作者归档 admin

ChatGPT is simpler than core-making? Zhou Hongyi: China can catch up with the United States in two or three years.

Because China’s science and technology started late, it is at a backward level in many scientific and technological fields, such as chips. However, western countries have already set up many technical barriers in the field of chips, so it is quite difficult to build a "Chinese core".

Of course, in addition to the chip field, the United States has also taken the lead in cutting-edge scientific and technological fields such as cloud computing, artificial intelligence and quantum computing, such as ChatGPT. With the global explosion of ChatGPT, technology companies all over the world are restless, and Google and Apple are scrambling to do it, and Chinese companies are no exception. So, in terms of ChatGPT, does China have a chance to catch up with the United States?

In response to this question, 360 Zhou Hongyi gave an unexpected answer. He believed that ChatGPT was easier than core-building. With the support of the state’s encouraging policies, China could catch up with the level of the United States in two or three years.

The reason why Zhou Hongyi can be so sure is because he knows enough about ChatGPT. As early as 202, when the first generation of ChatGPT came out, Zhou Hongyi had paid attention to it and conducted in-depth research. In his view, the success of ChatGPT has mainly done five things right, namely, open source, win-win cooperation, long-term persistence, ecology and user traffic.

First of all, when developing ChatGPT, OpenAI used a large number of open source algorithms and papers, and chose to cooperate with a large industrialized company like Microsoft, which saved itself a lot of trouble and solved the problems of computing power, data, business model, engineering and so on.

Secondly, from the very beginning, OpenAI was ready to fight a "protracted war", and like Apple, it worked hard to build an ecology and cultivated vertical applications based on ChatGPT. Thus attracting more users to use, and using user traffic can strengthen the learning and training of ChatGPT.

This is where China can learn. At present, if China wants to build its own ChatGPT, it can do it in terms of technology, strategy and language materials. Coupled with the advantages of open source technology and strong engineering and technology landing ability of China enterprises, it can already make ChatGPT-like products, so to some extent, making ChatGPT is easier than making cores.

However, after making a sample of ChatGPT, we need to spend more time on technological innovation, ecological construction and cooperation with Industry-University-Research, plus the support of the state’s encouraging policies. As Zhou Hongyi said, it will take two or three years for Chinese enterprises to catch up.

On the whole, ChatGPT is of great significance to the development of artificial intelligence in the future. Bill Gates describes ChatGPT as "as important as the Internet", Musk describes ChatGPT as "good and terrible", and Zhou Hongyi thinks that ChatGPT will bring a new round of information industry revolution. Technology bosses have high hopes for ChatGPT, and China version of ChatGPT can’t be left behind! Zhou Hongyi’s suggestion is worth pondering. What do you think? Welcome to comment!

Di livio: Support Mancini to go abroad to find a striker and solve the Italian national team’s front line problem

Live on March 12 th, Italian national team coach Mancini is going to look for a striker abroad to solve the problem of the national team’s front line. The Italian famous Sudilivio expressed support for this.

Mancini had previously publicly stated that there was a big problem in the Italian national team’s front line, but the media revealed that he might call up Compagno, who plays in Romanian league, and Retegie, who plays in Argentine league. Di livio said: "At present, he is in a delicate period. Obviously, the national team lacks strikers and needs to find a solution. I agree with Mancini that he went abroad to find players to solve this problem. We must find players in this position in the next few years and find excellent and young strikers, and they need to adapt quickly in a short time. "

Di livio went on to say: "When I played football, there were many excellent strikers who could not be selected for the Italian national team, because there were many candidates for this position, and the coach was even criticized for not selecting a striker. The problem now is that young players sometimes can’t take the final key step. They may get lost, but this is a big problem for Mancini. "

Asked by the media how the Serie A league and clubs should help the Italian national team, Di livio replied: "I think they are already trying to help Mancini. The club is signing many young players, something is happening, but maybe there is a lack of quality in the process of finding players. "

Application of AI customer service robot in beverage retail industry

With the rapid development of artificial intelligence technology, intelligent customer service robots have widely appeared in our lives, helping customers in various fields to deal with massive and complicated sales, consulting and service problems quickly. Intelligent customer service robots can accurately understand visitors’ intentions through autonomous learning, answer a large number of repetitive and standardized questions, and greatly improve service efficiency and user experience.

