One week of AIGC丨Large model financing is crazy, geniuses, famous schools, and big factories are vying to end

Entering the second half of the year, it is obvious that there are more financing news for the large-scale model track. Compared with popular tracks such as mobile Internet, blockchain, and metaverse that appeared in the past, entrepreneurs of large models have higher "technical content". The founders are either geniuses in AI research, or have been in charge of AI in large companies. Business executives. Is the large model a ticket for entrepreneurs to become famous, or a new battlefield for big factories to harvest?

Image source: Generated by Unbounded AI tool

Financing

AI start-up large-scale model company "Dark Side of the Moon" completed the first round of market financing, and the market valuation was revealed to be about 300 million US dollars

According to a report on June 12 by the Science and Technology Innovation Board Daily, the dark side of the moon recently completed its first round of market financing, and received bets from first-line VCs such as Sequoia and Zhenge. The market valuation was exposed to about 300 million US dollars. The Dark Side of the Moon is an AI start-up large-scale model company, whose main business is AGI, and is committed to maximizing the value of AI.

Ant and Baidu jointly invested in the AIGC track, and after completing the first round of financing of nearly 100 million yuan, Shengshu Technology was valued at 100 million US dollars

According to Titanium Media, on June 19, Beijing Shengshu Technology Co., Ltd. (hereinafter referred to as "Shengshu Technology") officially announced the completion of an angel round of financing of nearly 100 million yuan. This round was led by Ant Group and Baidu initiated BV Baidu. Venture capital and Zhuoyuan Capital followed suit, and the current valuation has reached 100 million US dollars.

This round of funds will be mainly used for the construction of the core R&D team to accelerate the development of multi-modal large models and application products. This is the first AIGC project invested by Ant Group since the release of ChatGPT in November last year. It is also the third AI content generation-related project invested by Baidu Ventures. The first two are Xihu Xinchen and Morph Studio .

Blockchain-based AI computing protocol Gensyn completed $43 million in Series A financing, led by a16z

According to CoinDesk, Gensyn, a blockchain-based AI computing protocol, completed a US$43 million Series A round of financing, led by a16z, with participation from CoinFund, Canonical Crypto, Protocol Labs, Eden Block, Jsquare and other angel investors. Funding from the round will be used to accelerate the protocol’s rollout and expand its workforce, including adding protocol and machine learning engineers.

The UK invested 100 million pounds to establish an AI working group and OpenAI to jointly build basic models

According to a report from IT House on June 13, the British Prime Minister announced the establishment of an AI working group at the recently held 2023 London Technology Week, and plans to invest 100 million pounds in artificial intelligence regulations and the development of responsible artificial intelligence. To ensure the healthy development of artificial intelligence in a safe environment.

AI protein design company "Yuxinshuo Bio" completed nearly 10 million yuan in angel round financing

According to the "Kechuangban Daily" report on June 13, the AI protein design company "Yuxinshuo Bio" has completed an angel round of financing of nearly 10 million yuan. The investor is Jiuhe Venture Capital. This round of funds will be used for technology research and development and market expansion. Founded in 2022, Yuxinshuo Biotech is a company that focuses on new protein design. It briefly summarizes its technology and applications, that is, using AI to design biological components, and using biological components to construct genetic circuits. By designing genetic circuits, it makes Strains and catalytic enzymes can undergo natural evolution to eventually generate new functional proteins.

French artificial intelligence startup Mistral AI raises 105 million euros

French artificial intelligence startup Mistral AI has raised 105 million euros ($113 million) in its first round of funding, making it a "new leader" in the field of artificial intelligence, the French news agency reported on June 14. global players". It is reported that the company is a new company formed by former DeepMind and Meta researchers, and will try to compete with OpenAI and Google to develop generative artificial intelligence infrastructure. Mistral’s funding is one of the largest seed rounds ever raised by a European generative artificial intelligence company.

TSMC invests in Silicon Valley AI chip startup SiMa.ai

According to news from IT House on June 14, SiMa.ai, an artificial intelligence chip start-up company in Silicon Valley, announced on Tuesday that it has raised an additional US$13 million from investors including TSMC VentureTech Alliance. SiMa.ai has raised a total of 2 One hundred million U.S. dollars.

