The AI Tool Collection uses a monthly format of 3 issues, updating the latest AI industry news daily. This issue is the collection of AI industry news for June Issue 3 (June 21 to June 30).
June 30
- Startup Inflection AI raised $1.3 billion from Bill Gates, Nvidia, LinkedIn co-founder Hoffman, and others, one of the largest funding rounds in Silicon Valley during the current AI boom. The company developed the chatbot tool Pi, and its current valuation has reached $4 billion. (Read the full article:Wallstreetcn）
- The AI startup focused on providing content creation services for enterprises, Typeface , raised $100 million from supporters including the investment arms of Salesforce and Alphabet. The deal values Typeface at $1 billion, bringing its total funding to $165 million. (News source:Cailian Press）
- Yesterday, led by Shanghai AI Laboratory and jointly released with top domestic and international research institutions, universities, and hospitals, the world's first medical multimodal foundation model group "OpenMEDLab" was launched. (Read the full article:Shanghai AI Laboratory Official Account）
- "SenseAvatar" digital human video generation platform was officially launched, fully available on the iOS / Android app markets. It is the industry's first AIGC full-stack video generation platform. (Read the full article:SenseTime Official Account）
- Credit rating and research company Moody's announced yesterday a new strategic partnership with Microsoft, using Microsoft Azure OpenAI to create an AI assistant to help clients analyze the large amount of information needed to assess risk. (Read the full article:National Business Daily）
- According to an updated post on the official Bing blog, Microsoft announced the integration of Microsoft Shopping features into Bing Search, Bing Chat, and the Edge browser. (Read the full article:PingWest）
- Microsoft today pushed the Build 23493 update to Windows 11 Dev preview users, one of the updates being the launch of the AI assistant Windows Copilot. Users can launch Windows Copilot by clicking a new button on the taskbar (or Win + C). (Read the full article:ITHome）
June 29
- Meituan announced that it has completed the acquisition of 100% equity of Light Year Beyond's domestic and overseas entities, with an acquisition price of approximately RMB 2.06 billion. Previously, Light Year Beyond co-founder Wang Huiwen left his post for medical treatment and recuperation due to health reasons. (Read the full article:JRJ）
- Domestic online document collaboration platform wolai has started internal testing of AI features, which can assist with writing, text processing and editing, code processing, creating mind maps, tables, etc. Individual Pro and above workspace versions can apply for the wol.AI internal test. (Read the full article:Welcome to use wol.ai）
- With the development of generative AI technology, content creation has become simpler and more automated. To help workers master the basic concepts and applications of AI, Microsoft today announced a new AI training program, offering free online courses and certificates. (Read the full article:ITHome）
- Well-known domestic interior design software Kujiale announced the launch of "Kujiale AI", which can help users quickly produce interior design renderings. Currently, Kujiale AI can be experienced on WeChat Mini Program and its official website. (Read the full article:Kujiale Official Account）
- The 2023 World Artificial Intelligence Conference will be held in Shanghai from July 6 to 8. The theme of this edition is "Intelligent Connectivity, Generating the Future." It will be held at the Shanghai Expo Center and the Expo Exhibition Hall, with sub-venues in Zhangjiang, Pudong, and West Bund, Xuhui, Shanghai. (Read the full article:China Securities Network）
- Dingdang Health officially released the large model Dingdang HealthGPT and launched application-oriented pharmaceutical AI products developed based on it—Dingdang Pharmacist and Nutritionist AI assistants. (Read the full article:Xinhuanet）
- OpenAI, the developer of ChatGPT, announced on its official website that it has established a new office in London, UK, which is also the first overseas branch of this American company. (Read the full article:PingWest）
- According to Nikkei News, SoftBank, the mobile communications business subsidiary of Japan's SoftBank Group, will independently develop generative artificial intelligence and will also install a supercomputer for development. (News source:Cailian Press）
June 28
