At this year's two sessions, "artificial intelligence (AI)" has become the focus of heated discussions. The government work report not only mentioned it many times, but also put forward the "artificial intelligence +" innovation action for the first time, demonstrating the deep insight and forward-looking layout of scientific and technological development.
Looking back on the past reports, from "Internet +" to "Smart +", each time it has led the trend of industrial transformation. This year's "artificial intelligence +" action has jumped to the strategic level, which is not only the promotion of technology, but also the call for new productivity. This action marks the solid pace of the mainland in the field of science and technology, and points out a new direction for industrial transformation. We need to respond positively, strengthen R&D and application, promote deep integration of industries, and pay attention to ethical safety to ensure that scientific and technological achievements benefit the people and promote social harmony and progress.
In 2024, the transformation and breakthrough of domestic large models
The top of the intelligent wave: the opportunities and challenges of domestic large models
At the top of the wave of intelligence, domestic large models are standing on the cusp of the times, and opportunities and challenges coexist. The sword of artificial intelligence has pierced old industries, stimulated new vitality, and become a new engine to promote the national economy.
At the Boao Forum for Asia Annual Conference, the "Opening a New Era of Intelligence - 2024 China AI Large Model Industry Development Report" released by People's Daily Online reflects the booming trend of China's AI industry. According to the data, the adoption rate of generative AI enterprises in China has climbed to 15%, and the market size is as high as 14.4 trillion yuan. Looking ahead, by 2035, generative AI will contribute nearly 90 trillion yuan to the global economy, and China will contribute more than 30 trillion yuan.
Many Chinese companies are bravely at the forefront of the trend and devoting themselves to the blue ocean of AI large models. The TASS model of real intelligence, Baidu's "Wenxin Yiyan", and Alibaba's "Tongyi Qianwen" not only have significant technological breakthroughs, but also show great potential in business applications and social productivity changes.
Specifically, let's take the TASS model of real intelligence as an example. As a quasi-unicorn in China's AI industry and a leading enterprise in the RPA industry, and a hyper-automation solution provider, Real Intelligence combines domestic self-developed AI technology and RPA products to help government enterprises achieve digital reform, transformation and upgrading. Real Intelligence is an artificial intelligence technology company based on RPA Agent (agent) + AGI large model + hyper-automation technology, leading the era of human-machine collaboration. Compared with traditional RPA, the threshold for the use of real agent (agent) is lower, and the automated process and software robot are generated in one sentence, that is, Wensheng digital employee, the intelligent assistant that solves the long-tail and low-frequency automation needs of individual users, and the employee office assistant for government and enterprises, so as to realize the "you say the PC does" and "you say the mobile phone does" for various application software on the PC and mobile phones, and fully open the era of the intelligent twin.
Real Intelligence launched the first product-level real AI Agent in the whole industry - "Wensheng Digital Employee", that is, through the generation of automated processes and software robots in one sentence, it brings intelligent assistants to individual users to solve the long-tail and low-frequency automation needs, and brings employee office assistants to governments and enterprises, realizing "you say PC does" for various application software on PC and mobile phones, and fully opening the era of intelligent twins.
In practice, the agent is able to simulate and surpass manual operations, not only perform repetitive rule-based tasks, but also handle more complex decision-making processes, such as identifying potential risks in the bank loan approval process, interpreting unstructured data, and synchronizing data across systems and platforms.
In 2024, the transformation and breakthrough of domestic large models
However, challenges also come with them. The demand for computing power has surged, and the limitations of mainstream model architectures have become prominent, such as high computing power consumption and high memory usage, which hinder the commercialization process of domestic large models.
But as the ancients said: "The road is long, and I will go up and down to seek." "In the face of challenges, we must be firm in our faith and rise to the occasion. Strengthen technological innovation, improve the training speed and output quality, optimize the model architecture, reduce computing power and memory consumption, promote industrial collaboration, form a joint force, and jointly promote the development of domestic large models.
Where is the future of large models?
In the great wave of science and technology, AI large models, as pioneers leading new quality productivity, are gradually unveiling the mystery of their potential and value. However, in order to make it play a greater role in production and life, we must deeply explore its application scenarios and future trends.
The report depicts a grand blueprint for the AI large model industry. The combination of general models and industry models will jointly promote the wide application of AI models in multiple fields. The general model shows strong capabilities in universal scenarios, while the industry model focuses on specific fields and provides customized solutions for the industry.
For individual users, the choice between cloud and device-side large models is crucial. The cloud-based model relies on powerful computing power to provide extensive knowledge acquisition and language understanding, while the device-side model focuses on user interaction experience and data privacy protection. In the future, the combination of cloud and device will become a new trend, bringing users a safer and more convenient AI experience.
In addition, the report also looks forward to the open source trend of general large models. This means that small developers will be able to take advantage of the capabilities of large models without worrying about computing power and complex model training, thus further promoting the innovation and development of the AI large model industry.
China's AI model is at a critical stage of technological breakthrough. A keen grasp of trends will be the key to breaking the game. We need to keep up with the pace of the times, continue to explore new application scenarios and technological innovations, and make AI models play a greater role in more fields.
In the future, AI models will be deeply integrated with all walks of life to create more unprecedented application scenarios and business models. At the same time, the continuous advancement of technology and the promotion of open source trends will make large AI models more user-friendly, efficient, and intelligent. We have reason to believe that in the near future, AI models will become an important force to promote social progress and economic development.
In 2024, the transformation and breakthrough of domestic large models
Break the game: face the challenges head-on and embrace the trend
At present, the digital employee solutions provided by Real Intelligence have been widely implemented in many fields such as government affairs, operators, finance, energy, manufacturing, and e-commerce, and their influence and practicability have been highly recognized by the industry. This solution not only serves more than 300 central state-owned enterprises and Fortune 500 companies, but also helps more than 2,000 leading digital enterprises achieve rapid business development.
Looking forward to the future, Real Intelligence will rely on its leading AI Agent technology to tailor exclusive digital assistants for every employee and position in the enterprise. These digital assistants will act like intimate assistants, providing intelligent assistance to employees, thereby greatly improving work efficiency and accuracy.
The vision of real intelligence goes far beyond that. It hopes that through the wide application of "artificial intelligence +" technology, it will inject strong impetus into the new quality productivity of the whole society. With the continuous advancement of technology and the deepening of applications, the real and intelligent digital employee solutions will show their great potential and value in more fields, leading a new round of leaps in social productivity.
In 2024, the transformation and breakthrough of domestic large models
We look forward to seeing that Real Intelligence will bring tangible benefits to more enterprises with its forward-looking technology and innovative solutions, and contribute to the sustainable development of social economy.