How Companies in China Can Fill the AI Talent Gap

Source: McKinsey

Authors: Wouter Maes and Alex Sawaya

Image source: Generated by Unbounded AI, style model (Song Yun Architecture)

Attracting and retaining artificial intelligence (AI) talent has become a global challenge, and China is no exception.

In McKinsey’s 2022 survey of global artificial intelligence business executives, 75% of Chinese respondents admitted that they encountered difficulties in recruiting data scientists [1]. More than half of the respondents said that it is difficult to find suitable talents to fill the vacancies of key AI-related positions, such as data engineers, data architects and machine learning engineers, etc., and these positions are crucial to the design, construction and promotion of advanced digital and AI capabilities indispensable.

Our latest research shows that even with the recent contraction in the market, talent acquisition will continue to be increasingly difficult. It is estimated that by 2030, the potential value brought by AI to China is expected to exceed US$1 trillion. As major companies compete to tap this value, China's demand for highly skilled talents will reach the current level (increased from 1 million to 6 million Human) 6 times【2】. (See sidebar "About this study" for our methodology.)

About this study

The survey and interviews covered 102 leading companies in China that have adopted artificial intelligence in at least one field. We also analyzed global and local reports, use cases, and recruitment databases to explore the demand for AI talent in China, the challenges and actions companies are taking to fill the talent gap. To assess talent needs, we considered the economic impact of AI on key industries (consumer, finance, manufacturing, business services, automotive, transportation and logistics, and healthcare and life sciences) and constructed per capita productivity for each industry. mold. On the supply side, we assess the number of domestic and foreign university graduates who tend to choose domestic employment, including science, technology, engineering and mathematics (STEM) majors, and the number of top scientific and technological talents available.

It is estimated that by 2030, domestic and foreign universities and existing top talent pools can only provide about 2 million (that is, one-third of the required) AI talents, and the gap will reach 4 million (see Figure 1). After 2030, as the birth rate declines, the number of college students will decrease, and the AI talent gap will become more severe.

Facing the huge gap that is about to emerge, how should enterprises ensure the talents and capabilities needed to compete? We conducted surveys and interviews with more than 100 leading companies in China, revealing two key insights:

Talent gaps vary. While every business will need to upskill existing workforces and move away from traditional recruiting methods to acquire the talent and capabilities it needs, investments and interventions will vary from company to company based on their level of digital maturity.

Local and multinational companies have their own advantages. Although Chinese college graduates are more optimistic about local companies and their innovation and performance-based incentive structures, multinational companies in China can effectively use their global networks to attract talents from a larger talent pool.

Aiming at the challenge of AI talents, this article discusses in depth the types of talents that enterprises should give priority to in each stage of digital maturity, and how to better acquire the required skills and capabilities.

Talent and skill needs vary with digital maturity

As important driving forces, digitalization and artificial intelligence are creating huge value for China, which requires a complete set of advanced skill bases. These skills come from roughly seven areas: customer experience, cloud, automation, platforms and products, data management, DevOps (an approach to optimizing software development), and cybersecurity and privacy. Although companies ultimately need to build talent pools in various fields, our research shows that companies should prioritize the talents they really need based on their digital maturity. The three common digital maturity levels are traditional, hybrid and digital (see Figure 2).

Traditional

Traditional refers to companies that have just started digital transformation. Such enterprises usually only have small-scale internal teams and face greater competitive pressure. They urgently need to start digital and artificial intelligence transformation. Their transformation focuses mainly on establishing data foundations, optimizing business processes, and focusing on specialized use cases that can rapidly improve business effectiveness (rather than building future innovative R&D AI capabilities). To this end, these companies should focus on two types of talent:

The first type of talent is a data management expert who is proficient in data architecture, data engineering, data analysis, and analysis translation. They can build data platforms, pipelines and processes, drive data openness, generate data-driven real-time insights, ensure data quality and governance, and manage the lifecycle of use cases. Enterprises can hire such experts to serve data product or use case teams to drive the delivery of new digital and AI capabilities. A data center of excellence also requires such experts to collaboratively design and oversee data management processes, ensuring proper access controls, data quality, and approval and retention policies.

An agricultural business built a centralized enterprise data center to support data management protocols and governance processes, providing access to thousands of employees in different departments to advance artificial intelligence and analytics use cases. Enterprises no longer need to repeatedly develop new data pipelines, thereby significantly reducing IT costs and modernizing business methods. For example, robots are used to track the breeding conditions of animals and automatically send out alerts when potential diseases and other problems are detected.

The second type of talent is platform and product experts, proficient in software development. They are able to customize "software as a service" (SaaS) or other external solutions to improve business efficiency and provide new customer-facing services.

For example, a consumer electronics manufacturer invested in a data platform development team after applying an AI use case to optimize production planning and labor productivity. The team will update the underlying model, user interface, data pipeline, and back-end infrastructure, continue to enhance current AI use cases, and introduce additional use cases.