British start-up Synthesia's latest financing is about 90 million US dollars, promoted to AIGC unicorn

According to the Financial Associated Press, on June 13, local time, Synthesia, a digital media platform headquartered in London, UK, announced that it had obtained approximately US$90 million in the latest round of financing, with a valuation of US$1 billion, and was officially promoted to "Single Horn". beast" enterprise.

Entrepreneurship

Mei Tao, former vice president of JD.com founded HiDream.ai, aiming to surpass Stable Diffusion in the basic model by the end of the year

According to the news from the Heart of the Machine, Jingdong AI expert Mei Tao resigned from the position of vice president of Jingdong Group before the Spring Festival this year, and is determined to devote himself to the large-scale model business. One month later, HiDream.ai was established, and the company's direction was determined as the basic model and application of generative visual multimodality. It is reported that HiDream.ai's model can not only support Wensheng map, but also support Wensheng video, picture video and Wensheng 3D content. Among them, Wenshengtu products support input to generate pictures, and can further edit the parts that you want to modify in any picture. For example, it is required to modify the background, subject, etc. of the generated picture. The initial goal of Mei Tao's team is to surpass the latest version of Stable Diffusion in terms of basic models before the end of the year, and to surpass MidJourney in terms of products.

Li Ming, former employee of Xiaomi No. 9, built the world's first AI Android desktop robot

According to a report by IT House on June 13, Li Ming, the original founding employee of Mi 9, announced his start-up in April this year and entered the field of AI. He later established Optimist as CEO. From June 12th to 13th, Li Ming successively announced more information on his first product "Optimist Desktop Robot" on Weibo. It is reported that this small robot is positioned as "an AI + robot for geeks and enthusiasts". In addition, Li Ming forwarded a netizen's response saying "618 should save some money to see this", implying that the robot will be released around June 18.

Meta has lost a third of researchers who published AI research results in the past year

According to the Financial Associated Press, citing the Wall Street Journal, analysis data shows that Meta has lost at least one-third of researchers who published AI research results last year. Many people expressed fatigue or lack of confidence in Meta and believed that Meta could not keep up with competitors. pace. A Meta spokesperson said the company has continued to recruit and bring in new AI talent.

big model

Baichuan Intelligent officially launched the first 7 billion parameters Chinese and English pre-training large model baichuan-7B

According to Xinzhiyuan, on June 15, "Baichuan Intelligence" formed by Wang Xiaochuan officially launched the first 7 billion parameter Chinese and English pre-trained large model baichuan-7B. It is reported that baichuan-7B not only surpasses other large models such as ChatGLM-6B by a significant advantage in the C-, AGI and Gaokao Chinese authoritative evaluation lists, but also significantly leads LLaMA-7B in the MMLU English authoritative evaluation list. At present, the baichuan-7B large model has been released on Hugging Face, GitHub and Model Scope platforms.

Chinese Academy of Sciences: Independently developed a new generation of artificial intelligence large-scale model Zidong Taichu 2.0 came out

According to CCTV news reports, on June 16, the Institute of Automation of the Chinese Academy of Sciences released a new generation of artificial intelligence model developed by the institute in Shanghai - Zidong Taichu 2.0. Compared with the first generation, the brand-new large model focuses on improving the decision-making and judgment capabilities, realizing the leap from perception, cognition to decision-making, and will play a greater role in the fields of medical care, transportation, and industrial production in the future.

Tiangai 100 completed tens of billions of parameter large model training

According to 36 Krypton reports, on June 10, at the AI system sub-forum of the 5th Zhiyuan Conference, Shanghai Tianshu Zhixin Semiconductor Co., Ltd. announced that the computing power cluster of the Tiangai 100 accelerator card is based on Beijing Zhiyuan artificial intelligence The Aquila language basic model with 7 billion parameters of the research institute continued to train using code data, and it ran stably for 19 days. The model convergence effect met expectations, which proves that Tianshu Zhixin has the ability to support large model training with tens of billions of parameters.

Huawei Shengsi AI framework has open sourced more than 400 models

According to the "Kechuangban Daily" report on June 13, the Artificial Intelligence Framework Ecological Summit 2023 will be held on June 16. At the meeting, the artificial intelligence framework ecological initiative will be jointly established, and the Shanghai Shengsi AI Framework & Large Model Innovation Center will be launched. The first batch of partners settled in, the Shengsi Community Council was established, and the world's first full-mode large-scale model was released. The reporter learned that at present, Shengsi has open sourced more than 400 models, and its partners have released large models such as Zidong, Taichu, Pengcheng, and Shennong. Over 4.74 million users have downloaded and installed them, served more than 5,500 enterprises, and taught more than 200 colleges and universities. , in cooperation with more than 40 scientific research institutions.

