As generative AI rapidly penetrates multiple areas, it is opening up entirely new opportunities to boost productivity. Recently, the global consulting giant McKinsey & Company released the artificial intelligence AIGC research report: "The State of AI 2023: The Year of the Explosion of Generative AI" report. Welcome to read~
It's still early days, but the use of a generation of AI is already widespread
Field survey results from mid-April 2023 revealed that while AI technologies are just becoming publicly available, experimentation with these tools is already common, with respondents expecting new capabilities to transform their industries. The new generation of AI has piqued the interest of those in the business world: individuals in different regions, industries, and qualifications are using the new generation of AI at work and outside of work. 79% of respondents said they had been exposed to at least some AI at work or outside of work, and 22% said they regularly use AI in their own work. While usage was similar among respondents of different seniorities, it was highest among respondents working in the technology industry and North America. (See Figure 1).
Figure 1 Percentage of respondents by region, industry, seniority, age and gender using generative AI tools
Figure 2 Percentage of respondents across business functions who regularly use generative AI tools
The survey shows that many organizations are not fully prepared for the widespread use of generative AI tools and the business risks they may bring. Only 21 percent of respondents said their organization has policies in place and governs the use of generative AI at work. When asked specifically about the possible risks posed by the use of generative AI, few respondents said their organizations were reducing a common AI risk, inaccuracy, which was mentioned more frequently than both cybersecurity and regulatory compliance, which were the most common risks to AI in previous surveys. Only 32 percent of respondents said their organization is reducing inaccuracy, which is lower than the 38 percent who are reducing cybersecurity risk. Interestingly, this number is significantly lower than the percentage of respondents (51%) who reported last year that reduced AI-related cybersecurity risks. Overall, consistent with the results of previous McKinsey surveys, the majority of respondents said their organizations were not addressing AI-related risks (see Figure 3).
Figure 3 Organizations' cognition and response to different types of AI risks
Second, leading companies are at the forefront of generative AI
According to the survey results, AI high-performing companies, that is, those that respondents say attribute at least 20% of their EBIT profits in 2022 to those that use AI, are fully engaged in AI. These companies that derive significant value from AI are already using generative AI tools in business functions more than others, especially in product and service development, risk, and supply chain management. Looking at all AI capabilities, including more traditional machine learning capabilities, robotic process automation, and chatbots, AI high-performing companies are more likely than others to use AI in product and service development, such as optimizing product development cycles, adding new features to existing products, and creating new AI-based products. These companies are also applying AI more to risk modeling and HR areas such as performance management, organizational design, and workforce deployment optimization than others.
Figure 4 The top target ratio of AI high-performing companies and other enterprises using generative AI tools
While AI high-performing companies also face challenges in extracting value from AI, the results show that the difficulties they face reflect their relative maturity in the field of AI, while others struggle with the fundamental, strategic elements of AI use. Respondents to AI high-performing companies believe that the top challenges for their organizations are models and tools, such as monitoring model performance in production and the need to retrain models over time. In contrast, respondents from other organizations mentioned more strategic issues, such as developing a well-defined AI vision associated with business value or finding adequate resources (see Figure 5).
Figure 5 Proportion of factors that AI high-performing companies and other enterprises face the greatest challenge in extracting value from Al
The findings provide further evidence that even high-performing AI companies are not mastering best practices for using AI, such as machine learning operations (MLOps) methods. For example, only 35 percent of respondents from AI high-performing companies said that their organizations would restructure existing components rather than create development when possible, but this is much higher than respondents from other companies (19 percent). To use some of the more transformative use cases offered by generative AI, many specialized MLOps techniques and practices may be required, and it is as safe as possible to do so.
The demand for AI-related talent is changing, and AI is expected to have a huge impact on the workforce
The latest survey results show that the job postings for companies have changed. Over the past year, companies using AI most often hired data engineers, machine learning engineers and Al data scientists, all common roles in McKinsey's survey last year. However, the percentage of companies hiring Al software engineers this year (28%) is much smaller than in last year's survey (39%). With the popularity of Al, there have been recent tips for engineering-related jobs, with 7% of respondents saying that companies have hired for these positions in the past year. The survey results show that hiring for Al-related positions remains a challenge, but has become less difficult to recruit over the past year, which may reflect a wave of layoffs in tech companies from late 2022 to the first half of 2023.
Figure 6 Percentage of organizations experiencing difficulties in recruiting AI-related positions
Figure 7 Expected impact of AI use on employee numbers in the next three years
In terms of the specific expected impact of using AI, the majority of respondents expect fewer employees in enterprise service operations. This finding is broadly consistent with recent McKinsey research: While the advent of Al has raised expectations for automating the percentage of work (from 50% to 60%-70%), this does not necessarily mean that all of the employee's work will be automated (see Figure 8).
Figure 8 The expected impact of AI use on the number of employees in different business functions in the next three years
AI high-performing companies are expected to reskill at a higher level than others. Respondents from these companies said that their organizations will reskill more than 30% of their workforce in the next three years as a result of the use of AI by high-performing companies (see Figure 9).
Figure 9 Likelihood of AI high-performing companies and other companies reskilling their employees in the next three years
Everyone is paying attention to generative AI, but the use and impact of AI will remain stable
While the use of generative AI tools is rapidly gaining popularity, survey data does not show that these new tools are driving the adoption of AI across organizations. For now, at least, the percentage of organizations using AI remains stable, with 55 percent of respondents saying their organizations are using AI. Less than one-third of respondents said their organizations have adopted AI in more than one business function, indicating that the scope of AI use remains limited. Consistent with previous survey results, product development and customer operations remained the two most frequently cited AI-enabled business functions by respondents. Overall, only 23 percent of respondents said their organizations attributed at least 5% of their EBIT last year to the use of AI, suggesting that there is more room for organizations to capture value (see Figure 10).
Figure 10 The proportion of business functions that use AI in an organization
Organizations are seeing a return in continuing to use AI's business functions and plan to increase their investments in the coming years. (See Figure 11).
Figure 11 Proportion of organizations using AI to reduce costs and increase revenue in 2022 by business function
Source: No. 18 Technology_https://mp.weixin.qq.com/s/7n9VGW2UybWdweYECR9pfg