Since the release of ChatGPT in December 2022, the “shockwaves” regarding the rapid development of AI threatening human jobs have been continuous: In March this year, director Lu Chuan stated that AI helped him create a movie poster, “The effect produced in 15 seconds is better than what a professional poster company can do in a month”; in April, an internal email from BlueFocus was exposed, requesting to “completely stop outsourcing copywriting” and switch to generative AI; in May, IBM announced it would pause hiring for positions that AI could handle, potentially replacing 7,800 jobs with AI; in June, the U.S. employment consulting firm “Challenger” disclosed that the number of layoffs in the U.S. has surged this year, with 3,900 jobs replaced by AI… People have gradually shifted from initial excitement and anxiety to facing the reality: AI is here — how should businesses, departments, and individuals respond?
AIGC Empowers Employees
Ren Zhengfei said that in the future, there will only be two types of people: those who are proficient in using AI tools and those who can produce AI tools.
The first priority is a change in cognitive levels. Corporate employees must actively embrace new things rather than passively resist. Currently, everyone has a certain awareness of large models represented by ChatGPT, which span a wide range of fields from text, images, to video and programming, showcasing impressive skills. We not only need to understand it but also need to create value together with it.
1. Acquiring Knowledge
Coming from the era of networking, we have become accustomed to using search tools like Baidu to acquire knowledge. From ChatGPT’s perspective, simply acquiring knowledge belongs to the “search” skill. The reason why AI large models have gained widespread acclaim in an instant is that they can not only search but also identify, interpret, summarize, and regenerate answers from similar databases. If we want to obtain a definition of a term, a quick Baidu search is sufficient; however, if we want to write a popular science article, Baidu alone is not enough. Knowledge includes scientific knowledge, social knowledge, event knowledge, numerical knowledge, and prototype knowledge, among others, in which large models perform well.
Thus, from the perspective of corporate employees, the first step in the workflow is often knowledge acquisition, making the use of large models an essential skill.
2. Creating Works
In the field of AIGC, text generation, image generation, and video generation are the most mature applications in businesses.
AI Writing: Most positions in a company cannot avoid document writing, which includes meeting minutes, promotional articles, hosting scripts, business contracts, notification documents, etc. Although there are usually specialized talents in copywriting within the company, talent is scarce, and with AI large models, everyone can collaborate to complete writing tasks.
AI Imaging: Besides the need for various images in corporate promotional roles, image creation is also quite common, such as designing a cover for a proposal, a certain type of display image in a PPT, or a meeting invitation.
AI Video Production: Short videos have become a common type of document in companies today, such as meeting promotions, product launches, and team-building videos. There are many generative video sites available, such as Tencent Zhiying, Wenxin Yige, etc.
3. Assisting Operations
AI can assist us in completing various operational tasks, such as drafting business plans, marketing plans, and collaborative processes. There are two necessary conditions:
First, the user must possess relevant operational knowledge. For example, to write a product marketing plan, you must accurately describe the product’s characteristics and customer types; otherwise, AI can only generate a vague plan that is nearly useless.
Second, the user must be skilled at interacting with AI, meaning they must use various prompts effectively. With these two points in place, using AI to assist operations is like adding wings to a tiger.
AIGC Empowers Departments
A business is primarily an organization that emphasizes economic value externally and efficiency improvement internally. AIGC can enhance internal management effectiveness in four main ways:
First is pure skill replacement; second is partial replacement plus human creativity; third is the integration of new skills and human skills; fourth is freeing up human energy with technological assistance to create new value.
So, what changes will occur in major departments of the business?
1. Marketing/Brand Department
The marketing/brand department is generally standard in enterprises, perhaps with different names, but similar functional responsibilities, differing only in scale. The marketing/brand department is responsible for product communication, brand building, market traffic, customer retention, etc. The high-frequency tasks can be categorized into three main types:
First is textual creation, including product introductions, corporate brochures, press releases, marketing articles, etc., where AIGC can serve as both an auxiliary tool and a primary player in certain structured texts. Especially for routine window operations, such as corporate websites, WeChat public accounts, and short videos, one can mimic business consultant Liu Run’s approach of using digital personas to disseminate through scripted digital avatars.
Second is image and video production, including corporate promotional videos, marketing posters, and product dynamic displays. AIGC can automatically generate videos or assist in synthesis, rapidly improving overall departmental efficiency and work quality.
