【Key Policies】The Fujian Provincial People's Government issued the "Implementation Plan on Promoting the High-quality Development of Digital Cultural Tourism" to deepen the application of artificial intelligence in cultural tourism scenarios
On April 24, the Fujian Provincial People's Government issued the "Implementation Plan on Promoting the High-quality Development of Digital Cultural Tourism" (hereinafter referred to as the "Implementation Plan").
The "Implementation Plan" proposes that by 2025, the cultural and tourism data resource network will achieve 100% coverage in the province's national tourist attractions above 4A level, cultural venues and art troupes at or above the municipal level, 100% coverage of digital public cultural services at or above the county level, and 100% coverage of the digital supervision of cultural relics protection units at or above the provincial level. 5G, metaverse, virtual reality, blockchain, artificial intelligence and other new technologies have been deepened in the application of cultural tourism scenarios, 100 smart scenic spots have been cultivated, 10 cultural industry demonstration parks (bases) with digital industry as the main business have been created, a number of digital museums, libraries, cultural centers, art galleries, etc. have been built, and a number of national smart tourism demonstration villages and towns have been created, and the development of digital cultural tourism in the province has reached the leading level in the country.
The "Implementation Plan" comprehensively uses big data, cloud computing, artificial intelligence and other technologies to gather, sort out, analyze, and display information data of the cultural and tourism industry and monitoring video of scenic spots, strengthen the monitoring and prediction and early warning of cultural and tourism economic trends, and provide decision-making reference for the development of cultural and tourism economy, emergency rescue and disposal, and market governance. Promote scientific research institutions, focus on breakthroughs in key common technologies such as big data in cultural tourism, the metaverse, and artificial intelligence, build application-oriented scientific and technological innovation carriers, and promote the coordinated development of the industrial chain innovation chain. Promote cultural and cultural venues, tourist attractions, resorts, leisure blocks and other places or related spaces, use digital technologies such as augmented reality, virtual reality, and artificial intelligence, and integrate elements such as cultural creativity to create immersive digital cultural tourism immersive consumption experience spaces such as night tours, exhibitions, and performing arts. Using new technologies such as 5G, artificial intelligence, and virtual reality, we can create experiential service scenarios such as slow live broadcasting, digital humans, AI tour guides, and stereoscopic projection.
Comments: Artificial intelligence, as the core driving force of a new round of scientific and technological revolution and industrial transformation, is committed to exploring and developing cutting-edge theories, innovative methods, cutting-edge technologies and their wide application systems that can simulate, extend and expand human intelligence. The development of artificial intelligence is of great significance for understanding the essence of intelligence, promoting scientific and technological innovation, and improving social productivity. In recent years, with the rapid advancement of technology, many breakthroughs have been made in the field of artificial intelligence. With the rapid development of technologies such as machine learning, deep learning, and natural language processing, AI systems have more powerful data processing, analysis, and learning capabilities. These technological breakthroughs provide a broad space and unlimited possibilities for the application of artificial intelligence in all walks of life.
At the same time, the progress of science and technology and the transformation of consumer demand have made the traditional cultural tourism industry face an urgent need for transformation and upgrading. Consumers' requirements for travel experience are becoming more personalized and diversified, and traditional service models can no longer meet their needs. The introduction of artificial intelligence has brought intelligent solutions to the cultural tourism industry. Through the use of artificial intelligence technology, it can accurately analyze the needs of tourists, provide personalized travel recommendations and services, and improve the satisfaction and experience quality of tourists.
The "Implementation Plan" conforms to this trend and aims to promote the digital transformation and intelligent upgrading of the cultural and tourism industry through the comprehensive application of artificial intelligence technology. The program emphasizes the comprehensive coverage and in-depth application of cultural and tourism data resources, and realizes the precise management and optimization of services for the cultural tourism industry through big data, cloud computing and other technical means. At the same time, the plan also clarifies the important role of artificial intelligence in cultural tourism safety supervision, emergency rescue, market governance, etc., providing a strong guarantee for the sustainable development of the cultural tourism industry.