- This afternoon, ByteDance's Volcano Engine held a launch event, officially announcing for the first time its R&D layout and cooperation progress in the field of large models, and released the large model service platform Volcano Ark, providing comprehensive functions and services such as model training, inference, evaluation, and fine-tuning. (Read the full article:GeekPark）
- During the Shanghai World Mobile Communications Conference, China Unicom released the image-text large model "Honghu Image-Text Large Model 1.0," the first large model oriented toward operators' value-added services. The model currently has two versions with 800 million training parameters and 2 billion training parameters, and can achieve functions such as text-to-image, video editing, and image-to-image. (Read the full article:Yicai）
- Unity, the world's largest game engine, announced that it will launch two new AI tools, Unity Sentis and Unity Muse. Unity Sentis is a cross-platform runtime inference engine that can help deploy AI models in any Unity project, while Unity Muse includes tools for accelerating real-time 3D content creation. (Read the full article:Unity Blog）
- The iOS version of the ChatGPT app has launched another new feature, namely the integration of Microsoft's Bing search engine, but this feature is only available to paying users, enabling the use of Bing search online to provide more accurate answers based on the latest data. (Read the full article:IT Home）
- PowerLaw AI and Zhipu AI jointly released PowerLawGLM, a vertical large model for law based on a Chinese hundred-billion-scale large model, with rich legal knowledge and legal language understanding capabilities for Chinese legal scenarios. (News source:TMTPost）
- JetBrains recently published a blog post announcing that this week's EAP versions of all IntelliJ-based IDEs and .NET tools will have AI assistant functionality, driven by JetBrains' self-developed AI model together with OpenAI services. (Read the full article:JetBrains Blog）
- Domestic HR service provider Moka officially released the first AI-native HR SaaS overall solution as well as the AI HR partner for the AGI era—Moka Eva. This is a milestone in Moka's comprehensive embrace of its AI strategic layout and will also redefine new trends in the future of HR digitalization. (Read the full article:Machine Heart）
- Booking, an online travel, accommodation, and hotel reservation website, announced the launch of an AI trip planning assistant, which will be launched in beta for some US travelers starting today. The AI travel assistant is based on Booking's existing machine learning models and is partly supported by OpenAI's ChatGPT to provide users with a conversational experience. (Read the full article:PRNewsWire）
- Online music sampling material website Splice has launched the AI music tool Splice Create, which uses AI to intelligently search over 1.9 million sampling materials on its platform and combine them together. This feature can be used for free, while exporting requires payment. (Read the full article:RA.co）
June 27
- China's largest cloud-based intelligent computing platform among universities—the Fudan University Intelligent Computing Platform CFFF (Computer for the Future at Fudan)—"Qiewen" No. 1 and "Jinsi" No. 1 were officially completed, launched, and opened today. The platform was jointly built by Fudan University, Alibaba Cloud, and China Telecom, and uses an advanced public cloud model to provide over a thousand-card parallel intelligent computing, supporting large model training with hundreds of billions of parameters. (Read the full article:Fudan Development Institute）
- Baidu CTO Wang Haifeng said that the ERNIE large model version 3.5 added a plug-in mechanism. Subsequently, ERNIE Bot will release more high-quality official Baidu and third-party plug-ins, and will gradually open the plug-in ecosystem to third-party developers. (News source:36Kr）
- Big data giant Databricks announced that it has signed a definitive agreement to acquire San Francisco-based generative AI startup MosaicML for $1.3 billion. (Read the full article:PingWest）
- MongoDB announced that it will cooperate with Google Cloud to accelerate developers' use of generative AI and the development of new types of applications. Developers can leverage MongoDB Atlas and its integration with Google Cloud Vertex AI large language models (LLMs) to speed up the software development process. (Read the full article:Sina Tech）
- Google recently launched an anti-money laundering tool, AML AI, combined with generative AI. It can be used to identify suspicious money laundering behavior in the financial industry and generate analysis reports that comply with industry standards. (Read the full article:ITHome）
