This article is intended to promote discourse regarding the interplay between AI and labor markets. The time frames referenced are meant to provide an illustrative scaffold, and certain concepts, including issues of distribution and governance, and nuances between capital, material, labor and total factor productivity, remain unaddressed here to maximize the reach of this discourse. Those further issues will be the subject of robust economic and policy analysis in additional S&P Global research.

Since the public launch of OpenAI’s ChatGPT in November 2022, conversations about AI and the future of work have polarized, with fear and exuberance in similar proportions. Both extremes reflect emotional responses rather than a practical understanding of how AI will transform the labor landscape. The reality is sure to be more nuanced. 

To frame our analysis of AI's effects on labor, we use three periods: the short term (one to three years), the medium term (four to six years) and the long term (seven to 10 years). Each phase will present challenges, ethical considerations and risks that must be balanced against the technology’s potential benefits.

As we progress along this timeline, the effects of AI on labor and labor markets will become increasingly evident across a range of key elements.

In the next few years, AI applications will likely focus on enhancing labor efficiency and productivity in existing workflows. This focus will be dictated by employers’ initial limited understanding of AI's potential in the context of current processes and will reflect experimentation as enterprise decision-makers learn about AI’s core competencies, develop integration approaches and refine their strategies. Companies will likely begin leveraging AI to streamline operations, reduce costs and improve decision-making. Stakeholders will need to determine adequate levels of prudence in adoption, change-management and governance. Adoption without sufficient checks could lead to negative consequences, such as job displacement and privacy concerns, and even where care is taken, the ride is likely to be bumpy and include some retrenchment.

Adoption and implementation of AI in various job families may add complexity to work, posing both opportunities and risks. A key challenge will be to balance the allocation of work between human strengths and AI capabilities in a way that enables the opportunities and abates the risks. Recent data suggests there is reason for optimism. S&P Global Market Intelligence 451 Research surveyed about 5,000 respondents regarding the impact of AI on jobs and on society at large. Between two waves of the survey in 2022 and 2024, respondent expectations regarding AI's impact on society notably shifted, with a significant increase in the expected degree of impact and a modest bias toward positive expectations. In that period, sentiment regarding AI’s impact on jobs remained stable. Notably, among those who expect AI to significantly impact their career, more than twice as many expect a positive impact on work versus a negative impact.

AI-driven automation will increase across capacities including manufacturing, logistics, customer service and data management. By combining information, systems and robotics to handle repetitive tasks, these developments should allow human workers to focus on more complex and creative activities or those with social aspects, such as management, coaching and inspiration. Application is likely to involve assessing the division of labor between tasks that require or benefit from human performance and those that are “machine ready.” For instance, AI-powered chatbots and virtual assistants will continue to evolve, in some cases providing more sophisticated support to human customer service agents, and in other cases providing direct customer support without human intervention.

Meanwhile, optimization agents will assess operations to reduce human input and maximize AI's potential for productivity gains. While this concept may evoke sci-fi tropes that fuel fears of AI, this assessment is a crucial step to maximizing the effectiveness of a mutualism between AI and humans: parsing the tasks that humans are especially good at versus those at which AI excels and facilitating interoperability between the two.

Diving deeper, we see that concerns about AI’s impact on jobs vary across industries, with some showing major concerns while others appear more optimistic. Measures that allay these concerns by abating negative impacts and supporting positive ones may present new opportunities in the work mix for existing positions and may even create new positions to help evolve the workforce toward a future that includes AI.

AI capabilities can improve workforce tooling, especially in minimizing administrative burden, reducing overhead and providing objective performance feedback. In another survey, S&P Global Market Intelligence 451 Research asked approximately 450 respondents about their preferences for workforce tooling innovations. The overwhelming majority focused on efficiency gains and reducing administrative overhead, presumably to allow better focus on "value-add" tasks. Other tools, such as digital twins of physical objects and processes, received less priority among respondents, but these innovations are crucial for providing the digital infrastructure whereby AI can operate.

Discriminative AI systems, which excel at classification, have proven adept at complex data analysis and assisting decision-making across industries such as automotive, manufacturing and logistics. We expect that the combination of discriminative AI and generative AI will also improve service economy activities by drawing on troves of service-oriented data collected over the past two decades. Financial institutions, for example, could use AI algorithms to expedite portfolio managers’ analysis of market trends, risks, correlation and other factors. In healthcare, AI could help diagnose diseases and recommend treatment plans based on vast amounts of medical data, potentially enhancing patient outcomes and reducing diagnostic errors.

Over the next three to five years, AI will significantly reshape labor markets, driven by increased investment in and knowledge of AI systems, and by competitive dynamics that should reduce AI's implementation costs. 

We expect significant advancements in machine learning, natural language processing and robotics, particularly due to improvements in agent-based systems (in which AI systems make decisions and take actions), inference (applying learning from patterns in data) and orchestration (connecting processes and systems). During this period, it will be crucial to address AI's ethical considerations and privacy concerns as regulatory scrutiny will also increase. Wider deployment of AI will also raise the risk of automating deficiencies, including bias, in downstream processes.