The release of the "Implementation Plan" not only provides a clear direction and path for the application of artificial intelligence in the cultural tourism industry, but also injects new vitality and impetus into the future development of the cultural tourism industry.
Figure 1: Market size of Chinese artificial intelligence industry from 2017 to 2023
Source: Zhiyan Consulting
【Key Policies】The Guangdong Provincial People's Government announced the "Guangdong Province Pilot Work Plan for Carrying out National Standardization Innovation and Development" to accelerate the promotion of independent intellectual property rights standards for artificial intelligence
On April 24, the Guangdong Provincial People's Government announced the "Guangdong Province Pilot Work Plan for Carrying out National Standardization Innovation and Development" (hereinafter referred to as the "Work Plan").
The "Work Plan" proposes to develop a number of advanced standards in the fields of industrial upgrading, scientific and technological innovation, rural revitalization, green and low-carbon, smart cities, modern services, and social governance, and puts forward a series of development goals until 2025.
By 2025, a number of high-quality "Bay Area Standards" will be issued and implemented, more than 200 new international standards will be formulated, more than 1,500 national standards will be newly formulated, and the proportion of standard research results will be formed by common key technologies and applied science and technology projects will reach 55%, and the average period of local standard formulation will be shortened to less than 15 months. Cultivate and introduce a number of high-level standardization research institutions, apply for a number of national technical standard innovation bases and national standard verification points with high quality, build more than 200 new standardization pilot demonstration projects, cultivate a number of standardized social groups and standard innovative enterprises with international advanced level, and steadily increase the number of registered international standard experts. Introduce and train more than 100 standardization leading talents, build a team of more than 1,000 standardization experts, cultivate more than 10,000 business backbones who can meet the needs of the full implementation of the standardization strategy, and improve the standardization expert think tank in the Guangdong-Hong Kong-Macao Greater Bay Area. Support the Guangdong-Hong Kong-Macao Greater Bay Area, Shenzhen Pilot Demonstration Zone and Hengqin, Qianhai, Nansha, Hetao four major platforms to establish international standards organizations, enterprises to participate in the formulation and revision of international standards, foreign advanced standards The ability of enterprises to participate in the formulation and revision of international standards and foreign advanced standards has been significantly enhanced, the important role of standards in foreign trade is more obvious, and the economic, social, quality and ecological benefits of standardization are fully revealed.
The "Work Plan" focuses on digital transformation, carries out standardized research and layout based on new business formats such as artificial intelligence and the metaverse, and accelerates the formation of a number of standards that are comparable to the international advanced level, have obvious leading and exemplary roles, and have independent intellectual property rights.
【Key events】Top 10 enterprises of Chinese artificial intelligence invention patents: Tencent first, Huawei fifth
On April 22, IPRdaily, China's authoritative patent media, released a ranking of Chinese artificial intelligence invention patent enterprises, which brings together artificial intelligence invention patents applied for and published in China in the past ten years. What is striking is that Tencent, Baidu, Ping An, State Grid, Huawei, Ant Group, Alibaba, JD.com, OPPO, China Mobile and other companies stand out and rank among the top 10.
Specifically, Tencent ranks first with 15,626 patents, followed by Baidu with 13,723 patents. Ping An Group ranked third with 13,139 patents, while State Grid ranked fourth with 11,567 patents. Huawei ranked fifth with 5,966 patents, while Ant Group and Alibaba occupied sixth and seventh positions with 4,381 and 4,103 patents, respectively. JD.com and OPPO ranked eighth and ninth with 3,815 and 3,743 patents, respectively, while China Mobile rounded out the top 10 with 3,709 patents.
Previously, Tencent Chairman and CEO Pony Ma reviewed the development of Tencent's various businesses in the past year and looked forward to future strategic plans. In the 35-minute sharing, Ma Huateng emphasized the importance of AI, pointing out: "AI was undoubtedly the focus of the industry last year, and Tencent has also made significant progress in this field, although it cannot be said to be the most advanced, but at least it is not too far behind. ”
In addition, Ma Huateng pointed out that in the future, we need to closely integrate the AI hybrid model with our actual scenarios and deeply integrate it into various businesses. I think that in the short term, pure native AI applications may not emerge yet, and more by improving the efficiency of our existing products, which is undoubtedly a huge opportunity. ”
【Key Event】Wild animals are frequently hit and killed, and Brazil plans to use AI to save them
On April 23, a graduate computer science student at the University of São Paulo proposed the use of artificial intelligence (AI) to identify wild animals and alert drivers.
