As artificial intelligence reshapes the global economy, a stark pattern emerges—one that would make even Victorian industrialists blush: the financial rewards of this technological revolution are flowing disproportionately to a narrow elite of AI owners, developers, and high-skill workers. This capital-biased technological change threatens to exacerbate existing wealth disparities, creating a two-tier society where the gains from increased productivity concentrate amongst those who control the technology, whilst large swathes of the population face stagnant wages or unemployment. One might say we’re building a brave new world, though Aldous Huxley might suggest we should have read the fine print (Piketty, 2014; Korinek and Stiglitz, 2019).
The economics of AI differ fundamentally from previous technological shifts. Unlike industrialisation, which eventually created broad-based employment opportunities (albeit after considerable suffering), AI systems can scale indefinitely with minimal additional labour input. A successful AI model can be deployed millions of times at negligible marginal cost, generating enormous returns for its owners without proportional job creation—a feature rather than a bug, as Silicon Valley might say (Brynjolfsson et al., 2018). This dynamic favours capital over labour in unprecedented ways, as shareholders and executives of AI companies capture value that would traditionally have been distributed more widely through wages (Acemoglu and Restrepo, 2022).
The concentration of AI capabilities amongst a handful of technology giants compounds this problem. Companies with vast computing resources, proprietary datasets, and top-tier AI talent enjoy substantial competitive advantages that smaller firms cannot easily replicate—creating barriers to entry higher than a medieval castle wall, and considerably more expensive to breach (Khan, 2017). This cements existing market power and ensures that AI-driven profits remain concentrated. Meanwhile, workers whose skills complement AI systems command premium salaries, whilst those in routine occupations face downward wage pressure or displacement, creating what economists politely term ‘skill-biased technological change’ and what affected workers might call something rather less academic (Autor et al., 2020).
Geographic inequality adds another dimension to this divide. AI development clusters in a few technology hubs in advanced economies, channelling wealth to these regions whilst leaving others behind. Within countries, urban centres with strong educational institutions and technology sectors pull ahead, whilst rural and post-industrial areas struggle to participate in the AI economy (Florida, 2017). This spatial dimension of inequality can destabilise entire regions and fuel political discontent—a phenomenon we might charitably describe as ‘suboptimal for social cohesion’ (Case and Deaton, 2020).
Addressing AI-driven inequality requires bold policy interventions that go beyond wishful thinking and corporate promises of ‘trickle-down innovation’. Proposals range from progressive taxation of AI companies and automation taxes to universal basic income schemes that redistribute productivity gains (Susskind, 2020; Atkinson and Luttrell, 2021). Strengthening worker bargaining power, investing in education and retraining, and ensuring broad access to AI tools and infrastructure also merit consideration. Without such measures, the promise of AI to improve living standards risks becoming a reality for only a fortunate few, whilst widening the gulf between society’s winners and losers—a gulf that may eventually require more than a metaphorical bridge to cross (OECD, 2021).
References
Acemoglu, D. and Restrepo, P. (2022) ‘Tasks, automation, and the rise in US wage inequality’, Econometrica, 90(5), pp. 1973-2016.
Atkinson, R.D. and Luttrell, C. (2021) ‘Why and how to tax robots’, Information Technology & Innovation Foundation, May.
Autor, D.H. et al. (2020) ‘The fall of the labor share and the rise of superstar firms’, The Quarterly Journal of Economics, 135(2), pp. 645-709.
Brynjolfsson, E. et al. (2018) ‘Artificial intelligence and the modern productivity paradox’, in Agrawal, A., Gans, J. and Goldfarb, A. (eds.) The Economics of Artificial Intelligence. Chicago: University of Chicago Press, pp. 23-57.
Case, A. and Deaton, A. (2020) Deaths of Despair and the Future of Capitalism. Princeton: Princeton University Press.
Florida, R. (2017) The New Urban Crisis: Gentrification, Housing Bubbles, Growing Inequality, and What We Can Do About It. London: Oneworld Publications.
Khan, L.M. (2017) ‘Amazon’s antitrust paradox’, Yale Law Journal, 126(3), pp. 710-805.
Korinek, A. and Stiglitz, J.E. (2019) ‘Artificial intelligence and its implications for income distribution and unemployment’, in Agrawal, A., Gans, J. and Goldfarb, A. (eds.) The Economics of Artificial Intelligence. Chicago: University of Chicago Press, pp. 349-390.
OECD (2021) Bridging Digital Divides in G20 Countries. Paris: OECD Publishing.
Piketty, T. (2014) Capital in the Twenty-First Century. Cambridge, MA: Harvard University Press.
Susskind, D. (2020) A World Without Work: Technology, Automation and How We Should Respond. London: Allen Lane.