Summary: According to the news on September 22nd, Google announced the launch of five new AI-assisted features for YouTube creators: AI-generated images, video enhancement (adding filters), video idea and suggestion generation based on audience preferences, automatic dubbing, and music accompaniment.
Image Source: Generated by Wujie AI tool
Hot News:
Liepin Releases "AIGC and Its Industry Chain Talent Demand Big Data Report 2023": Shenzhen Tops the Average Annual Salary at 470,000 RMB
According to China.com, Liepin recently released the "AIGC and Its Industry Chain Talent Demand Big Data Report 2023," analyzing the talent demand in the overall AIGC field and related industry chains in China. The report indicates that the average annual salary for new AIGC positions from January to August 2023 was 410,900 RMB, with a slight year-on-year increase of 0.88%. Among the four primary industries with the most new AIGC positions—IT/Internet/Gaming, Automotive, Electronics/Communications/Semiconductors, and Scientific Research Technology/Business Services—the average annual salaries for new AIGC positions were 431,300 RMB, 337,900 RMB, 413,600 RMB, and 448,800 RMB, respectively. In terms of the average annual salary for the top 10 cities with AIGC positions, Shenzhen, Beijing, and Shanghai ranked top three, all exceeding 440,000 RMB, at 474,200 RMB, 463,000 RMB, and 449,400 RMB, respectively. Hangzhou and Guangzhou ranked fourth and fifth in terms of salary, at 384,400 RMB and 357,200 RMB, respectively.
Qifull Technology Launches AI Expert Application "Yuzhi" Based on Qifull's Large Model
According to TechWeb, Qifull Technology announced that its large model application "Yuzhi" has achieved new breakthroughs in research and development efficiency, big data analysis, and other fields. Yuzhi AI expert, based on Qifull's large model, combines the professional capabilities of senior technical and big data analysts with data accumulation to form an "AI expert" in the financial field.
Microsoft Introduces EvoDiff Protein Generation Framework, Core Model Contains 640 Million Parameters
According to "Science and Technology Board Daily" on September 22nd, Microsoft recently launched a universal framework called EvoDiff, which can generate "high-fidelity" and "diverse" proteins based on protein sequences without the need for structural information of the target protein, thus eliminating the most cumbersome steps. This technology is expected to be applied in the creation of enzymes for new therapeutic drugs and drug delivery methods, as well as the development of new enzymes for industrial chemical reactions. The core of the EvoDiff framework is a model with 640 million parameters, trained using protein data from different species and functional categories.
Microsoft's Win11 Drawing Application Introduces Cocreator Image Generation Function
According to IT Home on September 22nd, Microsoft demonstrated the Cocreator function of the Win11 drawing application at the Surface 2023 event, which can generate images based on user input text. This function is similar to Bing Image Creator, but based on DALL-E. However, it is currently unclear whether it is based on DALL-E 2 or DALL-E 3. The report stated that Microsoft is currently inviting some users to test the Cocreator function, and users need to log in to their Microsoft account to join the waiting list.
Intel to Provide Supercomputer for Stability AI
According to "Science and Technology Board Daily," citing foreign media Tomshardware on September 22nd, Intel will build a supercomputer for Stability AI, one of the competitors of OpenAI, using Xeon processors and 4,000 Gaudi2 AI hardware accelerators to construct an AI supercomputer. Intel claims that this system will become one of the world's largest artificial intelligence supercomputers.
OpenKylin Operating System Officially Integrates Large Models, Providing Voice Assistants and Desktop Plugins
According to IT Home on September 21st, the OpenKylin operating system officially integrated large models, exploring the future development direction of intelligent interaction in desktop operating systems from the perspectives of intelligent input and output. The system's integration of large models includes two scenarios: voice assistants and desktop AI large model plugins. The OpenKylin voice assistant server uses the open-source model library paddlespeech for speech recognition services, recognizing and returning the streaming speech uploaded by the client. The client uses Qt's Multimedia module for speech recognition, then encapsulates the format and uploads it to the server, processing the corresponding operation based on the recognition result.
Google Announces AI-Assisted Features for YouTube
According to the news on September 22nd, Google announced the launch of five new AI-assisted features for YouTube creators: AI-generated images, video enhancement (adding filters), video idea and suggestion generation based on audience preferences, automatic dubbing, and music accompaniment.
Metaverse Recommended Reading:
"AI Water Army Tricks 'Andy Lau'"
"Why does Andy Lau rarely play the villain?" A recent hot topic on various social platforms has turned into a nationwide Turing test. Just because of this ordinary question, it attracted the siege of AI water army: "Lousy Huawei, using outdated chips to counterfeit Apple." Internet water armies have long been a mature gray industry chain, and now, with the support of generative AI, technology is once again doing evil?
https://www.8btc.com/article/6833626
"Microsoft's Late-Night Revolution, GPT-4 Office Suite Debuts! $30 per month, Snatching Jobs from 1 Billion Workers?"
Microsoft is really stirring things up! Suddenly exploding late at night, GPT-4 empowered the new Copilot to officially join the Office suite, Windows 11, and Edge. It can be said that what Microsoft boasted about six months ago has almost all been realized today.
https://www.8btc.com/article/6833614
"This List Reveals the 'Gap' in the Application of AI between China and Foreign Countries"
Recently, a16z released a list of the top 50 AI products by monthly traffic. After analysis and comparison, an interesting phenomenon emerged. Compared to AI companies in China, most of their overseas counterparts have already achieved profitability in the application layer of AI, especially in consumer-oriented applications. The vast majority of applications in China are still concentrated in the business-to-business (B2B) sector.
https://www.8btc.com/article/6833674
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。