According to the Brazilian Centre for Road Ecology (CBEE), part of the Federal University of Lavras, Brazil, about 475 million vertebrates die on roads in Brazil every year, most of them smaller species such as capybaras, armadillos and possums. CBEE coordinator Bager said road safety in Brazil has become one of the biggest threats to local wildlife. The various initiatives currently in place, such as erecting signs warning passing through animals and building animal bridges and tunnels, have had little effect, but have not solved the root cause of the problem.
To do this, Ferrante, a computer science professional, created a database of thousands of animal pictures and trained an AI intelligent model to identify them in real time. The test results show that the AI model is effective. Ferrante said that in order for the project to become a reality, the support of road management companies is needed to warn drivers using traffic probes and related computing equipment.
【Key Event】The country's first "AI voice infringement case" was announced: 250,000 yuan in compensation
On April 23, the Beijing Internet Court made a first-instance judgment on the country's first AI-generated voice personality infringement case.
The main issues involved in this case are whether the voice rights of natural persons should be extended to AI-generated voices, and how to determine the recognizability of AI-generated voices.
The plaintiff, Yin, was a dubbing artist and found that his dubbing works were widely circulated in multiple apps. The voices in these works came from text-to-speech products in a platform operated by defendant 1, a Beijing-based intelligent technology company.
The plaintiff had recorded audio recordings for defendant 2 and a cultural media company in Beijing, and defendant 2 provided the audio recordings of these audio recordings to a software company for defendant 3 to allow it to use, copy, and modify the data for its products and services.
Defendant 3 only used the audio recordings recorded by the plaintiff as materials for AI processing, generated the text-to-speech products involved in the case, and sold them on a cloud service platform operated by a network technology company in Shanghai.
The court held that the voice of a natural person is unique, unique and stable, and can form or cause ordinary people to have thoughts or emotional activities related to the natural person, and can show the individual's behavior and identity to the outside world.
If the AI-synthesized voice can be associated with the general public or the public in related fields based on its timbre, intonation, and pronunciation style, it can be found to be recognizable.
In this case, the AI voice is highly consistent with the plaintiff's timbre, intonation, pronunciation style, etc., which can cause ordinary people to have thoughts or emotional activities related to the plaintiff, and can link the voice to the plaintiff himself, and then identify the plaintiff's main identity. Therefore, the plaintiff's voice rights and interests extended to the AI voice in question.
Defendant 2, a Beijing culture media company, and defendant 3, a software company, used the plaintiff's voice in AI without the plaintiff's permission, which constituted an infringement of the plaintiff's voice rights and interests, and should bear corresponding legal responsibility.
Defendant 1, an intelligent technology company in Beijing, defendant 4, a network technology company in Shanghai, and defendant 5, a technology development company in Beijing, are not subjectively at fault and are not liable for damages. Comprehensively consider factors such as the circumstances of the defendant's infringement, the value of products in the same market, and the number of broadcasts of the products, and make a discretionary determination of damages.
In the end, defendant 1 and a Beijing intelligent technology company and defendant 3 a software company apologized to the plaintiff, and defendant 2 and a Beijing cultural media company and defendant 3 a software company compensated the plaintiff for losses totaling 250,000 yuan.
This case is the first time that the Civil Code of the Mainland has written the protection of "voice" into the Civil Code in the form of legislation, and clearly refers to the application of portrait rights to protect the voice of natural persons, reflecting the legislative spirit of comprehensive respect for and protection of personality rights and interests.