Meta releases I-JEPA, a "human-like" AI model that analyzes and completes unfinished images more accurately than existing models

According to news from IT House on June 14, social media giant Meta (formerly Facebook) announced on Tuesday that it will open up to researchers some components of a new "human-like" artificial intelligence model that can be better than existing models. Accurately analyze and complete unfinished images. The model, called I-JEPA, uses background knowledge of the world to fill in missing parts of images, rather than just inferring from nearby pixels like other generative AI models. The approach, which employs a style of human-like reasoning championed by Meta's chief AI scientist Yann LeCun, helps avoid common mistakes in AI-generated images, such as having an extra finger.

Microsoft launches LLaVA-Med AI model, which can analyze medical pathology cases

According to IT House's report on June 14, Microsoft researchers recently demonstrated the LLaVA-Med model, which is mainly used in biomedical research and can infer the patient's pathological condition based on CT and X-ray pictures.

360 officially releases new AI products "360 Smart Brain Model" and "360 AI Digital Human Plaza"

According to 36 Krypton, 360 officially released a new AI product "360 Smart Brain Model". It is understood that the 360 big language model is 360's self-developed 100 billion parameter large model, which already has cross-modal generation capabilities, and has four major capabilities of text, image, voice, and video processing. scenes such as video. In addition, 360 Zhinao 4.0 has been officially connected to 360 full-end products, including search, browser, security guard, etc. Zhou Hongyi, the founder of 360, said that 360 has passed the evaluation of the Ministry of Industry and Information Technology, and has become the first trusted AIGC large model in China.

application

Sichuan Changhong: AI large-scale model application "Changhong Super Brain" starts public beta

According to a report by "Kechuangban Daily" on June 15, some products of Changhong TV have been equipped with "Changhong Super Brain" artificial intelligence system, and the public beta has officially started for users. According to Sichuan Changhong, "Changhong Super Brain" is an artificial intelligence super brain system built on the basis of multi-modal interaction, semantic understanding and intent recognition, big data and other technologies. It interacts with users in the form of a digital image "Changhong Xiaobai". Execute instructions and realize user intent. As a result, "Changhong Super Brain" can perceive, understand, and learn, and can be applied to life encyclopedia knowledge, learning and education assistance, document creation and generation, and even picture design and other professional fields and personalized services.

ChatGPT officially entered the vehicle system: Mercedes vehicles will start testing AI voice assistant

According to the Financial Associated Press, on June 15 local time, Mercedes-Benz and Microsoft jointly announced that the two companies are cooperating to expand the use of artificial intelligence and plan to integrate this technology into the vehicle's voice control system. Mercedes said that this cooperation will bring the chatbot ChatGPT experience to car owners through Microsoft's Azure OpenAI service. The test program will begin on Friday local time, and 900,000 vehicles equipped with the MBUX infotainment system in the United States can participate in the test.

JD and Baidu create the first AI offline advertisement for e-commerce, which can save about 80% in production costs

According to Baidu AI official Weibo news, during the 618 period this year, JD.com and Baidu Wenxin Yige applied AIGC to e-commerce marketing, making the first large-scale offline advertising attempt, creating the first AI offline advertising in the e-commerce industry. According to Baidu's calculations, usually such a set of posters, from models, clothing, to design and typesetting, may cost close to 10,000 yuan per sheet, but with AI, the production cycle can be shortened by 70%, and the production cost can be saved by about 80%.

Meta releases voice generation AI "Voicebox", which supports voice generation for characters in the Metaverse

Babbitt News, on June 16, Meta announced the release of the speech generation AI model "Voicebox", which supports speech generation from text, can match audio styles based on samples that are only two seconds long, and convert text samples to another language, Given individual speech samples, it reads the translated text in the speaker's original voice, currently in six languages: English, French, German, Spanish, Polish, and Portuguese.

Supervision

UN Secretary-General backs creation of global artificial intelligence (AI) regulator

According to Reuters, United Nations Secretary-General Antonio Guterres (António Guterres) said yesterday that some AI executives have proposed the establishment of a global AI regulator similar to the International Atomic Energy Agency (IAEA), and he supported it.

