scenarios.
【AiBase Highlights:】
🌍 Genie2 can generate diverse virtual worlds in real-time through natural language commands, including complex 3D environments.
🤖 The model supports user interaction with AI agents for up to one minute, greatly enriching AI training scenarios.
📸 Starting from a single image prompt, Genie2 can quickly build a complete interactive world, even generating virtual scenes using real photos.
Details link: https://deepmind.google/discover/blog/genie-2-a-large-scale-foundation-world-model/
2. Fish Audio releases Fish Speech 1.5, adding 5 new languages and launching real-time seamless conversation feature
Fish Audio has recently made significant improvements to its Fish Speech 1.5 voice synthesis model. It not only supports five new languages but also features real-time seamless conversation capabilities, allowing for easier user interaction. The model has been trained for over 1 million hours in multiple languages, generating high-quality voice synthesis outputs and achieving second place in the anonymous TTS-Arena.
【AiBase Highlights:】
🎤 Zero-shot and few-shot voice synthesis: high-quality voice can be generated with just 10 to 30 seconds of voice samples.
🌍 Multi-language support: supports various languages including English, Japanese, Korean, and Chinese, breaking language barriers.
⚡ High accuracy and speed: achieves real-time voice synthesis on high-performance hardware with an error rate as low as 2%.
Details link: https://fish.audio/zh-CN/
3. Apple collaborates with Baidu: iPhone domestic version may integrate Wenxin 4.0 model
Apple has partnered with Baidu to integrate Baidu's AI large model Ernie 4.0 into the domestic version of the iPhone to enhance the user's AI experience. Despite facing some challenges during the adaptation process, Apple still has time for necessary testing and adjustments to ensure Apple Intelligence goes live as scheduled.
【AiBase Highlights:】
🤖 Apple collaborates with Baidu to integrate the AI large model Ernie 4.0 into the domestic version of the iPhone.
📈 Apple pays for the usage fees to enhance the AI experience for iPhone users.
📱 The iPhone 16 series is lagging behind in AI capabilities and needs to adapt to the market more quickly.
4. ByteDance's AI assistant Doubao now supports one-click generation of images with specified text
ByteDance's intelligent AI assistant Doubao has recently made significant upgrades to its text-to-image capabilities, allowing users to generate images containing specified text with one click. This new feature is a result of the efforts of the Doubao large model team, which integrates large language models and digital image technology, optimizing the learning ability of Chinese data and enhancing the ability to generate Chinese characters.
【AiBase Highlights:】
🆕 The new feature allows users to generate images with specified text with one click, enhancing operational convenience.
🔍 Doubao optimizes its learning and character generation capabilities for Chinese through the integration of large language models and digital image technology.
🚀 This upgrade demonstrates Doubao's in-depth exploration in the application of AI technology, looking forward to greater developments in the future.
5. Baidu: Wenxin Yiyan ranks first in user usage rate among generative AI products in China
As of June 2024, the user base of generative artificial intelligence in China has reached 230 million, with Wenxin Yiyan ranking first with an 11.5% usage rate. The daily call volume of the Wenxin large model has reached 1.5 billion, processing over 1.7 trillion tokens, reflecting its rapid development in technology and application. Conversational products account for 62% of the market share, with the Wenku AI feature being used over 2.8 billion times, attracting a large number of enterprises and developers to the agent platform.
【AiBase Highlights:】
📈 User base: As of June 2024, the user base of generative AI products in China has reached 230 million, accounting for 16.4% of the total population.
💡 Wenxin Yiyan: Baidu's Wenxin Yiyan ranks first in domestic netizens' usage rate at 11.5%, with a daily call volume of 1.5 billion.
🛠️ Wide application: Conversational products occupy 62% of the market share, with the Wenku AI feature used over 2.8 billion times, attracting 150,000 enterprises to the agent platform.
6. ChatGPT weekly active users exceed 300 million
OpenAI CEO Sam Altman announced at The New York Times' DealBook Summit that ChatGPT's weekly active users have surpassed 300 million, demonstrating its strong market appeal and user reliance. Since its launch in 2022, the number of ChatGPT users has experienced explosive growth, currently sending over 1 billion messages daily.