The voice of any natural person shall be protected by law, and the authorization of sound recordings does not mean the authorization of sound AI, and the unauthorized use or licensing of others to use the sounds in sound recordings without the permission of the right holder constitutes infringement
【Key Events】The Supreme People's Procuratorate: It will focus on strengthening the judicial protection of intellectual property rights in future industries such as artificial intelligence
On April 25, the Supreme People's Procuratorate held a press conference on "Deepening the Comprehensive Performance of Intellectual Property Procuratorial Duties and Promoting the Development of New Quality Productivity". Gong Ming, a member of the party group and deputy procurator general of the Supreme People's Procuratorate, said at the meeting that in the next step, the procuratorial organs will work hard to serve the healthy development of the future industry. The future industry is the main form of scientific and technological revolution and industrial transformation, and is the concentrated representative of the new quality of productive forces. The procuratorial organs will focus on strengthening the judicial protection of intellectual property rights in future industries such as artificial intelligence, biotechnology, green energy, and information and communication, not only to "strictly deal with them in accordance with the law" so that infringers dare not commit them again, but also to "act quickly in accordance with the law" to avoid continuous damage to the legitimate rights and interests of rights holders. Increase the extent of judicial protection for scientific research personnel involved in future industries, prudently handle cases involving scientific research funds and high-tech talents, create a good atmosphere that encourages innovation and tolerates failure, and stimulates the innovation vitality of the whole society.
OpenAI has hired its first employee in India to focus on the Asian market
On April 22, OpenAI is developing a strategy to increase its presence in different countries. OpenAI has hired its first employee in India, Pragya Misra, 39, who will be responsible for the company's public policy affairs and partnerships in India.
Misra is an industry insider with a deep understanding of India's technology sector, having previously served as Truecaller's Director of Public Affairs, working closely with government departments, investors, key stakeholders and media partners, as well as three years at Meta and WhatsApp's first employee in India.
OpenAI's move is to expand the company's presence in Asia and advocate for AI-friendly regulations. With its large population and growing economy, India is a major potential market for AI companies.
OpenAI faces competition in India from companies like Google, which is developing an AI model that can handle more than 100 Indian languages.
Meanwhile, OpenAI opened its first Asian office in Tokyo, Japan, this month and started its business activities, which means that OpenAI wants to aggressively expand the Asian market.
Tesla plans to sell Optimus Prime robots by the end of next year, aiming to solve the labor shortage
On April 23, Tesla CEO Elon Musk revealed to investors in a conference call that Tesla's humanoid robot "Optimus Prime" is still in the laboratory development stage, but the company plans to launch it to the market as soon as the end of next year.
With the growing demand for automation and intelligence in various industries, humanoid robots are seen as a possible solution to the labor shortage. These robots can perform dangerous or repetitive tasks in industries such as logistics, warehousing, retail, and manufacturing, increasing productivity and reducing labor costs.
Musk said he expects the Optimus Prime robot to begin performing tasks inside Tesla's factories by the end of the year, marking an important milestone for the robot's technology. Tesla's humanoid robot project has been steadily advancing since the launch of the first-generation Optimus robot in September 2022. Earlier this year, Tesla also showed a video of a second-generation bipedal robot folding a T-shirt inside a factory, showing the practicality and flexibility of the technology.
Commenting on Tesla's position in the field of humanoid robots, Musk confidently stated: "I think Tesla is in the best position among all humanoid robot manufacturers to be able to achieve mass production by performing effective reasoning on the robots themselves." "This reflects Tesla's ambitions in robotics.
However, Musk also has a history of failing to deliver on promises, such as in 2019 when he predicted that Tesla would operate a network of self-driving cars by 2020, but this plan has not yet been fully realized. Therefore, the market remains cautiously optimistic about the timing of the launch of the "Optimus Prime" robot.
At the same time, other automakers and technology companies are also actively deploying in the field of humanoid robots. For example, Boston Dynamics of Honda and Hyundai Motor has been conducting research and development in this area for many years. In addition, Figure, a startup backed by Microsoft and NVIDIA, has also partnered with BMW to deploy humanoid robots in factories in the United States.
Despite the fierce competition, Musk is confident in Tesla's future in the field of humanoid robots. He believes that robot sales have the potential to become an important part of Tesla's business compared to other areas such as automobile manufacturing.
