Why AI Isn't the Death of Jobs



Why AI Isn't the Death of Jobs
A probable outcome for companies that use AI only to increase efficiency. Companies using AI to drive innovation are likely to increase head count than reduce it.
Technology Briefing

Transcript


When pundits talk about the impact that artificial intelligence (or AI) will have on the labor market, the outlook is usually bleak, with the loss of many jobs to machines as the dominant theme. But that's just part of the story - a probable outcome for companies that use AI only to increase efficiency. As it turns out, companies using AI to also drive innovation are more likely to increase head count than reduce it.

That's what Jacques Bughin and his colleagues at McKinsey & Company recently learned through the McKinsey Global Institute's broad-based research initiative aimed at understanding the spread of AI in economies, industry sectors, and companies. The researchers polled 20,000 AI-aware C-level executives in ten countries to compile a sample of more than 3,000, mostly large, companies, identified distinct clusters within that pool, and ran a variety of scenarios on those clusters to project the effects of AI on employment, revenue, and profitability.

This research and analysis suggests that although AI will probably lead to less overall full-time-equivalent employment in those ten countries by 2030, it won't necessarily lead to massive unemployment. One major reason for this prediction is because early innovation-focused adopters are positioning themselves for growth, which tends to stimulate employment. Growth is a key part of the equation linking the performance and employment shifts we'll see from now through 2030 for five types of companies.

Here's how the McKinsey & Company experts expect things to play out in the five clusters of companies the authors examined.

Type #1: Enthusiastic Innovators are pioneering companies that make early investments in AI and embrace the disruption it can create in the quest for advantage, adopt a full range of AI technologies and use them to bolster innovation and efficiency.

These companies are analogous to what sociologist and communication theorist Everett Rogers called "early adopters" back when he coined the term - they're intrinsically motivated to use new technology to shape and open markets. While this approach is potentially complex in the short term, the McKinsey & Company analysis shows that by 2030, the profitability of enthusiastic innovators will grow 8 percent faster than that of the average company on an annual basis, their revenue will grow 4 percent faster, and their head count will rise 2.2 percent faster.

Currently, only about 2 percent of the sample companies qualify as enthusiastic innovators, but by 2030, the McKinsey & Company experts expect that figure to grow to 12 percent and account for 20 percent of total revenue of companies across all the clusters. Companies in this cluster include many digital natives: Google Inc., for example, is using AI to drive innovations in search and to pursue efficiency by reducing the energy consumption of its servers. They also include a lesser number of conventional companies: Chinese insurer Ping An Insurance Co. of China Ltd., for instance, has launched a variety of CEO-sponsored AI initiatives aimed at topline growth and has hired more than 600 data scientists to support these ventures.

Type #2: called Careful Innovators are somewhat slower to invest in AI and spread adoption than are Enthusiastic Innovators. They balance the risks of jumping into the new technology too quickly against the competitive threats they may face from more aggressive early adopters. Unlike enthusiastic innovators, these companies tend to focus their transformational initiatives more narrowly, mostly within their industry of origin - either because they are locked into legacy systems or because they see less opportunity elsewhere. The McKinsey & Company analysis suggests that, by 2030, the profitability of careful innovators will grow 3 percent faster per year than at the average company, and head count will rise almost 1 percent faster.

By 2030, careful innovators will account for about 12 percent of all companies and 14 percent of the overall company revenue. These companies tend to be the incumbents in business-to-business (or B2B) or less digitally mature sectors, rather than digital natives. Volvo Cars, for example, recently pegged 4 percent to 5 percent of its annual revenue to deploying new electric car innovations, many of which will be enabled by the focused application of AI.

Type #3: Efficiency leaders are using AI intensively. Like enthusiastic innovators, they are early adopters; however, their primary focus is profitability - they use AI to boost efficiency and replace labor. The McKinsey & Company analysis shows that, by 2030, the head count of efficiency leaders will fall roughly 3 percent faster per year than average, and their profitability will grow nearly 5 percent faster. But their revenue will grow only about 1 percent faster. The small amount of topline growth that efficiency leaders will capture stems from the market share gained by passing some of their cost savings on to customers.

By 2030, efficiency leaders will account for approximately 8 percent of companies and about 9 percent of overall revenue. In general, these are digitally savvy companies in industries such as banking, insurance, and manufacturing that are seeking to reduce the costs associated with manual processes. For example, in 2010, after Gastonia, North Carolina-based Parkdale Mills Inc., the largest buyer of raw cotton in the U.S., retooled its long-shuttered South Carolina plants with smart robotics, it was able to reduce its staff by more than 90 percent.

Type #4: Efficiency followers focus their AI efforts on efficiency but adopt AI very slowly. Consequently, the profitability of efficiency followers will grow only 0.2 percent faster than average annually, and their head count will fall 0.5 percent faster.

Efficiency followers represent the second largest cluster of companies in our sample: 18 percent of companies, accounting for 19 percent of the overall revenue. These companies are found in every sector, but they are less prevalent in high-tech and other AI-advanced sectors, such as media and financial services. They tend to use digital technologies for process optimization, but usually have not used them for business innovation.

Type #5: AI resistors are companies that will either not invest in AI at all or do so on a very limited scale (in one function, for example, or with a narrow set of technologies). Their revenue will shrink 2.7 percent faster annually than the average; their revenue share will go from 50 percent of all companies to 39 percent by 2030. Moreover, their cumulative profitability will fall 12 percent faster than average by 2030. The head count of AI resistors will fall 2.2 percent faster than average annually - a result of their limited employment prospects.

AI resistors are, by far, the largest cluster of companies, representing half of McKinsey's sample. These companies may be daunted by the complexities and costs of AI, or they may be locked into legacy strategies or systems. Although this is the only cluster that will grow its labor output compared with the average of all companies, AI resistors may actually have the least favorable employment prospects of the five clusters. Because of the profitability pressures they will experience, they will likely have to cut costs - particularly, head count - over the long term. So, their outlook for jobs may be even more troubled than McKinsey's analysis indicates.

As the McKinsey & Company experts compare the five clusters in terms of revenue, profitability, and employment, it's important to consider a couple of things.

First, in the average scenario, the overall effect of AI between now and 2030 is significantly less substantial than you might expect. For instance, the impact on the labor/output ratio is about a 1 percent drop each year. This is not much different from the trend in labor productivity reported by the Organization for Economic Co-operation and Development (or OECD) from 2001 to 2010.

Second, employment macro-dynamics will depend on AI activity within sectors and economies. More companies using AI to innovate will bolster overall head count. More AI resistors and companies pursuing AI in the sole pursuit of efficiency will reduce the number of employees in an industry or economy.

So, it's not an inevitable conclusion that AI will ratchet up unemployment, as many have suggested - at least between now and 2030. The outlook is more nuanced than that. Job losses will arise as the result of automation, as the labor-output ratio evolution suggests. But what often gets overlooked is that job losses are also a risk of companies' inability or unwillingness to use AI for innovative purposes, which leads to lower revenue and profit - and a lower absolute need for labor.

As is so often the case, the future is malleable. We all forge tomorrow's path with our actions today.

Comments

No comments have been submitted to date.

Submit A Comment


Comments are reviewed prior to posting. You must include your full name to have your comments posted. We will not post your email address.

Your Name


Your Company
Your E-mail


Your Country
Your Comments