The rise of the human–AI workforce

The widespread deployment of AI technology will necessitate a new understanding of interaction models, much like how people use smartphones without knowing their underlying mechanics. McKinsey Senior Partner Alexis Krivkovic and Partner Anu Madhavkar discuss the implications of people increasingly working with agents and robots. Their report, "Agents, Robots, and Us," emphasizes a constructive partnership between humans and these technologies, rather than a "humans versus robots" dynamic. Alexis highlights the profound, top-to-bottom scale of AI's impact across every function and industry, from frontline to C-suite. The primary challenge for organizations is absorbing such rapid and extensive change. Anu finds excitement in the opportunity for individuals to "super skill" themselves by working with AI, recognizing that shared skills between humans and AI mean everyone needs to leverage AI to enhance their abilities. This also presents a responsibility to upskill the entire workforce. Despite apocalyptic headlines about AI and job loss, the research indicates that while over half of US work hours could be automated by current technologies, this doesn't equate to job loss. Instead, human beings remain vital, especially for cognitive, social, emotional, interpersonal, and physical tasks. The adoption of AI creates new types of work, such as guiding, prompting, validating, refining, or building upon AI's output, and enables entirely new capabilities and demands, improving quality and expanding possibilities. High-value human contributions in the next five to ten years will be defined by critical thinking skills like problem-solving, negotiation, conflict management, and team leadership. Specialized areas like trauma care or complex surgery will still require human expertise, even with robotic assistance. Conversely, administrative tasks, customer onboarding, and client communications are ripe for increased automation. The example of radiology illustrates how AI can unlock scarce skills, enabling faster image processing and data inference, thereby expanding the supply to meet latent market needs. Beyond existing skills, the report identifies social-emotional, negotiation, coordination, and process management skills as increasingly important. AI will free up human time for these areas. Leaders, in particular, will need a "voracious learning mindset." Alexis notes that skills-based hiring is gaining traction as job descriptions shift. CHROs are grappling with rewriting job descriptions and finding that candidates are using AI to optimize their applications. This may lead to a return to "old school" assessment methods, like in-person tests or live coding. Capturing the estimated $3 trillion annual value of AI by 2030 requires organizations to scale pilot projects into company-wide initiatives. Leadership teams must align on a value thesis and marshal resources to attack opportunities at scale, especially since major opportunities often span multiple departmental domains. Anu suggests a "T-shaped approach," combining broad capability building with strategic bets on end-to-end process transformations. Companies must also consider market trends and how AI disrupts profit pools, whether through internal innovation or competitor threats. The widespread deployment of AI will require everyone to understand how to deploy new interaction models. While not everyone needs to understand AI's inner workings, like a smartphone user, working with agents demands learning to validate, provide judgment, redirect, and iteratively test and learn. This cultivates a culture of experimentation and critical judgment. Managing a hybrid team of human and AI agents will fundamentally question traditional productivity paradigms and KPIs. AI's ability to generate massive outputs, like 5,000 reports overnight, will create new human bottlenecks, driving further innovation. Effective managers will need an appetite for flux, resilience, and creativity. Redesigning education based on this research would involve preserving foundational cognitive and physical abilities, even with AI assistance, to avoid over-reliance on technology. The focus should shift from highly specialized learning to more transferable, foundational, and generalizable skills, given the dynamic nature of the workforce. AI itself can facilitate this reskilling through personalized learning paths and in-the-moment guidance, tailored to individual skill gaps and work contexts. Moving too slowly with AI poses the greatest risk. Speed is a strategy, and waiting for clarity amidst ambiguity is dangerous. Companies risk falling too far behind if they delay making bets. While ethical considerations, risk management, regulation, and public education are crucial, they must be addressed with speed to seize the opportunity. Ten years from now, a successful human-AI partnership would see AI augmenting and unleashing human potential, not replacing it. It would be a positive, integrated part of daily life, like smartphones or the internet, with focus shifting to control, equity, and access, rather than questioning its existence or inherent goodness. Anu hopes AI enhances the quality and experience of work, making it more fulfilling, and that its benefits are democratized, leading to transformational effects, wider access, and better quality for all.