Microsoft released Phi-3 Mini: 3.8B parameters are small enough to be stuffed into a mobile phone, and the performance is comparable to GPT-3.5
On April 23, Microsoft open-sourced the small language model (SLM) - Phi-3-mini on its official website. This is the first model launched in the Microsoft Phi-3 family of models.
As the fourth generation of Microsoft's Phi series, Phi-3-mini surpasses many models with tens of billions of parameters in language, reasoning, coding, and math benchmarks with a huge training data volume of 3.8 billion parameters and 3.3T tokens.
In addition, Phi-3-mini has two context-length variants, 4k and 128k tokens, which are pre-trained and instructionally tuned to better understand human language, expressions, logic, and execute different types of instructions.
Compared with large language models, Phi-3-mini models are easier to operate. The small size of the model allows it to be deployed and run on the device side. The model is said to be accessible offline as a chatbot and has performance that is not inferior to GPT-3.5.
The cost of the Phi-3-mini was significantly reduced. Sébastien Bubeck, Microsoft's vice president of generative AI research, said the Phi-3-mini could cost a tenth of what other models with similar features do.
According to the Microsoft research team, the innovation of Phi-3-mini's small size and high performance is due to a dataset composed of synthetic data. The dataset consists of a large amount of filtered web data and synthetic data from other AI-made children's books.
微软计划在未来几周内推出Phi-3系列的另外两个模型，分别是Phi-3-small（7B）和Phi-3-medium（14B）。
Amazon ended its drone delivery project in California and plans to enter Arizona by the end of the year
On April 23, CNBC reported that Amazon is shutting down its drone delivery operations in Rockford, California, one of the decade-long project's earliest testing sites in the United States.
The project, called Prime Air, has been hard to get off the ground since Amazon founder Jeff Bezos first elaborated his vision in 2013 for self-driving drones to deliver packages weighing less than 5 pounds in 30 minutes or less.
Amazon said in a blog post on the 22nd that it plans to continue to expand its drone delivery service to more U.S. cities in 2025, and plans to open it in parts of the Phoenix area later this year.
The company said it is working with the Federal Aviation Administration and local officials to obtain permission to make drone deliveries in Tolson, Arizona, west of Phoenix.
"We also decided to close the delivery site in Rockford as we look to the future and prioritize our resources to continue to grow the program," the company wrote.
Amazon said it will provide opportunities for employees at other locations, and Rockford residents can still place orders using alternative delivery methods.
Amazon says it is currently conducting test flights to prove the reliability of its new delivery drone, the MK30, which it unveiled at an event last year.
The drone is designed to be smaller, quieter than previous models, and can fly in light rain.
Amazon received Part 135 certification from the FAA in 2020, allowing it to use drones to deliver packages, with some restrictions. But the progress of Prime Air has been delayed.
Amazon said it would begin testing delivery services in college town, Texas, about 100 miles northwest of Houston, and Rockford, a small town south of Sacramento, in 2022, and the arrival of the plan was initially met with some skepticism from local residents.
Just when Prime Air looked like it was about to start operations, it suffered layoffs last year as part of Amazon's broader layoff plan. The group has also suffered regulatory setbacks and executive departures.
Amazon is continuing to push forward with the expansion of drone delivery business.
Last October, the FAA cleared a major regulatory hurdle by easing some restrictions on where its drones could operate, allowing the company to fly over roads and cars if necessary to complete routes.
Last week, Amazon signed an agreement with Embention, a developer of autonomous drone systems and components. As part of the agreement, Embention will provide Amazon with security-related hardware and software.
【Key Companies】Former Meta executives join Microsoft's AI supercomputing team: building next-generation systems
On April 23, according to foreign reports, Jason Taylor, a former Meta executive, is about to join Microsoft's AI supercomputing team.
Kevin Scott, Microsoft's chief technology officer, said in a LinkedIn post on Monday that Taylor will serve as the company's vice president and deputy chief technology officer, helping to build next-generation systems and push the frontiers of artificial intelligence.
Taylor worked at Meta from 2009 to 2022, most recently as the company's vice president of infrastructure.