【AiBase Highlights:】
🌟 ChatGPT's weekly active users have exceeded 300 million, showcasing strong market attraction.
📈 More than 1 billion messages are sent through ChatGPT daily, increasing user reliance.
🛠 OpenAI continues to roll out new features to enhance user experience, including AI search engines and new interfaces.
7. OpenAI announces "12-day live event" new technologies will be unveiled but GPT-5 will not be released
OpenAI recently announced on X platform the launch of a 12-day live event to showcase new technologies and innovative projects. Despite the ongoing discussions about GPT-5, OpenAI has made it clear that its successor will not be released this year. CEO Sam Altman denied rumors about the Orion model and promised to release more technological innovations.
【AiBase Highlights:】
🔍 OpenAI will launch a 12-day live event tomorrow to showcase new technologies and innovative projects.
🚫 OpenAI clarifies that GPT-5 will not be released this year, with the CEO calling reports about the Orion model fictional.
📊 The O1 model supports processing approximately 200,000 tokens and has image analysis capabilities, while the Sora video model is also highly anticipated.
8. Apple CEO Tim Cook states that charging for Apple's AI features has never been considered: AI is a fundamental feature
Apple CEO Tim Cook recently stated in an interview that Apple has never considered charging for its AI features. He emphasized that AI should be regarded as a fundamental feature of smart devices, just like multi-touch, and is an indispensable part. This statement effectively dispels speculation about Apple potentially charging for its AI services.
【AiBase Highlights:】
💡 Cook stated that Apple will not charge for AI features, viewing AI as a fundamental feature.
📈 Apple's strategic focus has shifted from hardware to services, but AI services will not incur charges.
🗣️ Cook expressed concern over retirement issues, emphasizing his close connection with Apple.
9. Elon Musk plans to expand Colossus AI supercomputer tenfold
Elon Musk's AI company xAI recently announced an ambitious plan to expand the capacity of its Colossus AI supercomputer tenfold. This expansion will take place in Memphis, Tennessee, integrating over 1 million GPUs to meet the growing computational demands.
【AiBase Highlights:】
🌟 Elon Musk plans to expand the Colossus AI supercomputer tenfold, expecting to integrate over 1 million GPUs.
💼 This expansion will create new job opportunities in the Memphis area and boost local economic development.
🚀 xAI aims to enhance its market competitiveness and drive innovation in AI technology through increased computational power.
10. Zhu Xiaohu responds to arbitration controversy over the project "The Dark Side of the Moon" and Loop Intelligent shareholders' resolutions
Zhu Xiaohu responded to the arbitration incident between the founder of "The Dark Side of the Moon" and Loop Intelligent, emphasizing that the project was developed internally by Loop Intelligent and did not receive formal shareholder resolutions. The incident reveals the complexities of delayed shareholder resolutions and conflicts of interest, reflecting the legal and ethical responsibilities of Chinese startups during the spin-off process.
【AiBase Highlights:】
📈 "The Dark Side of the Moon" is a project developed internally by Loop Intelligent, which has not yet received formal shareholder resolutions.
⚠️ Yang Zhilin and Zhang Yutao did not obtain consent from investors to initiate financing, triggering strong reactions.
💼 Zhang Yutong was dismissed for concealing conflicts of interest, revealing the complexities of shareholder rights and responsibilities in startups.
11. Generative AI threatens creators' income: $35 billion loss expected in the next five years!
A recent global economic research report reveals that the rapid development of generative artificial intelligence poses a serious threat to the income of music and film creators. According to a study commissioned by the International Confederation of Societies of Authors and Composers (CISAC), by 2028, the income risk for creators is expected to reach $35 billion. Music creators face a 24% income loss, while film creators could lose up to 21%.
【AiBase Highlights:】
📉 Income loss for creators: Generative AI will lead to a 24% income loss for music creators and a 21% loss for film creators.
💰 Market expansion: The market for AI-generated services is expected to grow from $4.8 billion to $103 billion.
⚖️ Policy calls: Experts and creator organizations are urging the government to take measures to ensure the rights and income of creators are protected.