- News website rating tool NewsGuard released the misinformation monitoring results for June 2023, saying that 141 brands are providing advertising revenue to low-quality artificial intelligence (AI)-generated websites to support the development of these unreliable websites. (Read the full article:The Paper）
- Cloud database company Snowflake announced that it will cooperate with Nvidia to tailor AI models for enterprises. Executives said this will ensure the security of proprietary data, and disclosed that it will cooperate with Microsoft to bring generative AI into Snowflake's data cloud platform. (Read the full article:MarketWatch）
- Radio Television Hong Kong's AI weather anchor Aida has officially debuted. Starting from June 26, she will report on local and global weather conditions in the station's "RTHK Weather Bulletin" program. Aida took about 3 months to produce and prepare, and was created using AI virtual avatar technology. (Read the full article:ITHome）
June 26
- Tencent Technology conducted an exclusive interview with legendary AI journalist Cade Metz. In the interview, Cade said that the industry has no moat, and China will quickly catch up with OpenAI, because the overall gap between the two in knowledge, funding, and resources is not large. (Read the full article:Tencent Technology）
- Li Yanhong revealed today while attending the World Internet Conference Nishan Dialogue on Digital Civilization that Baidu's ERNIE large model has been iterated to version 3.5. Compared with version 3.0, training speed has increased by 2 times, inference speed has increased by 17 times, and the model effect has improved by more than 50% cumulatively. In terms of data quality, generation effect, and content security, all have improved significantly. (News source:36Kr）
- The 5th Global Artificial Intelligence Product Application Expo opened in Suzhou yesterday. This year's Global AI Expo held 1 high-standard opening ceremony and main forum, set up 6 major brand exhibition areas, held 2 AI innovation and entrepreneurship competitions, presented 3 major awards, and released more than 20 new AI products. (Read the full article:Modern Express）
- The professional social networking platform LinkedIn earlier launched ad creation support, allowing advertisers to quickly generate marketing copy through generative AI. Now, the platform has begun applying generative AI to a broader scope — allowing users to directly use the technology to publish posts. (Read the full article:IT Home）
June 25
- According to a Gartner survey, only 8% of customers used chatbots in their most recent customer service experience. Of those, only 25% said they would use chatbots again in the future. (Read the full article:Geeknet）
- According to a report by Yomiuri Shimbun, Japan's Ministry of Education, Culture, Sports, Science and Technology recently stated that it has drafted guidelines for schools regarding generative AI represented by ChatGPT, and will release them as soon as possible after hearing expert opinions. (Read the full article:IT Home）
June 24
- Midjourney Launched version 5.2. This update includes improvements to aesthetic style, generation coherence, text understanding, prompt analysis, high variation mode, high resolution mode, zoom-out extended images, and more. (Read the full article:Machine Heart）
- YouTube is currently testing a new tool that can help creators use artificial intelligence to automatically dub videos into other languages. YouTube is collaborating with the AI dubbing service team Aloud, which is part of Google's internal incubator Area 120. (Read the full article:The Verge）
- In Marvel's new series Secret Invasion, the opening animation of the series was automatically generated using the latest AI technology. Although many fans marveled at the opening visuals, quite a few fans were also dissatisfied with it. (Read the full article:Tencent）
- Amazon announced this week that it will invest $100 million to establish an AI innovation center to help enterprise customers and partners plan and deploy AI technology. Through this technology, Amazon hopes to promote its AI cloud services, catch up with competitors from industry rivals such as Microsoft and Google, and expand the appeal of its AWS business to enterprises. (Read the full article:IT Home）
June 23
- Stability AI recently issued a press release announcing the launch of the SDXL 0.9 version update, upgrading the Stable Diffusion text-to-image model. The generated images are more realistic and more textured, and can correctly generate five human fingers. (Read the full article:Stability AI Blog）