According to his LinkedIn profile, he is responsible for artificial intelligence, data and privacy infrastructure, as well as managing the company's server budget. From 2015 to 2017, Taylor also served as president of the Open Compute Project Foundation, an organization dedicated to promoting open source design in data centers
【Key Companies】SK hynix plans to establish an artificial intelligence (AI) product research center in the U.S
On April 23, SK hynix plans to spend USD 3.9 billion to build an advanced packaging factory and artificial intelligence (AI) product research center in Indiana, USA.
【Key Enterprise】SenseTime threw out a large-scale model luxury family bucket, fully benchmarking GPT-4 Turbo
On April 23, SenseTime upgraded the "Ririxin SenseNova 5.0" large model system, and its comprehensive capabilities are fully benchmarked against GPT-4 Turbo.
For the first time in the industry, SenseTime launched a full-stack large-scale model product matrix of "cloud, edge, and device", including SenseTime's device-side large model, device-cloud collaboration solution, and the edge product "SenseTime Enterprise Large Model All-in-One Machine" for finance, code, healthcare, government affairs and other fields.
"RiRixin SenseNova 5.0" adopts a hybrid expert architecture, based on more than 10TB tokens training, covering a large amount of synthetic data, and the context window can support 200K during inference, which mainly enhances the knowledge, mathematics, reasoning, and code capabilities, and meets or surpasses the performance of GPT-4 Turbo in mainstream objective evaluation.
Xu Li, chairman and CEO of SenseTime, said that it is not how strong the model ability is, but that in different applicable scenarios, the decision-making speed of the small model is fast, when the large model is still calculating, the small model has completed the judgment and punched, regardless of whether its punch is optimal or not, it has really hit the opponent.
Its device-side large model achieves the fastest inference speed in the industry, with the flagship platform reaching up to 109.5 words per second, while the fastest reading speed of the human eye is only 20 words per second. SenseTime also officially released the end-side business SDK, which can be adapted to a variety of Qualcomm Snapdragon and MediaTek Dimensity chips.
The enterprise-level large model all-in-one machine supports 100 billion model acceleration and knowledge retrieval hardware acceleration, and can save 80% of the inference cost of 100 billion large model compared with similar products in the industry. The price of the light version of the small raccoon code large model all-in-one machine starts at 350,000 yuan per unit, and a single unit supports a team of 100 people, and the daily use cost is as low as 4.5 yuan per person.
At the same time, SenseTime released a Ascend-based industry model to work with Huawei to build an industry ecosystem for finance, healthcare, government affairs, and code.
In terms of its own application, SenseTime's "Ririxin SenseNova 5.0" has been updated in Miaohua, Ronin, Gewu, Qiongyu, Dayi, Little Raccoon Family and other products.
There is also One More Thing, a Wensheng video generation platform, where Xu Li brings three videos that are completely generated by large models, and emphasizes the controllability of characters, actions, and scenes.
In the future, a video can be generated by entering a text description, and the characters' clothing, hairstyles, and scenes can be pre-set to maintain the consistency and consistency of the video content.
Microsoft and Coca-Cola have reached a five-year partnership agreement on cloud and generative AI initiatives
On April 23, Microsoft and Coca-Cola issued a statement saying that they will establish a five-year strategic partnership to align Coca-Cola's core technology strategy system-wide, promote the adoption of cutting-edge technologies, and drive innovation and productivity on a global scale. As part of the partnership, Coca-Cola has committed $1.1 billion to the Microsoft Cloud and its generative AI capabilities. Through the partnership, the two companies will work together to try new technologies such as Azure OpenAI services to develop innovative generative AI use cases across various business functions, including testing how Microsoft 365 Copilot can help improve productivity.
【Key Enterprises】UK regulators: Open antitrust review of Microsoft and Amazon's AI investment activities
On April 24, the UK's Competition and Markets Authority (CMA) said it was seeking third-party opinions on the AI partnership between Microsoft and Mistral AI, as well as the partnership between Amazon and Anthropic.
