Too long to look at the version
Use large models to automatically take notes, extract PPT, organize interviews, Alibaba Cloud "Tongyi Listening Understanding" open public test, 20 hours of free time waiting for you to get
Alibaba Cloud has reached a cooperation agreement with the three districts of Guangzhou to jointly build a metaverse and AI technology industry
OPPO launches the MR headset "OPPO MR Glass" developer version
Xiaoice the first batch of GPT Internet celebrity AI clones were launched in advance, including "Hanzo Forest" and so on
"Chinese version of Adobe" Wondershare Technology's ink knife launched AI functions
OpenAI will train models in new ways to combat AI "illusions"
Australia plans to enact stricter laws on AI
Generative AI startup 3DFY.ai introduces the 3D Model Platform to create 3D models in natural language
Renault Korea plans to create a metaverse car experience showroom
The person in charge of the MOSS system of Fudan University: MOSS models with greater computing power and stronger reasoning capabilities will be launched in a planned manner
OpenAI founders share OpenAI's near-term roadmap: reducing GPT-4 API costs and more
Google has invested in AI startup Runway
Agricultural AI company "Fengma Technology" completed tens of millions of investment, led by Zhencheng Investment and Century Jinyuan Group
AI startup MiniMax is close to completing a new round of financing of more than $250 million, with Tencent-affiliated entities participating
Bright Dairy launched three digital asset images
Metaverse headlines
Use large models to automatically take notes, extract PPT, organize interviews, Alibaba Cloud "Tongyi Listening Understanding" open public test, 20 hours of free time waiting for you to get
On June 1, Alibaba Cloud announced the progress of Tongyi Big Model, and the new AI product "Tongyi Listening Understanding" focusing on audio and video content was officially launched, becoming the first large model application product open for public testing in China.
Zhou Jingren, CTO of Alibaba Cloud, introduced that Tingwu is a work-learning AI assistant that aims at audio and video content scenarios with high knowledge added value, such as meetings, classes, interviews, training, interviews, live broadcasts, watching videos, listening to podcasts, etc., and can quickly refine and precipitate knowledge through the latest AI technologies such as large models.
Tongyi Listening Access to the understanding and summarization capabilities of the Tongyi Qianwen model, it can become a powerful AI assistant in users' work and learning, helping to efficiently complete the transcription, retrieval, abstracting and sorting of audio and video content anytime, anywhere, such as automatically taking notes, organizing interviews, extracting PPTs, etc. with large models.
Finally, Thoughtful Jun will also give benefits to everyone! Scan the QR code on the poster, complete the registration of Tongyi Hearing, enter the exclusive password "Wander in the Metaverse" in "User-Equity Gas Station", and you can receive 20 hours of free transcription time!
There are 300 chances to claim the password, on a first-come, first-served basis! Tip, each user can stack different passwords and receive up to 100 hours of free credit!
Domestic dynamics
Alibaba Cloud has reached a cooperation agreement with the three districts of Guangzhou to jointly build a metaverse and AI technology industry
News on June 1, according to CaiLian News Agency, Alibaba Cloud and Guangzhou Haizhu, Baiyun and Nansha three districts today announced cooperation.
Among them, Haizhu District of Guangzhou and Alibaba Cloud cooperated to build a new computing power infrastructure for smart cities; Alibaba Cloud signed a cooperation agreement with the Baiyun District Government to establish partnerships in the fields of smart city, artificial intelligence, and metaverse, and promote the development of a new generation of AI technology industry ecology, including large models, in Baiyun. Nansha District of Guangzhou, AVIC Cloud and Alibaba Cloud will jointly build a digital foundation and industrial digital development platform for cross-border overseas industries facing the world.
OPPO launches the MR headset "OPPO MR Glass" developer version
Gyro Technology news, recently, OPPO launched an MR headset, called "OPPO MR Glass Developer Edition", equipped with the Snapdragon XR2 + Plus chipset. This device is designed for developers to create applications and explore the future of MR technology.
It is reported that OPPO hopes to attract developers to create more high-quality applications based on MR technology. The company will make the headset available to developers in China later this year. But for the time being, there are no plans to put the device into commercial use.
Xiaoice the first batch of GPT Internet celebrity AI clones were launched in advance, including "Hanzo Forest" and so on
According to the IT House news on June 1, two weeks ago, Xiaoice company announced the launch of the "Clone Project" in the Chinese and Japanese markets, claiming that as long as three minutes of data collection, it can create AI clones derived from their personality, skills, voice, and appearance for celebrities, experts and scholars or ordinary people.
"Chinese version of Adobe" Wondershare Technology's ink knife launched AI functions
Recently, the world's first ink knife AI/Mockitt AI under Wondershare Technology, a listed company in creative software, launched AI functions such as AI-generated prototype page, AI-generated prototype components, AI text intelligent filling, and AI dialogue, becoming the world's first integrated design collaboration platform that provides AI functions such as AI automatic generation of prototype components and AI intelligent filling of AI text.
At present, Ink Knife is the design collaboration platform with the largest number of domestic users, serving many well-known enterprises at home and abroad, the number of platform users has exceeded 3.5 million, and the user base of serving enterprises exceeds 130,000.
Not long ago, the overseas version of Mockitt has taken the lead in accessing the GPT model and officially launched MockittAssistant. In addition, Ink Knife V8.0 was launched earlier this year, supporting advanced functions such as infinite canvas, making operation smoother and collaboration more real-time.
Overseas news
OpenAI will train models in new ways to combat AI "illusions"
On June 1, OpenAI's latest research paper showed that the company is using a new method to train artificial intelligence models to combat artificial intelligence "illusions".