At present, the demand for customer service in the beverage retail industry is mainly concentrated on online customer service, with pre-sales consultation and after-sales service as the mainstay. Due to the large amount of consultation, repeated questions and difficult control of service effect, it is necessary to reduce the pressure of manual customer service through intelligent customer service robots, and at the same time improve the customer service experience and track and grasp the customer service effect in time.

1. Intelligent text robot

Based on NLP algorithm of deep learning, text robot has strong natural language understanding and generalization ability, and can accurately identify the real needs expressed by users; And based on the reinforcement learning model driven by big data, intelligent autonomous learning is realized, and the analogy is made. For the whole process of pre-sale and after-sale, independent reception and self-service save 85% of labor costs; 24×7 real-time standby, so that customers can enjoy intelligent services at any time; Intelligent auxiliary manual customer service, service efficiency increased by 100%; Business analysis, knowledge precipitation, release enterprise maintenance costs; Thousands of people, providing personalized intelligent customer service experience.

2. Intelligent voice robot

The intelligent voice robot responds intelligently according to the intention of the customer, and the voice recognition is rapid, which can accurately judge whether it is an intentional customer, and liberate the customer service staff from repetition, mechanical preliminary screening and intention recognition and invest in more valuable customers. Traditional outbound calls have high labor cost and management cost, and the service quality cannot be guaranteed, and the management is complicated. Therefore, mechanical, repetitive and boring telephone dialing tasks can be handed over to intelligent voice robots for assistance.

3. Intelligent outbound robot

Intelligent outbound robot is a typical application of artificial intelligence in speech recognition. It can automatically initiate outbound calls and actively introduce products to users in the form of natural human voice synthesized by speech. "Thousands of people and thousands of faces" outbound strategy to improve the success rate. Based on past operational data, intelligent voice robots can accurately sketch user portraits from the dimensions of age, gender, occupation, repayment ability, repayment habits and historical repayment records. Combined with the specific characteristics of different scenarios such as marketing, collection and return visit, the communication strategies and communication techniques for users with different attributes are continuously adjusted in each dialogue process node to improve the goal achievement rate.

GaussMind, an AI scene landing expert of Wofeng Technology, has rich AI products and years of experience in AI landing application, serving tens of thousands of customers including more than 60 Fortune 500 companies and more than 200 China Fortune 500 companies. GaussMind, based on deep neuroscience algorithm and convolutional neural network algorithm, integrates many cutting-edge technologies such as automatic speech recognition (ASR), natural language processing (NLP) and speech synthesis (TTS), and builds a typical product of voice interaction, AI outbound robot, hoping to provide outbound robot solutions for more industries, make enterprise marketing transformation more efficient and make enterprise service experience more heart-warming.

AI doctor assisted diagnosis and accurate classification of sinusitis

Chronic sinusitis is a highly heterogeneous disease involving nasal cavity and sinus mucosa, and it is one of the most common chronic respiratory inflammatory diseases. The prevalence rate in China is about 8%, and 20% of patients are chronic sinusitis complicated with nasal polyps. In order to make the classification and treatment of nasal polyps more efficient and accurate, the research team of the Third Affiliated Hospital of Sun Yat-sen University developed the first artificial intelligence pathological diagnosis system for chronic sinusitis. It is reported that the "Single Vision Home Edition" of the system has passed the clinical test, reaching the level of a senior pathologist, and has been tested and used in a number of 3A hospitals with 5G network systems in Guangdong-Hong Kong-Macao Greater Bay Area.

Patients with nasal polyps need comprehensive treatment mainly by surgery, but the curative effect is not satisfactory at present, and the postoperative recurrence rate is high. The reason is that there are different subtypes of nasal polyps, and the treatment methods and prognosis of different subtypes of nasal polyps are very different. Therefore, accurate classification and individualized treatment of nasal polyps is one of the main ideas to break through the bottleneck of curative effect of nasal polyps. However, at present, the commonly used manual diagnosis method is to randomly select 10 visual fields of nasal polyps slides under a microscope and calculate the number of inflammatory cells in them, which has huge workload, low accuracy and impossible to calculate the whole film.

The Third Hospital of Sun Yat-sen, in conjunction with Tsinghua Pearl River Delta Research Institute and other hospitals and research institutes, has developed and iterated (at the cellular level, molecular level/full film and single vision) an artificial intelligence pathological diagnosis system for chronic sinusitis, developed a "full film platform version" and a "single vision family version" of AI doctors for different application scenarios, and built a cloud platform.

