Critically assess the impact of technological advancement and automation on the nature of work and employment. (2024)
Automation alters not merely how many people work but what working is — the skill a job carries, who supervises it, and the legal form under which it is done. A sociological assessment must therefore separate three questions: the volume of employment, the quality of work, and the structure of control. On volume the alarmists have been wrong; on quality and control the critics have been right.
The optimistic case: automation upgrades work
- Robert Blauner (Alienation and Freedom, 1964) argued that alienation follows an inverted U across technologies: high in Fordist assembly work, but falling under continuous-process automation, where the worker monitors rather than serves the machine.
- Daniel Bell (The Coming of Post-Industrial Society, 1973) made theoretical knowledge the axial principle of the emerging order and predicted the pre-eminence of a professional and technical stratum.
- The job-loss forecasts have not survived scrutiny. Carl Benedikt Frey and Michael Osborne (2013) placed 47% of United States employment in a high-risk category, but Melanie Arntz, Terry Gregory and Ulrich Zierahn (OECD, 2016) showed that this rests on treating whole occupations as automatable; measured by tasks, only about 9% across 21 OECD countries were at high risk. The International Labour Organization’s 2025 exposure index agrees: about a quarter of global employment is exposed to generative AI, only 3.3% in the highest band, and just 11% in low-income countries.
The critical case: degradation, polarisation and algorithmic control
- Harry Braverman (Labor and Monopoly Capital, 1974) held that technology under capitalism is designed to separate conception from execution, transferring workers’ knowledge to management. Automation is a strategy of control before it is one of efficiency.
- The evidence supports a middle position — skill polarisation: routine tasks, manual and cognitive alike, are the automatable ones, so the occupational middle hollows out while professional work and low-paid personal services expand. The ILO index shows this is gendered: clerical occupations are most exposed, and 4.7% of women’s employment globally falls in the highest band against 2.4% of men’s.
- Shoshana Zuboff (In the Age of the Smart Machine, 1988) distinguished automating from informating: the system that replaces effort also generates a continuous record of the worker. A. Aneesh (Virtual Migration, 2006) named the resulting regime algocracy — governance through code rather than bureaucratic command or market price.
The Indian case
India’s automation is an enclave: 8,510 industrial robots were installed in 2023, two-fifths of them automotive, on an operational stock under 45,000. Yet the Periodic Labour Force Survey for 2025 records 56.2% of workers self-employed and 20.2% casual, against only 23.6% in regular wage employment. Technology here does not destroy jobs so much as informalise them. The Karnataka Platform Based Gig Workers (Social Security and Welfare) Act, 2025 obliges aggregators to fund welfare through a levy on payouts, give notice before termination, and disclose how automated decision-making affects workers — the first Indian recognition that the algorithm is an employer.
Conclusion
Neither Blauner’s optimism nor Braverman’s pessimism holds as a law; both describe outcomes that institutions decide. Automation’s real effect is to make control invisible and employment contingent, which is why the sociological question has moved from how many jobs survive to who writes the code that governs them.