laitimes

Xueersi responded to the "AI big model stealing data" incident, iFLYTEK Spark big model iOS closed beta version was launched, The Beatles used AI to extract Lennon's voice... | Metaverse refers to North Broadcast

author:Thoughtful client

Too long to look at the version

Xueersi responded to the "Pen God Composition said that its AI large model stole data" incident: the calls were all in line with the contract requirements

iFLYTEK Xinghuo Cognitive Large Model iOS closed beta version was officially launched

Hangzhou Asian Organizing Committee: Intelligent preparation of the Asian Games throughout the process, launching the first digital torchbearer in the history of the Asian Games

Tencent RoboticsX uses pre-trained AI models and reinforcement learning to improve the control of the robot dog Max

Mei Tao, former vice president of JD.com, created HiDream.ai with the goal of surpassing the latest version of Stable Diffusion on the base model by the end of the year

Obtained the new AI application software development of its business scope

The Beatles, using AI to extract Lennon's voice, will release its final single this year

Meta will add text-based "World Chat" to Horizon Worlds

Google's Gmail "Help Me Write Email" AI feature began to be rolled out to Android and Apple devices

Report: 70% of programmers have programmed with various AI tools

Report: Generative AI could add $4.4 trillion annually to the global economy

Research says ChatGPT doesn't understand humor

Avatar platform Hyper completes $3.6 million in seed funding

French AI startup Mistral AI completed a €105 million seed round of funding

British startup Synthesia recently raised around $90 million to promote the AIGC unicorn

American rapper Snoop Dogg has launched a new NFT collection for his touring event

Metaverse headlines

Xueersi responded to the "Pen God Composition said that its AI large model stole data" incident: the calls were all in line with the contract requirements

IT House news on June 14, after Penshen Composition released a statement on the infringement of the "learning and thinking" AI large model. Penshen Composition believes that Xueersi uses "crawlers" to train using the server data of the Penshen Composition App, and demands that Xueersi pay 1 yuan in compensation, publicly apologize and delete the data that has been obtained.

Xueersi replied, claiming that "the data calls were consistent with the pre-agreed contractual requirements".

Xueersi responded to the "AI big model stealing data" incident, iFLYTEK Spark big model iOS closed beta version was launched, The Beatles used AI to extract Lennon's voice... | Metaverse refers to North Broadcast

Domestic dynamics

iFLYTEK Xinghuo Cognitive Large Model iOS closed beta version was officially launched

According to iFLYTEK official micro-news, on June 14, the internal beta version of iFLYTEK Xinghuo cognitive large model iOS was officially launched, which preemptively realized the full coverage of "PC, iOS, Android, Mini Programs and H5" mainstream systems in the domestic large model, and supported cross-device history synchronization, allowing users to switch freely on mainstream system platforms.

Hangzhou Asian Organizing Committee: Intelligent preparation of the Asian Games throughout the process, launching the first digital torchbearer in the history of the Asian Games

According to a report by Cai Lian News Agency on June 14, the State Council's new office held a press conference today on the preparations for the Asian Para Games in Hangzhou. Hangzhou Mayor Yao Gaoyuan said at the meeting that 100 days later, the 19th Asian Games will be grandly opened in Hangzhou, which is the third time that Chinese cities have hosted the Asian Games after Beijing and Guangzhou.

The whole process of intelligent preparation for the Asian Games, iteratively upgraded the digital competition platform "Asian Games Online", created a one-stop digital viewing service platform - "Smart Asian Games One-stop Pass", launched the first digital torchbearer in the history of the Asian Games, launched the Asian Games Pass, landed autonomous driving, intelligent bus and other applications, improved the efficiency of preparation and service guarantee by digital means, and fully demonstrated the achievements of "Digital China" construction.