- Opera announced that it is officially opening Aria AI to users of the Opera browser on Android and desktop. Aria is an artificial intelligence service integrated into the Opera One desktop browser and its Android browser beta. (Read the full article:Opera Press Release）
- Beatles member Paul McCartney recently responded to the use of artificial intelligence to create a new Beatles song, and said that 'nothing has been created artificially or synthetically,' but still did not reveal the song's title or any lyrics. (News source: The Guardian)
- The All England Lawn Tennis Club and IBM recently announced that Wimbledon, one of the four Grand Slam tennis tournaments, which will be played from July 3 to July 16, will use AI technology to provide commentary and predictive analysis of players' chances of winning. (Read the full article:IT Home）
- According to the latest data released by the job recruitment website Indeed, the number of generative AI-related jobs in the United States increased by 20% month-on-month in May this year, indicating that more and more American companies hope to get on the trend and deploy or use this technology to drive corporate development. (Read the full article:IT之家）
June 22
- Microsoft announced a "milestone" progress in the field of quantum computing and released the roadmap for Microsoft's quantum supercomputer. Microsoft's Senior Vice President of Quantum Development said that the construction of the quantum supercomputer is expected to be completed within 10 years. (Read the full article:爱范儿）
- According to media reports, OpenAI, the company behind the chatbot ChatGPT, lobbied EU officials in an effort to influence the upcoming Artificial Intelligence Act, and succeeded. (Read the full article:腾讯科技）
- Figma announced the acquisition of the AI design startup Diagram. Figma's Vice President of Product Design shared his views on how artificial intelligence affects design and said that artificial intelligence has always been an area of greater focus for Figma. (Read the full article:Figma Blog）
- Dropbox launched two AI tools. One is called Dropbox AI, which helps users intelligently summarize and query document content; the other is called Dropbox Dash, an AI-powered universal search tool. (Read the full article:The Verge）
- Databricks recently launched Lakehouse Apps,which is the collective name for native software packaged using a series of models launched by the company. The company also launched an "AI model sharing mechanism," namelya series of models open-sourced by the company and third parties, which users can quickly deploy in Lakehouse Apps and their own content. (Read the full article:IT之家）
June 21
- Microsoft recently updated Bing Chat, adding a voice input feature. This feature allows users to use voice input for conversations, and currently supports Mandarin, English, Japanese, French, and German. Microsoft said that more new language types will be added in the future. (Read the full article:品玩）
- A reporter from the Science and Technology Innovation Board Daily learned exclusively that Ant Group's technology research and development team is independently developing a language and multimodal large model, internally named "Zhenyi." (News source:科创板日报）
- OpenAI is considering creating an app store for artificial intelligence software, where customers can sell customized AI models to enterprises. (Read the full article:IT之家）
- According to Gizmodo, DeepMind co-founder Mustafa Suleyman recently said that in his new book he proposed a completely new testing method to judge whether artificial intelligence possesses a human level of intelligence. (Read the full article:Gizmodo）
- JD.com officially announced today that the "2023 JD Global Technology Explorer Conference and JD Cloud Summit" will be held in Beijing on July 13. According to the introduction, this conference is themed "Leap · Industrial Intelligence," focuses on the deep integration of large models and industry, and will release the JD large model, launch a new generation of digital infrastructure, and upgrade products and solutions. (Read the full article:京东科技公众号）
- Snap CEO recently said that since its launch, about 150 million users have used Snapchat's chatbot My AI and sent it more than 10 billion messages. (Read the full article:界面新闻）
- An investigation report from Group IB shows that information of more than 100,000 ChatGPT users was stolen by hackers and sold. The report points out that the Asia-Pacific region has the most information. By region, the Asia-Pacific region has 40,999 records; the Middle East and Africa region has 24,925 records; Europe has 16,951 records. (Read the full article:PingWest）