At the end of February, Mistral AI, a European AI unicorn, announced that it had reached a cooperation with Microsoft, becoming the second company after OpenAI to provide large model services on Microsoft's cloud service Azure. At the end of March, Amazon made an additional investment in OpenAI's arch-rival Anthropic, bringing its investment in the latter to $4 billion.
The CMA also said it would like to hear from third parties on Microsoft's hiring of former Inflection AI employees and related arrangements.
In March, Microsoft appointed DeepMind co-founder Mustafa Suleyman as head of Microsoft's newly formed consumer AI division and hired several employees from his AI startup, Inflection AI.
The CMA believes that no conclusion has been reached as to whether these transactions are in compliance with the UK M&A rules or whether they will raise competition concerns.
【Key Enterprise】IPO of the world's first profitable large-scale model company
On April 24, Mobvoi (2438.HK), an "AI unicorn" that has undergone two submissions, was successfully listed on the Hong Kong Stock Exchange. Mobvoi's IPO was priced at HK$3.8 per share, and as of press time, Mobvoi was trading at HK$3.1 per share, with a total market capitalization of HK$4.624 billion.
As the world's first profitable large-scale model company, the Hong Kong stock IPO of Askou also means that it has become the "first AIGC stock" in 2024.
Mobvoi is a large-scale model company with AIGC and voice interaction technology as the core, mainly with three major business segments: AIGC solutions, AI enterprise solutions, smart devices and accessories, and has launched a multi-modal large-scale model "Sequence Monkey", which is applied to its Moyin Workshop, Wonder Yuan and other products.
According to the CIC report, in terms of revenue in 2022, the market share of AI speech technology and natural language processing (NLP) software in China is 1.4%, ranking third.
In 2022, the company's AIGC solutions accounted for approximately 0.9% of the global AIGC market share in terms of revenue, ranking 8th in the global market and 1st in Asia in terms of revenue from AIGC products and services in 2022.
Profluent, a start-up company, announced that it has open-sourced the world's first AI design gene editor and successfully edited DNA in human cells
On April 24, the startup Profluent announced that it had open-sourced the world's first AI design gene editor, successfully editing DNA in human cells.
On the most extensive dataset of CRISPR-based gene editing systems to date, the researchers trained LLMs. These LLMs produce proteins that expand the diversity of almost all naturally occurring CRISPR-Cas families by a factor of 4.8!
In addition, gene editors have shown comparable or better activity and specificity than SpCas9 (an example gene editor) in human cells, while remaining more than 400 mutations away.
This means that we have our own genomic code. The scientists of tomorrow will be able to fight diseases more precisely and quickly than they do today.
Moreover, the company has also decided to release these DNA molecules freely under the OpenCRISPR protocol.
【Key Enterprises】Huawei Developer Conference officially announced that Pangu Model 5.0 will be released
On April 25, the Huawei Developer Conference will be held in Songshan Lake, Dongguan from June 21 to 23. It is reported that at this press conference, Pangu Model 5.0 and HarmonyOS NEXT Hongmeng Galaxy Edition will be unveiled on the same stage for the first time.
The Pangu model is a series of pre-trained models launched by Huawei, including natural language processing (NLP) models, computer vision (CV) models, and scientific computing models. Trained on large-scale data, these models have high language understanding and generation capabilities and can be applied to a variety of natural language processing tasks, such as text generation, question answering systems, machine translation, and more.
【Key Financing】Aishi Technology completed the A2 round of financing of over 100 million yuan led by Ant Group
On April 23, Aishi Technology completed the A2 round of financing of over 100 million yuan, led by Ant Group, with Light Source Capital as the exclusive financial advisor. So far, Aishi Technology has received more than 200 million yuan in financing within a year, becoming the largest start-up company in the field of video models in China.
Aishi is a developer of AIGC visual multimodal algorithms, committed to exploring the unlimited potential of AI visual generation, empowering different industries and scenarios through AIGC technology, opening up new prospects for visual creativity and intelligent generation, allowing users to expand their imagination and unleash their potential. This round of financing will be used to further iterate the self-developed video generation model, upgrade the team, and accelerate the industrial application of AI video generation technology.