“The latest form of artificial intelligence, generative AI, is sounding deafening alarm bells, the loudest of which are the developers of generative AI, and we must take these warnings seriously,” Guterres said.

ASEAN is drafting an AI regulatory bill: Singapore takes the lead and will be completed by the end of this year at the earliest

According to a report by IT House on June 16, five sources recently told Reuters that Southeast Asian countries are drafting governance norms and ethical guidelines for artificial intelligence (AI) to set up "guardrails" for this booming technology.

Ministers from the 10-member Association of Southeast Asian Nations (ASEAN) agreed in February to develop an ASEAN version of AI guidelines for the region's 668 million people, but previous reports did not detail discussions among regional policymakers. Some officials from Southeast Asian countries said that ASEAN is drafting AI guidelines and will try to balance the economic benefits of the technology with its many risks: "The drafting work is in progress and may be completed by the end of the year before it will be accepted for discussion and adoption by ASEAN member states. Another official said the draft could be formally announced at the ASEAN Digital Ministerial Meeting early next year when it is completed.

policy

Shanghai Science and Technology Commission issued the action plan for the key technology research of Metaverse

Babbitt News, on June 13, the Shanghai Science and Technology Commission issued the "Shanghai "Metaverse" Key Technology Research Action Plan (2023-2025)". The main direction of the program includes immersive technology and Web3 technology. Aiming at the immersive, open, sustainable real-time, human-centered and other cross-border composite technical characteristics of the "Metaverse", it focuses on the content, storage, calculation, transmission and Terminal and other technical aspects, combined with domestic and foreign industrial development and the city’s research and development foundation, focusing on the two main directions of immersive technology and Web3 technology, and building a new heights.

The plan proposes that, facing the demand for the production of massive immersive audio-visual content in the "Metaverse", based on the technical route of AI empowering virtual and real fusion, the focus will be on the layout of artificial intelligence generated content (AIGC), cross-scale acquisition and reconstruction, digital human generation and driving, and three-dimensional virtual space. Engine and other key technology research and development, realize the automatic and low-cost production of immersive audio-visual content in the super-realistic "Metaverse", and support the ecological construction of "Metaverse" content creation.

big company

Accenture plans to invest $3 billion in data and AI businesses over three years

According to the Wall Street Journal, Accenture said it plans to invest $3 billion over three years to focus on data and artificial intelligence (AI) businesses. The Dublin, Ireland-based professional services firm said its data and AI business will double its AI talent to 80,000 through a combination of hiring, acquisitions and training. The company also said it is launching a new platform called AI Navigator for Enterprise that will help customers figure out how to use AI in their businesses.

Lenovo will invest an additional US$1 billion to accelerate the development of artificial intelligence

According to news from Jinshi on June 14, Lenovo will invest an additional US$1 billion to accelerate the development of artificial intelligence (AI). Annual AI infrastructure revenue has exceeded $2 billion.

AMD announced the new MI300X AI chip, the number of GPUs required for AI model training will be less

According to a report by the Financial Associated Press on June 14, AMD announced the new MI300X AI chip; the new chip means that the generative AI model no longer needs so many GPUs; the company also announced the MI300A chip, which is shipping in small batches. To be used in the "El Capitan" supercomputer in the United States.

Lenovo Group: The annual revenue of infrastructure business such as AI server exceeds 2 billion US dollars

According to Jiemian News, on June 15, Lenovo Group ISG (Infrastructure Solutions Business Group) announced that its global AI infrastructure business annual revenue exceeded 2 billion US dollars, and announced its next-stage growth strategy, that is, additional investment will be made in the next three years $1 billion to accelerate AI deployment in enterprises worldwide. As part of the additional investment, Lenovo will spend US$100 million globally to grow the Lenovo AI Innovator Program, which has developed more than 150 leading AI solutions in its first year with 45 industry-leading ISV partners . With the new solutions, customers can more quickly deploy and leverage leading AI capabilities, such as generative AI, computer vision, voice AI and virtual assistants.