Tencent RoboticsX uses pre-trained AI models and reinforcement learning to improve the control of the robot dog Max

According to the "Science and Technology Innovation Board Daily", on June 14, Tencent RoboticsX RoboticsX Robot Lab announced the latest progress in agent research, which greatly improves the flexibility and autonomous decision-making ability of Tencent Robot Dog Max by applying cutting-edge pre-trained AI models and reinforcement learning technology to the field of robot control.

Xueersi responded to the "AI big model stealing data" incident, iFLYTEK Spark big model iOS closed beta version was launched, The Beatles used AI to extract Lennon's voice... | Metaverse refers to North Broadcast

Mei Tao, former vice president of JD.com, created HiDream.ai with the goal of surpassing the latest version of Stable Diffusion on the base model by the end of the year

According to the Heart of the Machine news, Jingdong AI bull Meitao resigned as vice president of JD.com Group before the Spring Festival this year and is determined to devote himself to big model entrepreneurship. A month later, HiDream.ai was established, and the company's direction was determined to be generative vision multimodal basic models and applications.

It is reported that HiDream.ai's model can not only support Wensheng diagrams, but also support Wensheng videos, Tusheng videos and Wensheng 3D content.

Among them, the Wensheng image product supports entering Prompt to generate pictures, and can also further edit the parts of any picture that you want to modify. For example, it is required to modify the background, subject, etc. of the generated image. The initial goal of Mei Tao's team is to surpass the latest version of Stable Diffusion in the base model by the end of the year, and catch up with MidJourney in the product.

Obtained the new AI application software development of its business scope

According to interface news reports, the Tianyancha App shows that on June 6, Beijing Get Information Technology Co., Ltd. underwent industrial and commercial changes, and its business scope added digital technology services, artificial intelligence application software development, artificial intelligence basic software development, etc.

Overseas news

The Beatles, using AI to extract Lennon's voice, will release its final single this year

According to the BBC, Beatles member Paul McCartney said in an interview that the band has extracted John Lennon's voice from old transects with the help of AI and will release a new single this year, which will also be the band's "last song." "We just finished it and it will be released this year".

McCartney didn't name the song specifically, but it was likely a 1978 piece by Lennon called "Now And Then." A year ago, McCartney received the demo from Lennon's widow, Yoko Ono. It was one of several songs from a tape labeled "To Paul" that Lennon made shortly before his death in 1980.

Xueersi responded to the "AI big model stealing data" incident, iFLYTEK Spark big model iOS closed beta version was launched, The Beatles used AI to extract Lennon's voice... | Metaverse refers to North Broadcast

Google's Gmail "Help Me Write Email" AI feature began to be rolled out to Android and Apple devices

According to the IT House news on June 14, at I/O 2023, Google showed Gmail's generative AI function "Help Me Write".

This feature was previously only available on the desktop, and Google is rolling out this feature to mobile devices with Android and Apple operating systems. After the user clicks the "Write for me" button in the lower right corner, they type natural language prompts into the specified text box, and the AI can then draft the email as requested.

After AI completes this work, users can refine the finished message content according to their needs, such as adjusting the message content to be more detailed, formal, or shorter.

Meta will add text-based "World Chat" to Horizon Worlds

Recently, Meta announced that it will introduce the "World Chat" function in Horizon Worlds, aiming to continue to develop the social factor in the metaverse and provide users with another way to interact in the VR space, according to Yuanli News Agency.

The new World Chat option is a text-based thread, similar to a DM chain, that will pop up on the screen in Horizon. This is in addition to the existing voice chat elements in Horizon and provides another complementary contact flow to enable more connectivity within the virtual realm. But there may be another problem in VR, which will make users feel like they are being violated in the headset.

To address this, Meta has added a range of security elements to World Chat, including proactively scanning and deleting messages that violate its VR code of conduct, as well as user controls to ensure everyone has the option to turn off communications.