Expansion of AI-augmented roles

As AI technologies mature, the scope of roles augmented by AI will expand, especially in fields such as marketing and customer analytics. For example, improvements in AI's ability to meaningfully analyze vast amounts of data on consumer behavior and preferences will facilitate highly personalized marketing campaigns (akin to personalized treatment plans in healthcare). In education, AI-driven platforms will offer bespoke learning experiences, including educational content tailored to individual student needs and pacing, and corporate and client training that meets specific demands.

Emergence of new job categories

An increasing focus on governance, risk mitigation, bias mitigation and other AI-related issues will foster the emergence of new job categories. For example, AI trainers and explainability experts will be needed to ensure that AI systems are fit for purpose, transparent, understandable, and compliant with existing systems and governance frameworks. Skills to address and monitor AI ethics and quality will be crucial for responsible AI deployment, including the safeguarding of fairness and identification of biases. 

AI will become increasingly transformative over longer time frames due to the technology’s advancing maturity, humanity’s growing expertise in effective implementation and the regulatory landscape’s increasing clarity. Collaborative intelligence, a framework in which humans and AI work together, should notably increase.

In creative industries, AI will assist humans by generating ideas, drafts and prototypes. Architects, for example, could use AI-driven design software to explore innovative building designs that incorporate complex environmental and structural factors, leveraging datasets that encompass architectural history, art, design, materials science and weather patterns.

Autonomous systems

AI will support autonomous systems that transform industries. Autonomous vehicles will revolutionize transportation and logistics, reducing the need for human drivers and increasing efficiency. In agriculture, AI-powered robots with computer vision and sensors to detect air pressure, moisture, color and light could oversee planting, crop monitoring and harvesting, increasing productivity and sustainability. Challenges will include cost-effectiveness, which must surpass that of human labor to justify adoption.

Predictive maintenance

Manufacturing and heavy industries will benefit from AI-enabled predictive maintenance. AI will analyze data from sensors embedded in machinery to predict when maintenance is required, preventing costly breakdowns and minimizing downtime. This capability requires AI’s data processing and pattern recognition abilities, coupled with digital telemetry systems to detect acoustic, visual and even olfactory signals.

Autonomous data analysis

In fields such as finance and research, AI-driven data analysis is already indispensable. AI systems can sift through vast amounts of data, identify trends and generate insights with speed and accuracy far beyond human capacity. This capability will be increasingly crucial for making informed decisions in dynamic environments and will enable a better understanding of the world around us, including risks such as climate change, the energy crisis and geopolitical dynamics.

Shifting labor dynamics

Advancements in the capabilities of AI systems and the dynamics of their integration with human labor will necessitate broader rethinking to understand the adaptations required of both AI and human work paradigms to improve cooperation. 

Humans, for instance, outperform machines in relational tasks such as comprehensive (verbal and nonverbal) communication and, at least to date, innovation. Thus, in a framework of human-AI mutualism, we expect the balance of such tasks to fall to humans, while many data analysis and computational tasks will fall to AI. This may require workforce education in interpersonal skills such as emotional intelligence, empathy and clarity of communication. Humans will also need better human-to-AI communication skills, such as prompt engineering, and a deeper understanding of AI operational requirements.

Consider a typical customer service scenario: A customer is unsatisfied with the quality of goods received. In a world where AI and humans work together, required skills include understanding the customer’s problem (human), analyzing possible solutions (AI), relaying options with empathy to address the customer’s emotional and social needs (human), and executing a course of action to address functional needs (AI).

Leadership and the workplace of choice

Leadership in an AI-augmented world will evolve as workers look to their leaders for vision, systems thinking, and the ability to both analyze (extrapolate trends using context, qualitative judgments and data-driven insights) and synthesize (understand the interoperability of moving parts and connection to a greater whole). As leaders will manage humans and AI systems together, the importance of these abilities cannot be overstated.

Analytical functions will become more crucial as AI adoption takes hold, requiring leaders to adapt to increasing customization and diversification of product offerings, as well as to the interplay of skilled workers and AI systems required to deliver those products. An understanding of synthesis will be critical as many traditional work functions and systems become intersecting, looping and chained processes in increasingly complex designs.

Such complexity will likely change what humans demand of their workplaces. Traditional compensation schemes such as salary and variable pay will likely be augmented by purpose, culture and fit as an increasing proportion of tasks allocated to human workers will require distinctly human capabilities, such as emotional and social engagement. Organizations must recognize that their brand represents not only their product or service offerings or standard of quality to customers but also a workplace of choice for employees. Workplaces will need to earn their employees’ business, especially as employees’ nuanced personalities and skills become features in the workplace itself, as much of the labor characterized by homogeneity becomes the province of AI systems.

The idea of the modern-day John Henry will remain relevant as AI’s remit expands and matures. However, it is equally likely that many mundane, monotonous and undesirable elements of work could become increasingly automated, and that, as a result, humans — particularly those who learn to work in partnership with AI — may find increasing value and standing in the labor force by cultivating and expressing their most fundamentally human qualities.