Third is market activities, such as the high-frequency marketing events held annually, where the venue layout can utilize 3D virtual methods to enhance rendering effects.
2. Marketing Department
Marketing is the core work of a business; many CEOs or founders come from marketing backgrounds because this job is closer to customers and understands the market better. The marketing business department generally has three types of responsibilities:
First is product display. Compared to offline shopping, online shopping has the drawback of making it difficult to see the full picture of the product. Nowadays, AIGC technology can generate 3D models through visual algorithms, providing multi-dimensional visual experiences, lowering costs, and facilitating user transactions and conversions. Alibaba’s “Every Square Meter Business” has a relatively mature application in this area, using AIGC technology to simulate the online display of “products in the home,” allowing customers to see the actual effect directly.
Second is customer marketing. On one hand, attracting new customers involves basic methods such as building visits, telemarketing, community interaction, and targeted advertising. In recent years, with increasing information transparency, traditional marketing has struggled to attract customers, leading to the emergence of various methods such as content marketing, professional marketing, and word-of-mouth marketing. These methods typically require establishing “customer profiles” to analyze attributes, preferences, tastes, and purchasing power to formulate marketing plans that resonate with customers, an area where AIGC excels. On the other hand, maintaining old customers involves relatively fixed methods, such as customer birthdays and second purchase timings, which can be automatically generated and executed by AIGC.
Third is live-streaming sales. In the marketing process, live streaming has become an essential method in e-commerce. Faced with rising costs and the inability of hosts to work long hours, virtual hosts can live-stream 24/7 with strong controllability. For example, the digital persona “Liu Sanjie” from Guangxi, which was recently shortlisted for the IDC Best Case in July, acts as a promotional ambassador, not only familiar with scenic spots but also serving as a host and performing songs and dances in various activities, interacting with users effectively.
3. Financial Investment Department
The financial investment department is inherently linked to data processing and analysis, needing to search for various information, correct, analyze, and make decisions in complex markets, ultimately creating value. The responsibilities of the financial investment department generally fall into two categories:
The first is helping businesses with investments, mergers, and acquisitions. This type of work is best suited for AIGC, leveraging large models for information searching, data analysis, and decision support. Currently, quantitative trading and personalized robot investment advisors are very mature in the market. For example, Morgan Stanley is a seed client of GPT-4, managing and searching a vast wealth knowledge base, including hundreds of thousands of pages of investment strategies, market research comments, and analyst insights.
The second responsibility is promoting corporate financial products. AIGC can automate the production of video content for financial consulting and product introductions, and can also shape digital customer service personnel, like CITIC’s “Xiao Xin.”
4. Administrative Human Resources Department
Classifying the administrative human resources department encompasses many similar functional departments, such as the chairman’s office, corporate management department, and finance department. The responsibilities of these functional departments generally fall into three categories:
The first category is institutional documents, summarizing key success factors within the enterprise and formulating corresponding mechanisms to better replicate, innovate, disseminate, and prevent risks. In drafting institutional plans, AIGC can be directly used as a document assistant — you present your requirements, and it generates version 1.0, which can then be revised according to personalized needs.
The second category is data analysis. Whether it’s employee data, financial data, or corporate performance data, targeted collection, statistics, classification, and analysis are needed to ultimately form decision-making suggestions. The capabilities of AIGC in this area are self-evident.
The third category is customer service. For instance, handling employee onboarding/offboarding, reimbursement process inquiries, and process guidance can all be resolved through service robots.
5. Legal Regulatory Department
The legal department is a relatively independent entity, generally having three core functions:
First is drafting various legal documents, such as labor contracts, business contracts, and various correspondence in legal proceedings. Text generation is a strong suit of AIGC, and it is likely to handle ordinary document drafting competently.
Second is answering legal questions. When internal conflicts arise or disputes occur with clients, business departments will seek support from the legal department. Legal robots can answer common questions regarding legal terms and similar cases, providing certain handling suggestions.
Third is the lawyer function. This part requires a relatively high level of comprehensive ability, not only understanding professional content but also making judgments about people. The latest application is the “AI lawyer” developed by DoNotPay, an app that uses AI technology to provide legal services to ordinary people, initially automating appeals for parking violations. With advancements in AI technology, it can now act as a consumer agent, negotiating issues like network packages in dialogue boxes. GPT-4 has also passed a simulated bar exam, scoring in the top 10% of test-takers.
