Our reporter Xing Meng
The daily active users exceeded one million in less than a week after its launch, and 100 million within two months... The AI chat robot ChatGPT became popular all over the world, and became a phenomenon-level consumer AI application for a while.
ChatGPT is expected to open up a large number of application scenarios in thousands of industries and trigger the transformation of the AI industry.
Technology giants such as Microsoft and Google have stepped down one after another, taking the lead in laying out new AI tracks.
At the same time, ChatGPT recently started commercial monetization and launched a paid subscription version, which further verified the AIGC (artificial intelligence technology generated content) industry business model and opened up a broad space for the development of the AI industry.
ChatGPT, which is highly favored by capital, is also experiencing "growing pains". Its strong dependence on big data and computing power will limit its rapid development. New applications and new models that exceed expectations are causing market concerns and possible compliance issues. , intellectual property rights and other risks will also bring new challenges to digital governance.
AI moves from "behind the scenes" to "front of the stage"
The development of emerging things does not happen overnight. ChatGPT, which is popular all over the world, is an inevitable product of the development of the artificial intelligence industry chain to a certain stage.
After years of accumulation of AI technology research and development, benefiting from the breakthrough of deep learning technology, the improvement of computing power, the accumulation of big data and other factors, it has achieved the "out of the circle" of this killer AI application.
In fact, as a productivity tool to replace manual labor, chatbots are not a new concept.
Restricted by limited technology, the early chatbots were more entertainment in nature. In recent years, more intelligent chatbots have gradually been applied to business scenarios such as customer service and search engines.
Chen Xiaohua, an expert in the evaluation of major new infrastructure projects of the Ministry of Industry and Information Technology and executive director of the Metaverse Collaborative Innovation Center of Beijing University of Posts and Telecommunications Science and Technology Park, told the Securities Daily reporter, "Compared with other chat robots, ChatGPT can better understand the dialogue context and generate The text takes into account the content of previous conversations, and outputs high-quality texts that are more in line with language logic and human values. At the same time, it also increases the ability to understand and generate code, greatly improving practicability and broadening application scenarios.”
Essentially, ChatGPT is a large language model trained on large amounts of text data, which enables it to generate human-like responses to a wide variety of inputs.
"ChatGPT has greatly improved the artificial intelligence response function through large-scale language learning model training, and can comprehensively organize and output the content of related topics according to people's requirements." Zheng Lei, chief economist of Samoyed Cloud Technology Group, said The "Securities Daily" reporter said that after intensive training, ChatGPT can generate articles, computer codes, etc., and can also complete some knowledge work processed by humans, and transform it into artistic expressions such as drawings according to the text content.
ChatGPT is both practical and functional, and is regarded as one of the most representative AI applications in history, and it has also raised more people's expectations for the new blue ocean of AI.
"Although the current ChatGPT can't bring fundamental changes to our production methods, it represents that the development of artificial intelligence has entered a new stage." Zheng Ding, president of the Shenzhen Information Service Industry Blockchain Association, said "Securities Daily" reporter said that at this stage, users will no longer be passive recipients, but an object that can participate in the production process. Life.
Concentrated emergence of commercial monetization applications
The industry believes that the "out of the circle" of ChatGPT may mean the arrival of an era of large-scale commercialization of AI.
Gao Peijie, a lawyer from Beijing Jingsh Law Firm, told the "Securities Daily" reporter, "ChatGPT has huge commercial value in the intelligent application from To B to To C. The ChatGPT interactive platform can help companies establish effective communication solutions with customers. And it will bring about new industry development and changes in the fields of education, medical care, automobiles, smart venues, and smart homes, which will not only improve service quality, but also reduce the service cost of enterprises."
In early February, OpenAI announced the launch of the ChatGPTPlus paid subscription package, which charges $20 per month, opening the road to commercial realization.
Users who subscribe to this package can enjoy additional benefits such as skipping queues during peak hours, quick response, and priority access to new functions on the basis of free services.
At the same time, the Huaxi Securities Research Report shows that in addition to ChatGPT, other mainstream applications such as NovelAI and Stable Diffusion have also started commercial monetization, and set multi-level subscription packages according to the number of times and quality of content generation.
As these mainstream applications gradually develop commercialization models, the commercialization potential of the AIGC industry is gradually released.
"Internet and digital technology make it easier to share ideas and content, but at the same time, it also brings digital content creation into a spiral upgrade cycle of strong demand, video, and creativity. AIGC industry is undergoing technological integration and transformation at this stage. It just caters to the needs of users." Chen Xiaohua said that AIGC applications can release value through human-machine collaboration, and have been widely used as tools in industries with high digitalization and rich content requirements such as media, e-commerce, film and television, and entertainment. , will become an important content production infrastructure for the Internet and even the Metaverse in the future.
Technology giants such as Microsoft and Google are also optimistic about the business prospects of AI chatbots, and they have stepped down to lay out new AI tracks.
According to public information, Microsoft will invest billions of dollars in OpenAI and promote the commercialization of ChatGPT, planning to integrate it into a series of products such as Bing search engine, Office family bucket, Azure cloud service, and Teams chat program.
Google plans to launch a chatbot, Bard, to compete with ChatGPT.
Zheng Dingxiang believes that as an important part of the digital economy, the development prospects of artificial intelligence are beyond doubt.
As the latest achievement in the field of AI, ChatGPT also has enough room for imagination and development.
At present, technology giants such as Microsoft and Google are firmly investing in AI, showing broad development prospects for AI applications, and the AI industry is expected to start a new wave of development.
Domestic giants ahead of schedule
In fact, domestic Internet companies are also foresighted and have laid out the AIGC track ahead of time.
"China currently has the world's leading level in natural language recognition, and the relevant technologies are mainly in the hands of leading companies such as Baidu, HKUST Xunfei, and Hanwang." Zheng Lei said.
It is reported that the current cases of commercialization in my country include Baidu AIGC digital human anchor Du Xiaoxiao, Baijiahao TTV and so on.
In addition, the reporter learned that Baidu's AI chat robot ERNIE Bot is expected to complete internal testing in March this year and open to the public.
Gao Peijie said that from a technical point of view, domestic AI chatbots are relatively mature in technology, which needs to rely on huge database resources.
In the future, the AI industry will grow stronger in the competition of technology giants. From market research and development to database support, AI chatbots will achieve continuous iteration.
In addition, the current version of ChatGPT also has certain limitations, which has also caused controversy and concern in the market.
Gao Peijie believes that currently ChatGPT cannot carry out self-innovation, which makes it unable to understand logical thinking more deeply.
Zheng Lei said that the technical route of ChatGPT is different from the principle of human brain learning, and there may be bottlenecks caused by limited computing resources and algorithm limitations.
Zheng Dingxiang believes that the biggest hidden danger of ChatGPT is the lack of data. ChatGPT wants to obtain more extensive data or will face a series of difficulties such as privacy security, economy, and technology.
Chen Xiaohua analyzed that precisely because it is more advanced, the degree and breadth of risks brought by ChatGPT have also increased.
Limited by modeled learning materials, ChatGPT cannot guarantee the accuracy of its answers, which may mislead users. If it is later used in search engines, customer service and other fields, the authenticity of the information it replies may lead to legal risks.
At the same time, text and image data mining and re-creation involves compliance issues, and it is difficult to determine whether AI-produced content is infringing.
The current AI legislation at the national level is in the pre-research stage, and the regulation of the AI field needs to be improved.