Abstract: Google Deepmind has launched the artificial intelligence image generator Imagen 2. According to the introduction, Imagen 2 can generate the highest quality and most realistic images in all Google models to date, while strictly following user prompts; Baidu AI official WeChat account announced that the Wenxin One Word plugin mall is now officially online, containing plugins covering various usage scenarios such as office efficiency, multimodal content understanding and generation, and professional information retrieval.
Image source: Generated by Wujie AI
Hot News:
Nanjing Agricultural University develops the first national ancient book AI large model "Xunzi"
According to the official WeChat account of "Nanjing Agricultural University Alumni Association", recently, Professor Wang Dongbo's research team at the School of Information Management of Nanjing Agricultural University released the "Xunzi" ancient book large language model in Beijing.
It is reported that the "Xunzi" large model contains the vast majority of handed down ancient texts, including "Siku Quanshu", with a large corpus of over 2 billion characters. It is an intelligent tool launched by the school team with the support of the National Social Science Fund major project "Construction and Application Research of Cross-language Knowledge Base of Chinese Ancient Classics", in collaboration with Zhonghua Book Company's Gu Lian Company, specializing in the processing and research of ancient books. The model contains two parts: the base model XunziALLM and the dialogue model XunziChat. Users can also fine-tune the base model based on their own needs using local training data, thereby achieving superior processing performance in downstream tasks of ancient books.
Startup company Shiyun Technology applies for deregistration
Tianyancha shows that on December 11, Shiyun Technology (Shenzhen) Co., Ltd. applied for deregistration due to a resolution to dissolve.
It is reported that the company was established on November 2, 2020. Xu Zhuo is the legal representative. The AI generation tool HeyGen was launched by the company, which previously gained attention for creating videos such as "Let Taylor Swift Speak Authentic Chinese".
Midjourney has opened alpha version website testing to some users
On December 14, Midjourney confirmed in its Discord announcement channel that it has "started testing the alpha version website with image creation functionality". However, the company pointed out that the alpha version is only available to "users who have created over 10,000 images in Midjourney". The announcement also stated that it will be opened to more users within a month.
Google Deepmind launches artificial intelligence image generator Imagen 2
On December 13, Google Deepmind launched the artificial intelligence image generator Imagen 2. According to the introduction, Imagen 2 can generate the highest quality and most realistic images in all Google models to date, while strictly following user prompts.
Imagen 2 was initially provided to developers and cloud customers through Google Cloud's Imagen API. Google plans to provide Imagen 2 to end users, but first hopes to minimize potential risks and challenges. This includes security measures such as SynthID (a toolkit for tagging and identifying AI-generated content) and technical safeguards to restrict problematic outputs.
Mozilla launches new AI project Solo to help zero-coding users create websites with AI
Reported on December 14, Mozilla, the developer of the Firefox browser, recently launched a new project called Solo, aimed at users with no coding experience. Through integrating AI capabilities, websites created with Solo can achieve development effects comparable to professional developers.
Mozilla stated that the project is mainly aimed at small and medium-sized enterprises and individuals. In the official demonstration, users only need to input natural information in text, images, and layout aspects, and the website can be processed and processed into the desired website.
Baidu launches Wenxin One Word plugin mall, providing functions such as mind map, audio and video extraction
On the evening of December 13, the official WeChat account of Baidu AI announced that the Wenxin One Word plugin mall is now officially online, containing plugins covering various usage scenarios such as office efficiency, multimodal content understanding and generation, and professional information retrieval.
Users can easily achieve various multi-scenario multimodal needs such as PPT generation, audio and video extraction, and mind map creation through simple commands. The plugins currently launched include mind map tools, PPT generators, and "video assistant" that can extract key information from a URL.
Google introduces new AI model MedLM for healthcare
Google announced on Wednesday the launch of a new set of AI models for healthcare called MedLM, aimed at helping clinical doctors and researchers conduct complex research and summarize doctor-patient interactions. This move marks Google's latest attempt to monetize AI tools in the healthcare industry. The MedLM suite includes a large and medium AI model, both built on Med-PaLM 2. Med-PaLM 2 is a large language model trained on medical data announced by Google in March this year.
Google opens Gemini Pro to cloud computing customers, allowing developers to build applications with it
On Wednesday, Google announced a series of upgraded AI features aimed at providing better services to its cloud computing customers. It is reported that Google has released Gemini Pro for enterprises, allowing developers to build applications using Google's latest AI models. Gemini is a large AI system trained on a large amount of data, capable of generating new content according to user requirements.
Google Cloud customers can use Gemini Pro to create AI chatbots, easy-to-query inventory databases, and marketing presentations, among other applications. The company also emphasized that Gemini Pro will initially be provided free of charge to cloud customers, but with some limitations. However, Google stated that it plans to ensure that the pricing of its cloud AI products is "competitive".
According to the parameters released by Google, the text-based functions of Gemini Pro have a 4x lower input cost and a 2x lower output cost compared to its predecessor, the AI model PaLM 2, released in June.
Investment and Financing:
AI company Glean, founded by former Google search team, is seeking at least $200 million in financing with a pre-investment valuation of $2 billion
Reported on December 14, two informed sources revealed that the startup company Glean, founded by former Google search engineers, is in negotiations with investors, hoping to raise at least $200 million with a pre-investment valuation of $2 billion.
According to one of the sources, Kleiner Perkins has also invested in Glean's first three rounds of financing and is discussing jointly leading the latest financing. This financing will double Glean's valuation from May 2022, and as of this quarter, the company's annual recurring revenue has increased to over $30 million, at least three times higher than the same period last year.
Insiders: Legal AI startup Harvey is negotiating a new round of financing with a valuation of $700 million
Reported on December 14, two people familiar with the financing situation revealed that the legal artificial intelligence startup Harvey is negotiating a round of financing of $70-80 million, with a valuation including investment reaching $700 million. It is reported that this financing is more than four times the valuation announced in April.
Insiders said that venture capital firm Kleiner Perkins and individual investor Elad Gil will jointly lead this round of financing, the latter having invested in the company's seed and Series A rounds. In addition, existing investors Sequoia Capital and OpenAI Entrepreneurship Fund will also participate in this round of investment.
Great Reads:
"Pika is on fire, but AI video has not reached the 'GPT moment' yet"
Wensheng's video has always been regarded as the "Holy Grail" of multimodal AIGC, and while there are cool demos in the current AI video track, there are still many challenges facing the industry, similar to the practical experience of participating in the production of Oscar-winning films.
This article mainly discusses three key issues related to Wensheng's video:
- What is the technical route behind Wensheng's video?
- Why is it said that AI video has not yet reached the real GPT moment?
- Who has a leading advantage in the current industry competition?
https://www.aixinzhijie.com/article/6841020
"The cake is almost all divided by OpenAI, AI startups are in a financing dilemma"
According to overseas media reports, the investment train for AI startups is slowly slowing down. Some young startups have found that achieving the goals of a new round of financing has become increasingly difficult, whereas a year ago, investors would certainly have been eager to try.
https://www.aixinzhijie.com/article/6841022
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