View

Zhou Hongyi: The basic level of domestic large models has caught up with GPT-3.5

According to The Paper, on June 13, at the press conference of the 360 Smart Brain Model, Zhou Hongyi, the founder and chairman of 360, said that the iteration speed of the domestic model has basically caught up with or approached the international advanced level. At present, the basic level of domestic large-scale models has caught up with GPT-3.5, and the speed of catching up with GPT-4 is also beyond my imagination. He predicted that the GPT-5 launched in the future will fully have multi-modal capabilities, and multi-modality will become the focus of competition in the future.

Wedbush: AI is driving tech into a '1995 moment', could rise 10% to 12% in second half of this year

According to the "Technology Board Daily" report on June 13, Wedbush Securities analyst Dan Ives wrote in a report on Monday that technology stocks are preparing for a long-term rebound and may rise by 10% to 12% in the second half of this year. While tech stock valuations are a key concern, Ives said he still believes AI is driving the tech industry into a "1995 moment," with future growth not seen since the 1990s. Currently in the AI arms race period, technology suppliers of software/chip/digital media ecosystem must build platforms around ChatGPT/other artificial intelligence models, which will change many industries in the next few years.

Musk: Humans are already cyborgs

Babbitt News, on June 12, Musk published a blog post on a16z partner Marc Andreessen "Why AI will save the world?" " commented. Musk believes that humans are already cyborgs because most of our memory is outsourced to computers. He also pointed out that there are still many things that humans cannot do with artificial intelligence, such as "thinking", which "has not been outsourced yet".

Researchers confirm that training AI with AI-generated results will lead to model degradation and even crash

According to news from IT House on June 14, a research team observed and recorded the "matryoshka training" of using AI-generated results to train AI. The results showed that "the use of model-generated content in training will lead to subsequent The resulting model is irreversibly flawed". Ilia Shumailov, one of the lead authors of the paper, also stated that "over time, errors in the generated data force the AI to further misperceive reality, and we were surprised to observe that the model collapse occurs quite quickly, and the model can quickly forget what they were originally. Much of the raw data from which to learn."

In addition, if the results generated by AI are manually polished and then put into model training, model "degeneration" cannot be avoided. For any large model, with too much learning data, they will inevitably be exposed to other AI-generated data, so the researchers said that "AI identification should be introduced to pick out the learning data that may be wrong" to improve the model learning ability and accuracy.

Yang Likun, Chief AI Scientist of Meta: ChatGPT does not yet have human-level intelligence

According to CNBC, Meta's chief artificial intelligence scientist Yang Likun (Yann LeCun) talked about the limitations of generative AI at the Viva Tech conference in Paris, France on Wednesday, saying that current AI systems like ChatGPT do not have human-level intelligence. AI systems aren't even as smart as dogs. The future of machines smarter than humans should not be seen as a threat.

Nikon: The beauty of nature captured through the lens and the creativity of photographers cannot be replaced by AI

According to news from IT House on June 16, creative companies represented by Adobe have begun to embrace artificial intelligence and integrate it into their own photography and design software. Camera maker Nikon does not recognize the AI trend in its latest video, arguing that the natural beauty captured through the lens and the creativity of photographers cannot be replaced by AI. It is reported that Nikon first launched the ad in Peru, encouraging photographers not to turn to generative AI, but to see the world through a creative lens. The bottom tagline is "Don't give up the real world."

Research Report

Bloomberg Intelligence: Generative AI Market Size Expected to Grow to $1.3 Trillion by 2032

According to a report by "Kechuangban Daily" on June 13, a Bloomberg Intelligence report pointed out that in the next 10 years, the market size of generative AI will grow from 40 billion US dollars in 2022 to 1.3 trillion US dollars in 2032, with a compound annual growth rate. It is expected to reach 42%. Generative AI is expected to account for 10%-12% of IT hardware, software services, advertising spending, and gaming market spending by 2032, up from less than 1% today.

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

According to a report by IT House on June 14, the latest developer survey report released by the programming question-and-answer website Stack Overflow shows that although programmers are skeptical about the accuracy of various AI-driven tools, it does not prevent them from being used in the development process. Tools like ChatGPT.

According to the survey results, 77% of programmers surveyed welcome the use of AI in their workflow, and 70% of respondents said they have programmed using AI tools this year. Thirty-three percent of respondents believe that the introduction of AI tools can increase productivity; 25% of respondents believe that it can speed up the learning process.