AIGC Empowers Senior Managers
1. Decision Mechanism
AIGC possesses excellent information integration and analysis capabilities, enabling the rapid processing and understanding of large amounts of data and text information from multiple sources. By interacting with AIGC, comprehensive data analysis and background information can be obtained, understanding the decision-making context and potential impacts. It also plays a role in problem identification and resolution, helping us recognize issues, provide solutions, and predict potential problems and risks based on historical data. As part of a decision support system, AIGC provides real-time information and analysis to decision-makers, reducing subjective bias and uncertainty by simulating and predicting the possible outcomes of different decision options. Here are a few examples for illustration.
Market Trend Analysis: Through AIGC, business managers can gain insights into market trends, competitor dynamics, and consumer behavior. AIGC can integrate and analyze this information from multiple data sources, helping managers better understand market conditions and formulate corresponding market strategies and product positioning.
Sales and Demand Forecasting: AIGC analyzes historical sales data, market trends, and other relevant data to provide forecasts for future sales and demand. This assists business managers in making decisions regarding inventory management, production planning, and marketing strategies to meet market demand while avoiding overproduction or shortages.
Supply Chain Optimization: AIGC can integrate data within the supply chain, including supplier information, logistics data, and production plans, helping business managers optimize supply chain management. It can provide analysis on inventory levels, delivery times, and supply chain risks to assist managers in making better supply chain decisions, thereby improving supply chain efficiency and flexibility.
Risk Assessment and Management: AIGC helps business managers identify and assess potential risk factors, including market risks, competitive risks, and legal risks, assisting managers in formulating risk management plans and measures to mitigate the impact of risks on the business.
Employee Performance Evaluation: AIGC can analyze employee performance and related data, providing suggestions and opinions regarding employee performance and training needs, helping managers make more accurate performance evaluations and decisions.
2. Talent Deployment
The impact of AIGC on talent deployment within businesses is significant, with the most direct effect being the reclassification and segmentation of business talent. With the development of AIGC, business talent can be categorized into two main types: natural persons (carbon-based individuals) and reconstructed individuals (silicon-based individuals).
Natural persons can be further divided into three categories: full-time digital talent, partners, and relational individuals.
Full-time digital talent includes digital management, professional, and application talents. Digital management talent corresponds to the high-level, middle-level, and grassroots managers originally set within the enterprise. Digital professional talent corresponds to technical talents such as big data engineers and algorithm engineers. Digital application talent corresponds to talents involved in operations, marketing, production, and other aspects within the enterprise.
Partners refer to cooperative partners and professionals on shared platforms who can provide unique technical expertise, project management capabilities, etc., offering professional services to businesses. For instance, companies can hire independent designers, consultants, and marketing experts through online platforms.
Relational individuals refer to external users interacting within communities who accumulate experience and knowledge while using the company’s products or services. Companies can gain feedback and insights by interacting with such talent for product improvements and innovations. For example, businesses can interact with users through social media platforms to gather user needs and opinions. Users can also directly create value, such as UGC (User Generated Content) contributors posting videos on Douyin.
Reconstructed individuals are categorized into robots and digital personas. Robots can be divided into hardware robots and software robots (RPA). Hardware robots include industrial robots, service robots, and special robots, applied in manufacturing, service industries, and specific fields. Software robots (RPA) are process robots that handle repetitive, rule-based, and tedious tasks, and their work scope can be expanded by incorporating AI.
Digital personas are categorized into human-driven and computationally driven. Human-driven digital personas can replicate the appearance and voice of real people through mapping and interaction. Computationally driven digital personas are entirely virtual, achieving voice and expression effects through computation.
In summary, AIGC’s impact on talent deployment is to reclassify and refine talent. This classification and refinement help businesses better utilize the expertise and capabilities of various talents, enhancing work efficiency and innovation capabilities.
3. Operational Models
The impact of AIGC on business operational models primarily manifests in three aspects: intelligence, automation, and integration. Through intelligent interactive methods, it provides a better user experience; through automation technologies, it improves work efficiency and reduces costs; and through integrated consolidation, it promotes internal collaboration and cooperation within the enterprise. For instance, how does AIGC operate in the production of a car?