Hybrid

Hybrids refer to established players in industries that have invested heavily in digital transformation. Such companies already have strong internal technical strength and a solid foundation, and are now focusing on simplifying the development process, accelerating the delivery of new digital and AI products, and expanding domain expertise to provide an excellent customer experience. Hybrid enterprises need DevOps experts who specialize in software development, such as agile product management, continuous integration/continuous delivery (CI/CD) practices, and microservices for faster deployment. Customer experience experts are also what they need. These experts are proficient in various predictive analysis, design thinking and automated testing capabilities, and have strong prototyping capabilities to create new experiences for customers.

Of course, IT efficiency and server spending will also be a challenge as hybrid enterprises continue to expand capabilities and host more AI models and applications in the cloud. Our cloud survey in 2022 found that more than 75% of enterprises in China plan to use multiple cloud services, and 90% plan to use a mix of public and private cloud services by 2025 [3]. To clarify capability requirements and how different cloud services will operate, enterprises need cloud experts with experience in Kubernetes, Docker, and multi-cloud architectures.

Number

Digital refers to digital native businesses such as tech giants, artificial intelligence and tech start-ups. Such enterprises already have sufficient talent reserves in most digital and AI fields, but they still need to further expand their reserves to meet changing industry expectations and technological advancement needs.

These companies focus on cybersecurity and data privacy. In China, due to the increased security and privacy protection of enterprises, which may have an impact on AI and digitalization, digital enterprises need experts with a global perspective and a systematic approach to solve problems, giving priority to security testing in the early stages of product development (often referred to as shift-left security), zero-trust security frameworks, and data protection laws and practices. Another category of talent that should be prioritized are automation experts with skills in generative artificial intelligence, robotic process technology, machine learning, AI-enabled analytics, and quantum computing. They drive end-to-end automated development, testing, and deployment to improve the efficiency and speed of bringing new features to market.

Various multinational companies

Regardless of the level of digital maturity, multinational companies operating in China must ensure that their AI talent is equipped to work smoothly across their global networks. For example, the team needs to be fluent in Chinese and foreign languages, understand the working mode of other regions, and be able to communicate smoothly with global colleagues. Leadership needs to be good at building partnerships and ensuring that everything works in line with the company's global IT and AI standards, while effectively meeting local business needs. Product owners need to understand which data and designs in different regions can be reused and scaled, and which data and designs need to be rebuilt locally to meet the needs of China's digital ecosystem.

For example, the European branch of a multinational company developed a global transportation app that uses consumer traffic data from Google, Facebook and Instagram to optimize routes. Although most of the branches of this branch around the world can use this application, in order to obtain data from the domestic platform, the product leader in China needs to lead the team to adjust the application first and then deploy it.

Fill gaps by upskilling and expanding talent sources

Through interviews on topics related to talent selection and retention, we found that traditional and hybrid companies have a lot of work to do in each talent management stage (see Figure 3). Digital businesses only need to strengthen in a few areas to maintain their talent management advantage.

After in-depth research on corporate strategies, we found that when all enterprises promote the development of digital and AI talents, there are two most critical points: 1. Improve the skills of existing talents; 2. Diversify and expand talent sources. Our research shows that different types of businesses need to take different actions on these two fronts.

Upgrade the skills of existing employees

Upskilling employees is a common strategy for companies to acquire the talent they need. Our research suggests that companies in China can build the required skills through targeted capacity building of their existing bench of business and AI talent (see Figure 4).

Of course, our interviews also showed that the best skills to improve are key skills that are hard to find, outsource, or obtain, such as an understanding of legacy applications or existing product functions (for details on how to start an employee skills improvement plan, see Figure 5).

Traditional

Analytical translation is a skill that traditional organizations should focus on. Our research shows that without these skills, business units will struggle to convince new digital and AI initiatives. Improve the skills of business experts in different fields to identify and evaluate potential digital and AI use cases, assess potential business value and support later deployment, allowing traditional enterprises to gain value from digital and AI investments faster. This type of training is best delivered in-house as an “analytics academy,” where companies can customize training and offer apprenticeships so experts can apply what they’ve learned.

For example, in order to improve the skills of employees, an advanced manufacturer established an analytics academy at the beginning of the transformation, helping more than 200 employees to transform into analytics translators.

The courses include: 1. Weekly half-day lectures (for 2~3 months), including problem solving, talent and use case requirements; 2. Best practices in agile delivery and change management; 3. For use cases in the company's roadmap, Conduct on-the-job training.

Since taking up the post, these translation talents have promoted the implementation of more than 50 new digital and AI use cases.

Hybrid

At present, only 8% of domestic AI talents have advanced AI-related skills, such as edge computing, big data and machine learning, and cognitive artificial intelligence [4]. For hybrid businesses, upskilling existing employees is a key part of transformation. But such businesses need to increase investment in online courses and certification programs. In McKinsey’s 2022 Global Artificial Intelligence Survey, only about one-third of Chinese companies surveyed used such programs (31% of companies used their own online courses, and 29% used certification programs) [5].

A leading financial institution provides a customized learning journey based on the employee's position and career path, focusing on online learning. Every employee can use a mobile learning app to take the key skills-building courses needed for their role. The app offers a wide range of courses, including Python programming, multi-cloud architecture deployment, leadership skills required for digital transformation, and more.