【Key Financing】Airchat received strategic investment
On April 24, Airchat is an AI social media application that allows users to create chat rooms that communicate asynchronously with voice, text or video, with built-in AI voice translation, which can translate the user's language and imitate the user's accent for expression, reducing the cost of communication between users in different countries. Airchat receives a strategic investment, with OpenAI CEO Sam Altman investing.
【Key Investment】To promote the "cost reduction and efficiency increase" of AI chips, Nvidia acquired two start-ups
On April 24, Nvidia announced that it had signed a definitive acquisition agreement with Run:ai. The company believes this will help customers make more efficient use of their AI computing resources.
Founded in 2018, Run:ai is a workload management and orchestration software provider based on the open-source container orchestration platform Kubernetes. Today, AI deployments are becoming increasingly complex, with workloads distributed across cloud, edge, and on-premises data center infrastructure. Managing and orchestrating generative AI, recommender systems, search engines, and other workloads requires complex scheduling to optimize performance at the system level and underlying infrastructure.
Run:AI's products can achieve efficient resource utilization of GPU clusters. Put simply, its products enable developers to run multiple AI workloads in parallel, thereby improving the efficiency of AI chips and reducing costs.
目前，Run:ai的解决方案已与英伟达的DGX、DGX SuperPOD、Base Command、NGC容器和AI Enterprise软件等产品结合，英伟达DGX和DGX Cloud的客户将能够使用Run:ai的AI工作负载功能。
Nvidia officials did not disclose the details of the terms of the acquisition transaction, such as the amount of the acquisition. However, according to Israeli media Calcalist and other relevant media reported a few days ago, the transaction amount is 700 million US dollars.
Deci, another AI startup that has also received the favor of NVIDIA, is also committed to achieving "cost reduction and efficiency increase" of AI chips. According to The Information, Nvidia agreed to acquire Israeli startup Deci, and it was not possible to know the exact amount of the acquisition.
Founded in 2019, the company provides efficient generative AI and computer vision models that can be deployed in a variety of environments while keeping data secure and controlling inference costs, according to Deci's official website.
It is understood that its solution allows it to run cheaper on AI chips by resizing the AI model. At present, Deci has reached cooperation with many technology giants such as Microsoft, Intel, AMD, and Amazon.
It is not difficult to see that the purpose of NVIDIA's two investments is to help its customers make more effective use of their own AI chip products.
According to a report by The Information, two people with knowledge of the deal said Run:ai's ability to improve the efficiency of AI chips could attract more customers to NVIDIA DGX Cloud. In addition, Nvidia may offer Deci as well as CUDA software, allowing developers to build AI-powered applications more cheaply.
【Key Financing】Parloa received €61.7 million in Series B financing
On April 25, Parloa, a developer of a conversational AI platform that offers natural-sounding AI voices, uses conversational AI to automate customer service, such as phone bots or chatbots. Parloa announced that it has raised €61.7 million in Series B funding, led by Altimeter Capital, with participation from new and existing investors EQT Ventures, Newion, Senovo, Mosaic Ventures and La Famiglia Growth.
【Key Financing】Cognition raised $175 million in Series A financing
On April 25, venture capital firm Founders Fund led a $175 million investment in Cognition, a startup that recently launched an artificial intelligence programming assistant called Devin, according to an investor who had seen the terms of the deal. The investor said the Founders Fund values the startup at $2 billion, which includes the new funding. Just a month ago, Founders Fund led the startup's Series A funding round at a valuation of $350 million, according to PitchBook.
Zhiyan Consulting has been focusing on industrial consulting for 15 years and is a professional service organization in the field of industrial consulting in China. The company takes "driving industrial development with information and empowering enterprise investment decision-making" as its brand concept. Provide professional industrial consulting services for enterprises, including high-quality industry research reports, special customization, monthly topics, feasibility study reports, business plans, industrial planning, etc. Provide weekly/monthly/quarterly/annual reports and other regular reports and customized data, covering policy monitoring, corporate dynamics, industry data, product price changes, investment and financing overview, market opportunities and risk analysis, etc.