Page 767 - ISC PROCEEDINGS 21.4

substantial lever for improving financial metrics and satisfying investor demands for
profitability.
2.3. Technological disruption, ai emergence, and labor substitution
The emergence of generative artificial intelligence during 2022-2023, particularly
large language models and diffusion-based image generation systems, introduced a novel
dimension to the layoff wave that complicates straightforward cyclical or monetary policy
explanations. While previous automation waves primarily affected routine manual and
cognitive tasks—a pattern extensively documented by Autor et al. (2003) in their seminal
work on skill-biased technological change—generative AI's capabilities extend to creative
and knowledge work traditionally considered automation-resistant. The author and
colleagues demonstrated that computerization disproportionately substituted for routine
cognitive and manual tasks while complementing non-routine analytical and
interpersonal tasks, predicting a polarization of labor markets between high-skill and low-
skill employment.
Yet the relationship between AI adoption and employment remains theoretically
and empirically contested. Acemoglu and Restrepo (2019) provide a more nuanced
framework distinguishing between displacement effects, where technology substitutes
for labor in existing tasks, and productivity effects, where technology complements
human capabilities and increases demand for workers through new task creation and
productivity-driven expansion. Their framework suggests that the net employment effect
of AI depends on the relative magnitudes of these countervailing forces, which vary
across occupations, industries, and time horizons. The current period may represent a
transitional phase where displacement effects dominate short-term employment
decisions and generate immediate cost savings, even as longer-term productivity gains
and task reinstatement remain uncertain and difficult to quantify for risk-averse
executives facing investor pressure.
Acemoglu and Restrepo (2019) further argue that the distributional consequences
of automation depend critically on how productivity gains are shared between workers
and capital owners, and whether displaced workers can transition to newly created tasks
or industries. In the technology sector specifically, the question becomes whether AI
displaces specific occupational categories while creating demand for AI trainers,
supervisors, and complementary roles, or whether the technology fundamentally reduces
the labor intensity of software production. The answer carries profound implications for
long-term employment trends in one of the developed world's largest employment
sectors.
2.4. Geographic concentration, agglomeration economies, and spatial
vulnerability
Research on geographic concentration highlights the dual-edged nature of industry
clustering. As Glaeser and Gottlieb (2009) demonstrate, agglomeration economies foster
innovation through knowledge spillovers, specialized labor pools, and firm linkages,
transforming regions like Silicon Valley and Seattle into highly productive technology hubs
that attract continuous employment growth.
Conversely, this same concentration creates severe vulnerability to sector-specific
shocks, as regional economies become over-reliant on a single industry (Glaeser &
Gottlieb, 2009). When tech companies simultaneously reduce headcount, the negative
multiplier effects cascade through local economies, depressing consumer retail,

766