Number

The biggest challenge for digital businesses will be keeping pace with the rapid development of emerging technologies such as generative artificial intelligence and quantum computing. Such enterprises can encourage employees to actively keep up with the latest technological developments (such as arranging employees to attend academic conferences, participate in relevant research, apply for patents, participate in hackathon competitions, etc.), and help them narrow the gap with new talents.

A tech company gave employees the time, space and budget to research and develop new capabilities using emerging technologies outside of existing projects, which brought artificial intelligence, blockchain and cloud computing, and new products to the company Numerous patents and patent applications in innovative fields.

Diversified sources of talent development

Job outsourcing and acquisition of basic technical capabilities (and corresponding talent) are also ways for companies in China to fill talent gaps. Multinational corporations have an obvious advantage in this regard because of their global influence. They can leverage existing solutions developed by colleagues in other regions, or new capabilities developed in countries such as Vietnam and India. Of course, businesses need to consider various financial and regulatory issues, such as ensuring compliance with all data protection regulations in China. Our research shows that different types of businesses have different best practices.

Traditional

Traditional players must act quickly to catch up with AI and digital leaders to remain competitive. Initiating a digital transformation by recruiting and training new people, especially in a tight labor market, can take a lot of time. One way to quickly acquire AI talent and capabilities is to partner with vertical IT and SaaS providers. Some business leaders first advance through such partnerships while seeking out new talent. For example, the consumer electronics manufacturer mentioned above outsourced the development of new AI-optimized models while building out its talent strategy. In this way, the company put new capabilities into production (and generated value) within 8 weeks, which may take several times longer if it relies entirely on training new people.

Others may work with external suppliers who build the overall infrastructure of their digital systems. For example, a Chinese industrial vehicle supplier hired a leading software company to integrate more than six business and factory systems, including enterprise resource planning, manufacturing execution, product lifecycle management, supplier management, human resources, and business intelligence. After the project took more than three years to complete, the company rolled out a range of use cases, including a collaborative product design system that improves R&D efficiency and speeds up new product launches.

When outsourcing work, ensure that all relevant data and technology strategies align with the company's strategic priorities upon which the vendor can base design decisions. In this way, companies can involve multiple vendors in different tasks and projects, and ensure that all solutions share data and insights seamlessly.

Hybrid

In the next phase of digital transformation, outsourcing can be extremely valuable for hybrid enterprises, increasing the reach and productivity of existing technical experts. Outsourcing can also reduce the burden on technical staff, so that they do not need to spend a lot of time maintaining legacy systems in the middle and back office of the upgrade.

Today, enterprise software solutions related to human resources, finance, communication, and business process automation have matured in China. Enterprises can quickly migrate these systems to the cloud and redeploy AI talents to high-value use case projects. In other cases, companies can use third-party resources to build parts of new digital or AI solutions for teams.

Number

Many digital-native companies are finding that frequent expansions and reorganizations lead to high tech brain attrition and high recruitment costs, threatening their continued growth. For digital companies, entering new markets or business areas through strategic acquisitions will be a better strategy (rather than building new capabilities internally) as the talent gap widens.

Take ByteDance as an example. Through this acquisition, it has obtained new virtual reality (VR) capabilities, its applications have been expanded, and it has also obtained a team of VR experts to continue to build new capabilities for it.

picture

Looking ahead, China's demand for AI talents will be in short supply. Leaders need to inspire creativity and ensure that the organization has the talent pool and capabilities to remain competitive for the next decade. Companies can prioritize upgrading the skills of existing talent and strategically fill talent gaps through outsourcing and acquisitions to build competitive advantage in major global markets.

Notes:

[1] "The state of AI in 2022—and a half decade in review", McKinsey, December 6, 2022. The survey covers 102 interviewed companies in China.

[2] Based on the following research: Shen Kai, Tong Xiaoxiao, Wu Ting, and Zhang Fangning, "Exploring the new frontier of artificial intelligence: China's economy welcomes another $600 billion opportunity", McKinsey, June 7, 2022; "Notes from the AI frontier : Applications and value of deep learning”, McKinsey Global Institute, April 17, 2018; National Bureau of Statistics of China, 2021.

[3] Kai Shen, Anand Swaminathan, Xiaoxiao Tong, and Wei Wang, "China in the Cloud, Looking to 2025", McKinsey, July 8, 2022.

[4] "2021 China IT Service Talent Supply Report", iSoftStone and iResearch, August 2021.

[5] “The state of AI in 2022,” December 6, 2022.

author:

Wouter Maes

McKinsey Global Managing Partner, Beijing Branch

Alex Sawaya

Global senior managing partner of McKinsey, resident in Hong Kong branch

The author thanks Tong Xiaoxiao and Wang Lingyi for their contributions to this article.

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Share
Comment
0/400
No comments
Trade Crypto Anywhere Anytime
qrCode
Scan to download Gate app
Community
English
  • 简体中文
  • English
  • Tiếng Việt
  • 繁體中文
  • Español
  • Русский
  • Français (Afrique)
  • Português (Portugal)
  • Bahasa Indonesia
  • 日本語
  • بالعربية
  • Українська
  • Português (Brasil)