Intelligent Aspect: AIGC collaborates with the car design team to provide intelligent design suggestions. It can analyze market data, user feedback, and the latest technological trends to help the design team optimize the car’s appearance and performance to meet consumer demands and enhance competitiveness.
Automation Aspect: AIGC interacts with automated devices on the production line to achieve automated control of production processes. It can interact in real-time with robots and automated equipment to monitor and adjust production progress. For example, AIGC can communicate with welding robots to guide them in completing body welding tasks and ensure that assembly robots accurately install components.
Integration Aspect: As an internal communication tool for businesses, AIGC promotes collaboration between different departments. For example, design teams, engineers, production planners, and supply chain teams can share information, solve problems, and coordinate work with AIGC. This helps ensure smooth production processes, reducing communication errors and information delays.
Additionally, AIGC can collect and analyze customer feedback and demands. Businesses can use AIGC to understand customer preferences, expectations, and opinions regarding cars, thereby optimizing product design and providing better after-sales service.
4. Organizational Models
After digitization, the ideal organizational form is “operating body + empowering middle and back office,” colloquially known as platform-based organizations. The Daren Think Tank proposes a more systematic “cell-based team”. Before the emergence of AIGC, enterprises adopted a hierarchical structure due to the irreplaceability of certain responsibilities, but AI applications have accelerated organizational platformization, as intelligence makes collaboration easier, allowing organizations to become flatter. This is specifically reflected in several aspects.
Empowering the Frontline: AIGC serves as an intelligent assistant, providing support for frontline departments (such as customer service teams). AIGC can help frontline employees quickly access product information, answer common questions, and provide personalized suggestions. This can enhance customer satisfaction, strengthen customer relationships, and promote business growth.
Building the Middle Office: AIGC can play a role in building the middle office. Middle office teams utilize AIGC’s intelligent capabilities to integrate and manage core data, business logic, and service capabilities. Middle office teams can quickly access data and analysis results, providing consistent decision support and service capabilities. This improves data accessibility and consistency, promoting collaboration and sharing between different departments, enhancing overall efficiency and flexibility.
Enhancing the Back Office: AIGC supports back office departments (such as R&D, technical support, etc.), enhancing professional depth. Back office departments can access relevant knowledge, technical guidance, and problem-solving solutions. AIGC provides intelligent training and learning, helping back office departments update knowledge and improve professional levels.
It is important to note that introducing AIGC does not automatically mean that organizational structures will change. Enterprises need to comprehensively consider their business needs and cultural characteristics when applying AIGC, adjusting organizational structures and processes to better leverage AIGC’s advantages. This may involve changes in training employees, workflows, and employee roles and responsibilities.
5. Knowledge Management
Knowledge management is a systematic approach and process aimed at effectively acquiring, organizing, storing, disseminating, and applying knowledge assets within and outside the organization. It involves integrating and utilizing the organization’s internal expertise, experience, information, and external knowledge resources to support organizational goals and strategies. With the support of AIGC, it can promote the creation, sharing, transmission, and application of knowledge, further enhancing organizational performance.
Knowledge Discovery: AIGC can serve as an intelligent search engine, helping users discover relevant knowledge and information. Users can pose questions or topics, and AIGC provides relevant knowledge and information based on the context and semantic understanding of the questions, helping users find the knowledge they need more quickly.
Knowledge Acquisition: AIGC quickly acquires and integrates a large amount of knowledge and information. It can obtain data from multiple sources, including texts, papers, web pages, etc., integrating this information into useful knowledge, thereby accelerating the knowledge acquisition and integration process and improving knowledge management efficiency.
Knowledge Indexing: AIGC organizes and classifies knowledge to make it more accessible and understandable. Users provide questions or topics, and AIGC offers relevant knowledge and information, enhancing the usability and accessibility of knowledge management.
Knowledge Assessment: AIGC provides intelligent assessment and feedback, helping users evaluate and verify the accuracy and reliability of acquired knowledge. Users can pose questions or validate hypotheses, and AIGC provides corresponding assessments and feedback based on its training data and algorithms. This helps users better evaluate and utilize the knowledge they acquire.
Knowledge Distribution: AIGC acts as an intermediary to promote knowledge sharing and collaboration. Users share knowledge and experiences with others and obtain knowledge and experiences from others. AIGC helps users better allocate and share knowledge, enhancing team and organizational collaborative efficiency and innovation capabilities.