Among them, the "full-film platform version" of AI doctor is an artificial intelligence pathological diagnosis system for nasal polyps based on full-film, and related research results have been published in high-level journals such as JACI, eBioMedicine, a The Lancet sub-journal, and Chinese Journal of Otolaryngology and Head and Neck Surgery. The "Single Vision Home Edition" system has been developed and tested. The AI doctor "Sandy" can accurately identify the common inflammatory cells of chronic sinusitis and generate the first artificial intelligence pathological diagnosis report of chronic sinusitis. According to the standard of senior pathologists, its accuracy has exceeded 90%, which is helpful for accurate diagnosis of chronic sinusitis and subsequent individualized medication guidance.

□ Zhou Jinan Zhen Xiaozhou

Let AI write its own importance.

With the continuous development of science and technology, artificial intelligence (AI) has been widely used in various fields, and its application in enterprises has been paid more and more attention. AI technology can use big data, machine learning and natural language processing technology for intelligent analysis and decision-making, so as to improve production efficiency, reduce costs, shorten processing cycle and optimize management, and help enterprises better cope with market changes and competitive pressures. This article will discuss the importance of AI in the work from the following aspects.

In the traditional manufacturing industry, a large number of production tasks require the input of manpower and material resources. However, with the application of AI technology, intelligent equipment such as robots can replace manpower and complete a lot of repetitive, dangerous and tedious manual labor in the production process, which improves the production efficiency.

At the same time, AI technology can also optimize and manage the production process of enterprises, and improve the management level of production planning, logistics distribution and quality control of enterprises through big data analysis and intelligent decision-making.

For example, a chemical company under the German industrial manufacturer Siemens has created an intelligent factory by using AI technology, which can independently control and optimize its own production process, thus improving production efficiency and quality.

With the help of AI technology, enterprises can more accurately grasp the quantity and time of materials and accessories needed for production, thus realizing the optimization of supply chain and cost control. In addition, AI can also predict the time required for equipment failure and maintenance by analyzing data, so as to better schedule machine maintenance and replacement, thereby reducing operating costs and production downtime.

Tesla, an American automobile manufacturer, widely uses AI technology in the production process. Through fine scheduling and optimization of automation equipment, the processing cycle of its production workshop is greatly shortened, thus improving production efficiency and reducing production costs.

AI technology can create new business models and opportunities, thus helping enterprises to stay ahead and innovate in the market competition. For example, intelligent customer service based on AI technology, through speech recognition technology and natural language processing technology, can deal with a large number of user feedback and problems, improve customer satisfaction and user stickiness, and also create new service benefits and business opportunities for enterprises.

In addition, in the medical field, AI technology can also assist doctors in diagnosis and treatment, thus improving medical level and reducing medical costs. Shenzhen Traditional Chinese Medicine Hall has developed a big data platform based on traditional Chinese medicine by using AI technology, which comprehensively absorbs the information of national traditional Chinese medicine medical records, prescriptions, experiences, modern science and technology, and helps doctors to diagnose and treat diseases through AI intelligent analysis.

AI technology can realize intelligent recommendation and search according to customers’ needs and feedback, and improve customer experience and service quality. Large-scale e-commerce enterprises using AI technology can provide personalized recommendations and offers to customers through information such as customer purchase history and behavior data, and improve customer shopping experience and loyalty.

In addition, AI technology can also help enterprises to identify and analyze customers’ voices, emotions and attitudes, so as to better understand customers’ needs and emotions and optimize customer service.

Through the integration of AI technology, enterprises can better cope with risks and security issues. For example, in the financial industry, AI technology can help enterprises reduce risks and losses through customer identification and fraud identification. In the field of production, enterprises can avoid production accidents and safety problems through intelligent sensing technology and autonomous control technology.

With the continuous progress of science and technology, AI technology is more and more widely used in enterprises, which can improve production efficiency, reduce costs, optimize management, create new business models and improve customer service quality. However, there are also many problems and risks to be considered in the application, such as data security, privacy protection, failure of artificial intelligence technology and so on. We need to fully realize the advantages and limitations of AI technology, so as to make better use of this technology and promote the development and innovation of enterprises.

China version of chatGPT is coming, what surprises will Baidu bring us?