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

According to a report by the McKinsey Global Institute, generative AI will generate $4.4 trillion annually for the global economy, according to the New York Times. Half of all jobs will be automated between 2030 and 2060, the report says. McKinsey had previously predicted that AI would automate half of all jobs between 2035 and 2075, but generative AI has accelerated the agency's predictions.

Report: China's cloud service spending increased by 6% in the first quarter, and the AI wave may bring new growth opportunities to the cloud service market

According to a report by the Financial Associated Press on June 15, a report by the analysis agency Canalys shows that in the first quarter of 2023, spending on cloud infrastructure services in mainland China increased by 6% year-on-year to US$7.7 billion, accounting for 12% of the overall global cloud spending. Enterprises' demand for cloud migration remains sluggish, and the overall market has maintained single-digit growth for three consecutive quarters. The need for telecommuting and online meetings has subsided as pandemic restrictions have eased. Businesses remain cautious about their IT budgets. The consumption of cloud computing in the stock market has slowed down, and the investment in cloud computing in the incremental market has been limited, resulting in a decline in the overall growth rate of China's cloud computing market. As ChatGPT leads the AI wave, generative AI and AI large models are highly sought after, which is expected to bring new growth opportunities to the cloud service market.

GitHub report: 92% of US programmers are using AI tools

According to a report by IT House on June 15, GitHub, an online code hosting platform, recently conducted a survey of American companies with more than 1,000 employees. The questionnaire surveyed 500 enterprise developers, hoping to understand the impact of artificial intelligence on developers' work.

The survey results show that 92% of programmers are using various AI tools, and 70% of programmers believe that the introduction of AI tools has significantly improved work efficiency. Programmers surveyed said AI tools can improve code quality, speed up output, and reduce production-level incidents. This means that AI programming tools are an essential part of modern business IT.

McKinsey: Around 2045, 50% of jobs will be taken over by AI

According to news from IT House on June 15, a report by consulting firm McKinsey called "The Economic Potential of Generative Artificial Intelligence" shows that by around 2045, 50% of jobs will be taken over by AI. In the report, McKinsey claims that "the time for AI to replace human jobs" has been brought forward, "The new AI adoption scenario considers the development of generative artificial intelligence, modeling the time spent on work activities in 2023, and the 50% of occupations will be replaced by AI between 2060 and the midpoint is 2045, an acceleration of about a decade compared to previous estimates."

The report also shows that the current AI may add "US$2.6 trillion to US$4.4 trillion" to the global economy every year, which is equivalent to adding the GDP level of a new country on the planet.

Report: AI server shipments will be less than expected, accounting for less than 1% in 2023

According to a report on June 15 by the Science and Technology Innovation Board Daily, although generative artificial intelligence is very hot, many technology companies and cloud services plan to purchase AI servers and integrate them into existing systems. However, shipments of such products will not grow as significantly as expected, and it is expected that they will account for less than 1% of all server shipments in 2023.

Report: 70% of HRs are concerned about the application of AI in the fields of intelligent talent assessment analysis and intelligent resume screening

On June 15th, 36 Krypton Enterprise Services Dianping released "Organizational Change and Recruitment Upgrade under Human-Machine Symbiosis-AI + Recruitment Application Guide". According to the report survey, 75% of the interviewed HRs believe that their positions will be partially replaced by artificial intelligence; only 3% of the interviewed HRs believe that artificial intelligence will completely replace their positions; about 70% of HRs care about AI in talent intelligence assessment Applications in the fields of analysis and intelligent resume screening; salary performance management and employee training are also AI application scenarios that many HRs care about.

Data: About US$8.7 billion has been invested in AI startups in the past three months

About $8.7 billion has been invested in artificial intelligence startups in the past three months, including CharacterAI’s $150 million Series A round, according to data from startup database Crunchbase, the Science and Technology Innovation Board Daily reported on June 16. , Cohere's $270 million Series C financing, MistralAI's €105 million seed round financing, and Runway's $100 million Series D financing.

Quark released the 2023 college entrance examination volunteer report: the attention of AI-related majors is rising the fastest

According to Jiemian News, on the afternoon of June 14, Quark released the "2023 College Entrance Examination Volunteer Report" based on Quark search and volunteer form data. The report shows that among the popular undergraduate majors, the three traditional popular majors of computer science and technology, law and clinical medicine still firmly occupy the top three. With the popularity of large models and generative AI, majors related to artificial intelligence have the strongest upward momentum.