Xueersi responded to the "AI big model stealing data" incident, iFLYTEK Spark big model iOS closed beta version was launched, The Beatles used AI to extract Lennon's voice... | Metaverse refers to North Broadcast

Institutional perspective

Report: 70% of programmers have programmed with various AI tools

According to IT House on June 14, the latest developer questionnaire published by programming Q&A website Stack Overflow shows that while programmers are skeptical about the accuracy of various AI-driven tools, it does not prevent the use of tools such as ChatGPT in the development process.

Report: Generative AI could add $4.4 trillion annually to the global economy

According to a report by the McKinsey Global Institute, generative AI will generate $4.4 trillion annually in revenue to the global economy, according to the New York Times.

According to the report, between 2030 and 2060, half of all jobs will be automated. McKinsey had previously predicted that AI would automate half of all work between 2035 and 2075, but generative AI accelerated the agency's predictions.

Research says ChatGPT doesn't understand humor

Recently, scientists in Germany discovered that ChatGPT is not good at telling jokes. In more than 1,000 tests, 25 old terriers were repeated 90% of the time. Therefore, ChatGPT, although funny, is not funny and cannot create hilarious original content. Although ChatGPT came out, it is widely believed that it is good at writing, chatting, etc.

However, the sense of humor belongs to one of the human emotions, and if ChatGPT understands humor one day, it means that it has emotions.

Financing news

Avatar platform Hyper completes $3.6 million in seed funding

Recently, Hyper, an San Francisco-based avatar platform, announced the completion of a $3.6 million seed round of funding, according to Yuanli News Agency. The round was led by Two Sigma Ventures, with participation from MakersFund, Amazon's AlexaFund, Trevor McFedries, Robin Raskza and Dan Romero. It is reported that the company will use the financing to develop existing businesses and develop new products with avatar technology as the core.

Virtual YouTubers, also known as VTubers, have grown in popularity in recent years. Inspired by animation powered by motion capture technology, they use computer-generated avatars to stream content on platforms like YouTube, Twitch, and interact with viewers in real time, such as being able to stream games, commentary, and more.

Currently, the platform has launched Hyper AI, a tool for creating characters. It creates characters that can be used as personal chatbots, storytellers, and even on their own when live streamers want to take a break but fans don't.

Xueersi responded to the "AI big model stealing data" incident, iFLYTEK Spark big model iOS closed beta version was launched, The Beatles used AI to extract Lennon's voice... | Metaverse refers to North Broadcast

French AI startup Mistral AI completed a €105 million seed round of funding

On June 14, according to Techcrunch, French artificial intelligence startup Mistral AI has completed a 105 million euro ($113 million) seed round of funding, led by Lightspeed Venture Partners, Xavier Niel, JCDecaux Holding, Rodolphe Saadé, Motier Ventures, La Famiglia, Headline, Exor Ventures, Sofina, First Minute Capital, and LocalGlobe participated. Such financing is valued at 240 million euros ($260 million), the sources said.

The company, a new company formed by researchers from former DeepMind and Meta, will try to compete with OpenAI and Google to develop generative AI infrastructure. Mistral's funding is one of the largest seed rounds ever raised by a European generative AI company.

British startup Synthesia recently raised around $90 million to promote the AIGC unicorn

According to CaiLian News, on June 13, local time, Synthesia, a digital media platform headquartered in London, England, announced that it had received about $90 million in the latest round of financing, with a valuation of $1 billion, officially promoted to a "unicorn" company.

Digital collections

American rapper Snoop Dogg has launched a new NFT collection for his touring event

News on June 14, American rapper Snoop Dogg announced the launch of a new NFT series "Snoop Dogg Passport" with a price of 0.025 ETH. The collection aims to provide holders with behind-the-scenes content from Snoop Dogg's exclusive tour event, while also receiving future NFT launches, new song releases, merchandise and event ticket discounts.

In addition, Snoop Dogg also announced the launch of the "Snoop Select" program, which will select NFT works from some crypto artists such as Coldie, Terrell Jones, and Alien Queen and make them available to Passport NFT holders for free.