ChatGPT-like chat intelligence AI launched by Baidu has better performance and applicability in Chinese context, and may gain more users and application opportunities in China market. This also reflects China’s continuous efforts and innovations in the fields of natural language processing and AI.

increase efficiency

Baidu version of ChatGPT can handle conversations faster and more accurately than before. This improvement in speed and accuracy can lead to more effective interaction between human and artificial intelligence.

The improvement of the efficiency of Baidu chatGPT can also bring more cost-effective solutions to enterprises that need to use chat bots for customer service or other tasks.

More natural interaction

Baidu version of ChatGPT can better understand the context of the conversation, which can lead to a more natural interaction between people and AI. This can make the interaction more pleasant and efficient.

The higher accuracy of Baidu version of chatGPT also means that it can better answer a wider range of questions, thus obtaining more accurate answers and better user experience.

Baidu version of ChatGPT can produce more human-like reactions to dialogue, which leads to a more natural interaction between human beings and artificial intelligence. This can make the conversation more fascinating and enjoyable.

The improvement of the accuracy of Baidu chatGPT also means that it can better understand the context of the conversation, thus bringing more accurate response and better user experience.

Smarter solutions

Baidu version of ChatGPT can generate smarter dialogue solutions, thus making the interaction between people and artificial intelligence more effective. This can make the dialogue more effective and efficient.

The improvement of the accuracy of Baidu chatGPT also means that it can better understand the context of the conversation, thus bringing more accurate solutions and better user experience.

"Talk" and AI will know what you are talking about.

Transfer from: Hohhot Daily

It is conceivable that due to the huge potential demand in the fields of public welfare, public safety and national security, and the strong promotion of the rapid development of AI technology, in the near future, AI lip reading is expected to achieve rapid promotion and deep popularization, and the industrial prospects are very promising.

According to the continuous breakthrough of the bottleneck, it has become a reality that AI technology has achieved great success in the field of lip language recognition.

Many problems need to be broken.

However, Yan Huaizhi also said that at present, China’s artificial intelligence lip language recognition technology is still in its infancy, and there is still a long way to go if we want to use artificial intelligence to accurately recognize lip language.

From the perspective of language itself, human language has a high complexity. Of all the phonetic symbols involved in human speech, only about 30% are directly controlled by human lips, and 70% are teeth sounds, tongue sounds and throat sounds that are difficult to distinguish by naked eyes or even machine vision. Moreover, different people’s tone of voice, dialects, conjunctions, accents, and even beard cover will all lead to subtle changes in mouth shape, and it is this subtle change that will seriously affect the recognition and judgment of lip language by artificial intelligence.

From a technical point of view, the environment for artificial intelligence to collect lip language is usually complicated, and it is very difficult to accurately identify it. As far as the current artificial intelligence technology is concerned, the recognition level of long sentences and complex sentence patterns is not satisfactory, not to mention the problems of multi-scene recognition and lip recognition of multi-person images.

Yan Huaizhi said that only by solving the above problems can AI achieve a breakthrough in lip reading and move towards a mature stage of development.

There are many differences between different languages of human beings. Can AI read the lips of each language?

Yan Huaizhi introduced that most of the successful AI lip reading systems were limited to English models, because most AI models were trained based on English data. However, from the technical framework, the training models of different languages are basically the same, or they can be realized by the same kind of technical means.

Of course, in order to adapt to lip language recognition in different languages, some adaptive adjustments need to be made: on the one hand, data in corresponding languages should be selected for targeted training; On the other hand, the AI model needs to be adjusted, such as incorporating time masking, optimizing language model and improving superparameters.

In addition, the same language will have different mouths, even if the mouths are similar, they may represent completely different meanings. Therefore, a mature AI lip reading system needs a large number of lip feature sample data, and covers as many application scenarios and different types of speakers as possible, so as to improve the generalization ability of the trained lip recognition model and improve the recognition accuracy of AI lip reading for different mouth shapes and different ideographic languages.

Technical double-edged sword in urgent need of supervision

Despite all kinds of difficulties, more and more AI companies have begun to set foot in and plan to deepen the artificial intelligence lip recognition track. At present, the choices of major AI giants are different, which can be divided into lip language data, lip language video recognition, lip language understanding and so on.

Yan Huaizhi also said that at present, many artificial intelligence lip recognition technology fields have achieved initial breakthroughs, the prospect of full chain integration is expected, and industrial clusters are gradually taking shape.

