The conversation around artificial intelligence and its impact on employment has reached a fever pitch, with headlines warning of massive job losses and societal upheaval. In a recent guest essay for the New York Times, David Solomon, the chief executive officer of Goldman Sachs, pushes back against the notion of an imminent AI‑driven job apocalypse, characterizing such fears as overblown. He argues that while AI will undoubtedly reshape the labor landscape, the United States possesses the institutional flexibility, entrepreneurial spirit, and policy tools needed to adapt and even expand its workforce in response to these technological shifts. Solomon’s perspective is grounded in a belief that history offers repeated examples of economies absorbing disruptive innovations without enduring long‑term unemployment spikes, provided that proactive measures are taken to harness the technology’s productivity gains.
To contextualize Solomon’s optimism, it is useful to revisit past waves of technological anxiety. When the steam engine, electricity, and later the personal computer entered the mainstream, economists and commentators alike warned of widespread displacement. John Maynard Keynes famously speculated in 1930 that advances would eventually allow a 15‑hour workweek by 2030—a vision that remains unfulfilled, yet it serves as a reminder that forecasts of joblessness often overlook the capacity of new technologies to spawn entirely new industries and roles. Solomon invokes this historical precedent to suggest that the current apprehension may be missing the bigger picture: AI’s potential to boost productivity could fuel economic growth that, in turn, creates demand for labor in sectors we have not yet imagined.
Solomon bases his confidence on three interconnected arguments. First, he contends that AI will automate routine, repetitive components of many jobs, thereby freeing human workers to devote more time to complex, creative, and strategic tasks that machines cannot easily replicate. Second, he believes that rather than rendering entire professions obsolete, AI will raise the standards and effectiveness of those professions by handling data‑intensive chores, allowing professionals to focus on judgment, client relationships, and innovation. Third, he anticipates the emergence of new job categories centered on the development, deployment, oversight, and maintenance of AI systems themselves—roles that will require a blend of technical expertise and domain knowledge, thus absorbing workers displaced from more automatable functions.
A critical piece of evidence supporting Solomon’s outlook comes from Goldman Sachs’ own analysis, which estimates that AI could automate roughly a quarter of current work hours over the next ten years. This figure, while significant, is far from the total replacement scenario that some alarmists predict. A 25 % reduction in hours translates into a shift in how work is organized rather than a wholesale elimination of positions. For many organizations, this could mean shorter workweeks, more flexible scheduling, or the reallocation of saved time toward higher‑value projects, provided that companies and policymakers manage the transition thoughtfully.
The essay highlights that white‑collar fields such as accounting, banking, and law are particularly susceptible to having specific tasks automated. Routine bookkeeping, basic tax preparation, contract review, and due‑diligence research are areas where AI excels due to its pattern‑recognition capabilities and ability to process large volumes of structured data quickly. However, Solomon notes that the core functions of these professions—strategic financial advising, complex litigation strategy, and personalized client counseling—remain firmly within the human domain. The implication is that professionals in these sectors will need to upskill in areas like AI literacy, data interpretation, and interdisciplinary collaboration to stay competitive.
When it comes to blue‑collar occupations, Solomon acknowledges greater uncertainty about the magnitude and nature of AI’s impact. Manufacturing, logistics, and maintenance roles involve a mix of physical dexterity, situational awareness, and problem‑solving that current AI systems have not yet mastered at scale. Nevertheless, he points out that AI can augment these jobs by providing predictive maintenance alerts, optimizing supply‑chain routing, or offering real‑time safety monitoring via computer vision. The net effect may be a transformation of job descriptions rather than outright elimination, emphasizing the importance of training programs that teach workers to interact confidently with AI‑enhanced tools.
Beyond the white‑collar sphere, recent research from McKinsey underscores a growing trend that could affect early‑career entrants. In a 2025 survey, 51 % of organizations reported that generative AI is diminishing their need for entry‑level positions, particularly those centered on repetitive data entry, basic customer service inquiries, and preliminary report generation. This shift raises concerns about the traditional pipeline that feeds junior talent into more senior roles, suggesting that companies may need to redesign internship programs, apprenticeships, and graduate rotations to incorporate AI‑augmented tasks that still provide meaningful learning experiences.
Goldman Sachs economists have identified specific occupations that face a high risk of being outright substituted by AI. Roles such as telephone operators, insurance claims representatives, and bill collectors involve highly scripted interactions and rule‑based decision‑making that are readily handled by natural‑language processing and robotic process automation. As these functions become increasingly automated, workers currently employed in them will need accessible pathways to reskill into adjacent areas like customer experience design, AI‑assisted case management, or technical support for AI platforms.
Conversely, certain professions appear poised for augmentation rather than replacement. Education administrators, physicians and surgeons, construction managers, and chief executives are cited as examples where AI can enhance decision‑making through predictive analytics, imaging assistance, project‑schedule optimization, and strategic scenario planning, respectively. In these contexts, the technology serves as a force multiplier, enabling professionals to handle larger caseloads, undertake more complex projects, or pursue innovative strategies that would be impractical without AI’s computational power.
The debate is not one‑sided, however. Daron Acemoglu, an MIT economist known for his measured assessments of AI’s economic effects, has warned against “excessive automation”—a scenario where firms deploy AI primarily to cut labor costs without reinvesting the productivity gains into new tasks or roles. Acemoglu argues that if AI merely replaces existing work without creating complementary activities, it could lead to a work shortage, lower labor‑force participation, and a proliferation of low‑skill, menial jobs. He emphasizes that the desirable outcome is one where AI augments human labor, expanding the set of tasks workers can perform and thereby sustaining or even increasing overall employment.
Given these divergent views, Solomon advocates for a coordinated response between the public and private sectors should AI begin to displace jobs at an unprecedented pace. He calls for joint initiatives that invest in large‑scale reskilling programs, strengthen unemployment insurance safety nets, and incentivize businesses to redesign jobs around human‑AI collaboration rather than pure substitution. Public policy could play a role by funding community colleges and technical schools to offer AI‑relevant curricula, while tax credits or grants might encourage firms to pilot internal mobility programs that transition at‑risk workers into emerging AI‑focused positions.
For individuals navigating this evolving landscape, the actionable advice is clear: cultivate a habit of continuous learning, focusing on digital fluency, data literacy, and the ability to work alongside AI tools. Professionals should seek out projects that involve AI implementation, even in a peripheral capacity, to build relevant experience. Employers, meanwhile, ought to conduct systematic audits of their workflows to identify tasks ripe for automation and then redesign those roles to emphasize higher‑order skills, offering clear pathways for internal mobility. Policymakers should prioritize investments in education infrastructure, portable benefits systems, and social dialogue mechanisms that ensure the gains from AI are broadly shared, ultimately turning the technology into a catalyst for inclusive economic growth rather than a source of widespread dislocation.