AI illusion refers to the ability of AI models to generate content that is not based on any real-world data, but is the product of the model's own imagination. There are concerns about the potential problems that this hallucination can bring, including moral, social, and practical issues.
When OpenAI's chatbot ChatGPT or Google's competitor Bard are purely fabricating false information, AI hallucinations occur as if they are gushing about facts. Some independent experts have expressed skepticism about OpenAI's effectiveness in doing so.
Australia plans to enact stricter laws on AI
According to the Golden Ten, the Australian side said that due to concerns about the possible abuse of AI technology, it plans to regulate the technology, including possible prohibition of deep counterfeiting and content that seems real but false.
A previous report by Australia's National Science and Technology Council showed that AI-generated content could be misused in parliamentary consultations, sparking numerous submissions and misleading public opinion. Australia's Minister for Industry and Science, Ed Husić, said: "The government has a clear role to play in recognising the risks... and impose restrictive measures. ”
Generative AI startup 3DFY.ai introduces the 3D Model Platform to create 3D models in natural language
According to Sina Finance, generative AI startup 3DFY.ai has launched a 3D model platform, 3DFY Prompt. Similar to Midjourney, users can create realistic 3D models from natural language text, with detailed text descriptions and a choice of seed number (the more the better).
Currently, 3DFY.ai is in open beta and can be tried for free.
Renault Korea plans to create a metaverse car experience showroom
According to Sina VR, Renault Korea announced that it plans to build and operate a car experience metaverse space called "Renault Korea Center" this fall.
The metaverse space will be operated by K-verse, which has entered The Sandbox, a metaverse platform dedicated to Korean brands.
Metaverse perspective
The person in charge of the MOSS system of Fudan University: MOSS models with greater computing power and stronger reasoning capabilities will be launched in a planned manner
According to a report by Shanghai Securities News on June 1, Qiu Xipeng, head of the MOSS system at Fudan University, said on "the performance of MOSS in the past month of operation" that the MOSS system currently ranks high in the number of downloads in the open source community, and the number of following users on the GitHub platform has exceeded 10,000.
OpenAI founders share OpenAI's near-term roadmap: reducing GPT-4 API costs, etc
Sam Altman, CEO of OpenAI, said at the closed-door seminar that OpenAI is currently severely constrained by GPUs, causing them to postpone many short-term plans, and most of the problems with ChatGPT reliability and speed are due to GPU resource shortages.
In addition, OpenAI's near-term roadmap will reduce GPT-4 API costs; Longer ChatGPT context window (up to 1 million tokens), and in the future there will be a version of the API that remembers the conversation history; GPT-4's multimodal capabilities won't be made public until 2024, and the visual version of GPT-4 can't be extended to everyone until more GPU resources are available.
OpenAI is considering open source GPT-3, and part of the reason they haven't done it yet is because they don't feel many people and companies have the ability to properly manage such large, large language models.
Many recent articles claim that "the era of giant AI models is over" is not true, OpenAI's internal data shows that the law of scale proportional to performance still holds, and OpenAI's model size may double or triple every year (multiple sources suggest that GPT-4 parameter size is 1,000 billion), rather than increasing by many orders of magnitude.
Financing news
Google has invested in AI startup Runway
According to two people familiar with the matter, Google invested in Runway, a New York-based artificial intelligence startup, and including the new funding, the company was valued at $1.5 billion. Runway allows users to leverage artificial intelligence to generate videos from text descriptions.
The investment is part of a round of about $100 million in funding that highlights stiff competition among cloud vendors to get close to companies with cutting-edge AI services that could become major cloud customers or acquisition targets in the future.
Agricultural AI company "Fengma Technology" completed tens of millions of investment, led by Zhencheng Investment and Century Jinyuan Group
According to the "Science and Technology Innovation Board Daily" reported on June 1, agricultural AI technology company Fengma Technology completed a new round of tens of millions of yuan financing.
It is reported that this round of financing is jointly led by Zhencheng Investment and Century Jinyuan Group. Fengma Technology, founded in 2019, was incubated and established by Turing Artificial Intelligence Research Institute, and has successively received investment from Turing Venture Capital, Kaifeng Venture Capital, IMO Ventures and Chixing Venture Capital.
After the completion of the new round of financing, Fengma Technology will continue to invest in scientific and technological innovation and accelerate large-scale development. The company's management team is mainly from Bayer, Didi, Metro and other well-known enterprises at home and abroad, and integrates compound talents with multidisciplinary backgrounds such as agronomy, biology, artificial intelligence, software design, mechanical automation, and marketing.
AI startup MiniMax is close to completing a new round of financing of more than $250 million, with Tencent-affiliated entities participating
MiniMax, a Chinese AI startup focusing on general large models, is close to completing a round of financing of more than $250 million, and its valuation is expected to reach nearly $1.2 billion, according to people familiar with the matter. The company is developing an AI solution similar to ChatGPT.
According to two sources, MiniMax's latest round of funding has attracted new investors, including entities linked to Tencent. It is reported that MiniMax attracted Mihayou to participate in the previous financing. Other early investors include Yunqi Capital and MindTrend Capital.
Digital collections
Bright Dairy launched three digital asset images
On June 1, as the first dairy enterprise in China to build its own digital asset issuance platform, Bright Dairy first launched three digital asset images: "Changyou Rabbit", "Suixiao" and "Bright Quality Rabbit". Each one carries multiple surprise rights: "Changyou Rabbit" can be exchanged for Guangming Changyou 0 belly flavored ™ fermented milk (original flavor), a variety of Changyou Rabbit peripheral gifts and other benefits.