From the perspective of application scenarios, AI lip reading has begun to emerge in the fields of social welfare and public safety. Judging from the current layout of the giants and the development trend of related technologies, AI lip reading is expected to have broad application prospects in identity recognition, national security, intelligent systems and so on. "It is conceivable that due to the huge potential demand in the fields of public welfare, public safety and national security, and the strong promotion of the rapid development of AI technology, in the near future, AI lip reading is expected to achieve rapid promotion and deep popularization, and the industrial prospects are very promising." Yan Huaizhi said.

Of course, technology application is a double-edged sword. Many people worry that lip-reading by AI will reveal the private content in people’s conversations, whether the parties are speaking publicly, whispering or talking to themselves. "Zhang Zhangkou" was stolen by others, and it was really terrible to think about it carefully.

Yan Huaizhi said that this kind of worry is not unfounded. On the one hand, the privacy leakage caused by AI lip reading may be caused by malicious lip reading, on the other hand, it may be the normal use of AI lip reading system, but the improper protection of storage and use leads to the theft or abuse of relevant data, which in turn causes damage to personal rights and interests. Moreover, because it involves the conversation content of the parties and has obvious directionality, this kind of privacy disclosure may be more harmful than ordinary personal information disclosure.

Therefore, Yan Huaizhi suggested that from the perspective of privacy protection, we should strengthen the formulation of relevant laws and regulations at the management level, strictly regulate and restrict the application scenarios, scope and purposes of AI lip reading, and increase the supervision and punishment of malicious use of technology. In addition, it is necessary to strengthen the construction of the security protection system of AI lip-reading system at the technical level, improve the recognition accuracy of the system by technical means, avoid technical abuse, and effectively ensure the content security of user conversations.

ChatGPT triggers change, dragon information builds data analysis and empowerment tools

You can code, write poems, translate novels, even take exams and make online consultations … ChatGPT, an artificial intelligence chat robot that is responsive, has exceeded 100 million monthly active users only two months after its launch.

ChatGPT’s popularity has once again triggered widespread concern about the development of artificial intelligence technology around the world. ChatGPT adopts the route of "big data+big computing power+strong algorithm=big model" in the technical path, and explores a new paradigm in the direction of "basic big model+instruction fine-tuning", in which the basic big model is similar to the brain, and instruction fine-tuning is interactive training, and the combination of the two realizes the language intelligence approaching human beings.

At present, the development of AI mainly relies on large-scale model technology, and it is necessary to train very large-scale models with huge parameters based on massive natural language or multimodal data sets. To successfully train a large model with larger parameters, higher accuracy and higher capability, not only a huge amount of high-performance AI computing power is needed to support it, but also a large number of high-quality data sets obtained by careful cleaning and an efficient data platform are needed to ensure a long-term model training process.

After more than 20 years’ experience in data management, Julong Information adopts technologies such as big data, cloud computing and artificial intelligence combined with hierarchical architecture design of data warehouse to build a smart cube platform. Through algorithms such as relationship mining, time series mining and spatio-temporal mining, the standard data model is uniformly built in the way of business backward, so as to realize in-depth mining, serial-parallel analysis, early warning and prediction, etc., and meet the goal of "artificial intelligence driving business to improve efficiency".

Introduction to the Application Scenarios of Smart Cube Platform

Portrait analysis: based on the ontology conceptual model of the entity, build the entity portrait and refine the label. Through natural language processing technology and structured data processing technology, the emotion mining of characters is carried out by using template rules, automatic mode and mixed mode. Finally, according to the business needs, we use tag mining, emotional polarity analysis, similarity analysis, relationship analysis and other algorithms to create a portrait of the entity for analysis.

Correlation analysis: Based on the behavior, relationship and portrait data of knowledge map construction, the potential correlation between objects is portrayed by using random walk of the map, factor association, community discovery, intention recognition, semantic search, relationship mining and similarity algorithm, which helps the police to make in-depth judgment and analysis.

Clue mining: Deep-level mining is carried out in combination with emotional business scenarios, and hidden information such as people, clues, elements, relationships, behavior patterns, etc. are found through frequent item mining, association mining, classification and clustering, and anomaly detection technologies to assist the police in deep judgment and analysis.