A Dialectical Analysis of AIGC’s Impact on Enterprises
1. Limitations
While the application of AIGC in enterprises has many potential benefits, it also has certain limitations, mainly focusing on data privacy, knowledge graphs, and talent bottlenecks.
Data Privacy: AIGC requires large amounts of data for training, which may involve sensitive internal data. When using AIGC, businesses need to ensure data privacy and security, avoiding the leakage and misuse of sensitive information. This requires establishing appropriate data protection and privacy policies and implementing corresponding security measures, such as data encryption and access control.
Knowledge Graph: The knowledge of AIGC is learned from pre-training data and cannot directly obtain real-time internal enterprise or industry-specific knowledge. For businesses, it may be necessary to build their own knowledge graphs to integrate and manage internal expertise and experience. This requires additional work and investment, including knowledge collection, organization, and tagging, as well as the technology and platforms for building knowledge graphs.
Talent Bottleneck: Using AIGC in enterprises requires corresponding technical and professional knowledge. However, the technology and resources needed to train and deploy AIGC may exceed the capabilities of ordinary enterprises. Moreover, businesses may need specialized teams or personnel to manage and maintain AIGC, including data scientists, natural language processing experts, etc. This may lead to a talent bottleneck, limiting the development and application of AIGC in enterprises.
2. Differences
National conditions, industrial differences, and the personalized situations of enterprises will all affect the application of AIGC. Enterprises need to comply with relevant regulations and customize development and management based on their scale and resources, in conjunction with industry characteristics.
When using AIGC, businesses must ensure data compliance. For example, financial enterprises need to pay special attention to protecting customer privacy to comply with relevant personal information protection regulations.
In terms of industrial differentiation, companies in different industries may have different application scenarios and needs. For example, e-commerce companies can utilize AIGC to provide personalized recommendations and customer support services. Businesses can recommend related products and services based on user needs and preferences, enhancing the shopping experience. In the manufacturing sector, AIGC applications may focus more on technical support and troubleshooting, helping businesses improve production efficiency and quality control.
Furthermore, the personalized situations of enterprises will also affect the application of AIGC. Large enterprises may have more resources and technical capabilities to develop and deploy AIGC. For instance, a large bank can leverage AIGC to build an intelligent customer service system that provides 24/7 online support. Meanwhile, small and medium-sized enterprises may face challenges related to talent shortages and technical limitations, needing to rely on third-party AIGC solutions or share AIGC with partners.
3. Developmental Aspects
From a developmental perspective, the impact of AIGC on enterprises can be discussed in four aspects.
First, further technological innovation. Businesses can utilize AIGC for technological innovation while protecting data privacy and ensuring compliance with data sharing. On one hand, privacy protection technologies such as data encryption, differential privacy, and secure computing can be adopted to safeguard user data privacy. On the other hand, exploring federated learning and secure multi-party computation technologies can promote broader innovation and collaboration while protecting data privacy.
Second, customized AI development for enterprises. Businesses can conduct customized AI development according to their specific needs and business scenarios to better adapt to their unique demands. Utilizing some foundational models, specific domain training and fine-tuning can enhance the accuracy and applicability of AIGC. Additionally, combining other technologies and tools, such as automated processes, data analysis, and machine learning algorithms, can create tailored comprehensive solutions based on large models to meet specific business needs.
Third, effective talent cultivation. To better apply AIGC, enterprises need to focus on talent development, including training internal data scientists, natural language processing experts, and AI developers to enhance their capabilities in large model development and application. Moreover, businesses can collaborate with universities, research institutions, and professional teams to conduct talent training and knowledge sharing, facilitating talent exchange and cooperation to meet the demand for AI technology and applications.
Fourth, the development of AIGC in the Chinese context. China’s AI large models have rapidly proliferated, ranking first internationally in parameters and leading in certain fields, especially in the application of Chinese language models, which far exceed foreign AI models. With user growth and the maturity of data training, these advantages will become even more pronounced.
The arrival of AIGC presents unprecedented opportunities and challenges for employees, departments, and decision-makers in enterprises. Its existence is not a threat but a chance to inspire us to surpass ourselves. Let us embrace the changes of this digital age with brave hearts.■
The author is the general manager of Daren Think Tank and the secretary-general of the Yangtze River Delta Transparent Business Ecosystem Industry Alliance.