Prediction and early warning: According to the temporal and spatial distribution characteristics of historical events, the future development trend of events is predicted and judged by using node intimacy calculation, narrative event evolution diagram, element extraction, correlation analysis, relationship mining, comparison collision, behavior pattern analysis and temporal and spatial analysis algorithms, such as event trend prediction and spatial hot spot prediction.

Case sharing: At present, the crime of electric fraud continues to be high, resulting in huge asset losses, accounting for nearly 50% of the criminal cases. At the same time, electric fraud has the characteristics of strong concealment of crime, great difficulty for the masses to prevent publicity, and rapid renovation of means. Based on more than 12,000 case transcripts, 615 million population data and more than 20,000 phone bill data collected by law enforcement users in a certain place as data input, the platform was analyzed, modeled and mined. After multi-level data screening and analysis and modeling, 10+ high-value and vulnerable portrait features were finally analyzed and mined, which was highly recognized by users.

Julong Information Artificial Intelligence organically organizes billions of data into a knowledge network that conforms to people’s cognitive style in a scientific, reasonable and efficient way, making the data easier to be understood and processed by people and machines, and providing application support such as search, analysis, mining, application, presentation, prediction and early warning in various business scenarios.

The development of artificial intelligence is inseparable from the combination of human and artificial intelligence, and our future will also be an era of co-evolution of human and artificial intelligence.

ChatGPT depicts the future world of artificial intelligence

Robots are like humans.

Thought and wisdom have surged

In this dreamy world

Seamless combination of technology and human beings

Life is like a poem.

Calm and beautiful, never stop.

one

Two

three

four

five

Liverpool 0-1 Bournemouth relegation team Salah missed the point and the whole team was in a downturn.

"I have always hated losing," Salah said in a conversation with Gerrard this week. "I don’t like losing."

Liverpool supporters traveled long distances to the south coast of England, only to see the performance of their team once again. After klopp’s team was defeated by Bournemouth, they would undoubtedly agree with the Egyptians.

When substitute jota’s header hit Smith’s arm, Liverpool got an excellent opportunity to change the result of the game. The referee awarded Liverpool a penalty through VAR replay.

Milner, so reliable on the court, only played a few minutes ago, and his cross created this penalty opportunity. However, Salah is Liverpool’s first player. However, his shot was far from the right post of the goal, which greatly reduced the possibility of Liverpool becoming the top four for the first time this season.

Salah’s performance before he missed the penalty was hard to convince. He kept losing the ball, made the wrong choice and was easily intercepted.

Think less than a week ago, the 30-year-old player surpassed Robbie Fowler in the 7-0 victory over Manchester United to become the player who scored the most goals in Liverpool’s Premier League history, and he was happy to tear off his jersey. How high-spirited he was at that time.

However, Salah was not the only Liverpool player with a low performance. After 20 minutes of hopeful opening, the performance of the whole team was disappointing.

In five Premier League away games against Nottingham Forest, brentford, Brighton, Wolves and Bournemouth, Liverpool conceded 11 goals and scored only one. The recent setbacks are becoming the norm on Liverpool’s way forward.

The gloomy expression on Liverpool players’ faces on their way back to the dressing room after the game reflects their mood. I wonder if they can improve before the second leg of the Champions League against Real Madrid on Wednesday.

Six core technologies, the small I robot has great commercial value potential.

The advantages of artificial intelligence small I robot are not only in corpus data, but also in its own hard technology.

Compared with other big manufacturers, cognitive intelligence is still in the stage of research and development. Small I robots have been put into practice for a long time, and on this basis, they have quietly established their own business empire.

Xiaoyi commercialized all his six core technologies, and formed nine product series including dialogue AI, knowledge fusion, intelligent voice, super automation, data intelligence, intelligent construction support, visual analysis, intelligent hardware support, and meta-universe. The market share of products in several vertical industries is in the forefront.

With outstanding technology, the small I robot has blossomed rapidly in intelligent services. At present, there are more than 800 million end users of the small I robot worldwide. Nearly a thousand large and medium-sized enterprises and governments, hundreds of thousands of small enterprises and developers are using the services and intelligent robot products provided by the small I robot, covering communications, finance, government affairs, legal affairs, medical care, manufacturing and other industries.

In the banking industry alone, more than 40 of the top 50 banks in China have adopted the small I intelligent customer service system. According to the official data of China Construction Bank, the workload of intelligent customer service is almost equivalent to 9000 employees, which can save a lot of astronomical figures only in labor costs.