AI's Impact on Job Market: Destruction and Creation

AI will, in due course, destroy jobs, but it will also create new ones - The Economic Times

Will AI kill millions of jobs? Optimists hope AI will create more jobs than it destroys. One hopeful sign comes from the severe labour shortage globally in manufacturing chips for AI. This may hurt India in the short run, but it should be an advantage in the long run. TSMC (Taiwan Semiconductor Manufacturing Company) is the world's largest maker of chips needed for the AI revolution. Its sales rose 45% in July to $7.9 bn, thanks mainly to increased demand for its chips after the arrival of AI. However, it now fears a manpower crisis. Corporations earlier assumed that money was the main hurdle for the very high investment needed for chip manufacturing. Governments worldwide began viewing chip-making and AI-enabled chips as strategically vital and offered subsidies galore for new investment. India has offered subsidies of a whopping 70% for an investment by Micron that is not even for chip fabrication - it's for assembly, testing, marking and packaging. Tata is building a fabrication plant in Gujarat and a chip assembly plant in Assam. Adani has long expressed determination to get into this sector too. Sums required are humungous. TSMC is investing $65 bn in three chips plants in Arizona. It will get $6.5 bn in grants and $5 bn in cheap loans from the government, and has applied for investment tax credits of 25% of capital costs. But cash, it turns out, is not the only hurdle. Setting up and running a chip plant requires specialist workers. A McKinsey report estimates that the US AI sector requires more than 1.6 lakh new engineers and technicians in the next five years. Yet, barely 1,500 engineers and 1,000 chip technicians join the chip industry each year. The excitement over AI is one reason why the US Bureau of Labor Statistics estimates that the US will need an additional 1 mn workers in STEM in 2030 than it did in 2020, without even counting the workers who retire before then. Hence, US corporations are lobbying hard for more skilled migrants from India. Almost 80% of global chip manufacturing has shifted to South Korea and Taiwan in the last 25 years. However, even South Korea will have a shortage of 56,000 workers by 2031. Both South Korea and Taiwan have falling populations. Students enrolling in higher education have been declining in numbers every year since 2012. What does this mean for India? Gradual wave: The AI revolution will take much longer than expected. It will not be a tidal wave crushing employment in a thousand sectors. It will be far more gradual, given intrinsic difficulties and the coming skilled manpower shortage. Manpower-hungry: This will affect all Indian plants in the semiconductor sector. Like their compatriots in the US, South Korea and Taiwan, Indian companies are going to have a tough time hiring and training enough skilled manpower. Importing technicians from South Korea and Taiwan will not be easy, given the shortage of technicians in both countries. Low-end technologies are easily available in semiconductors, but AI is at the cutting edge. Train, train: India needs to embark on a massive training programme to develop the brainpower needed for an AI revolution. Experience in high-level chip manufacturing and operating is scarce globally. India needs partnerships with top global companies for both expertise and investment. STEM the tide: India has a huge underlying advantage in manpower. It produces an estimated 3.2 mn STEM graduates a year. At least a quarter are unemployable graduates from third-rate colleges. But no other country, save China, which is becoming globally untouchable, produces remotely as many STEM graduates. Even if one-third are unusable, that means India produces over 2 mn employable STEM graduates a year. Indian IT companies complain of a shortage of at least 1 lakh software engineers, but the shortage is of skilled techies, not rookies. Only a fraction of fresh STEM graduates will find IT jobs. The best of them, coming from IITs, often go abroad. But more than enough remain in India to be hired by the 1,500 MNCs that have set up GCCs in India. GCCs were originally called in-house centres, doing low-level wage-intensive jobs like call centres and medical transcription. They have now developed into design and R&D centres, with an emphasis on AI. Microsoft, Google, Amazon and IBM are among those using GCCs to develop AI. This means invaluable high-level training for Indian techies who can later be hired by Indian companies. No training programme devised by the government could ever match the learning-by-doing that GCCs are providing. Even better, analysts expect hundreds of additional MNCs to set up GCCs in India in the next five years. This should further expand the skilling of Indian techies. India is in a sweet spot. AI may in due course destroy jobs, but will also create jobs. Every hi-tech job is associated with several low-end jobs, so AI will be a job multiplier. There lies hope.

AI will, in due course, destroy jobs, but it will also create new ones

Will AI kill millions of jobs? Optimists hope AI will create more jobs than it destroys. One hopeful sign comes from the severe labour shortage globally in manufacturing chips for AI. This may hurt India in the short run, but it should be an advantage in the long run. TSMC (Taiwan Semiconductor Manufacturing Company) is the world's largest maker of chips needed for the AI revolution. Its sales rose 45% in July to $7.9 bn, thanks mainly to increased demand for its chips after the arrival of AI. However, it now fears a manpower crisis. Corporations earlier assumed that money was the main hurdle for the very high investment needed for chip manufacturing. Governments worldwide began viewing chip-making and AI-enabled chips as strategically vital and offered subsidies galore for new investment. India has offered subsidies of a whopping 70% for an investment by Micron that is not even for chip fabrication - it's for assembly, testing, marking and packaging. Tata is building a fabrication plant in Gujarat and a chip assembly plant in Assam. Adani has long expressed determination to get into this sector too. Sums required are humungous. TSMC is investing $65 bn in three chips plants in Arizona. It will get $6.5 bn in grants and $5 bn in cheap loans from the government, and has applied for investment tax credits of 25% of capital costs. But cash, it turns out, is not the only hurdle. Setting up and running a chip plant requires specialist workers. A McKinsey report estimates that the US AI sector requires more than 1.6 lakh new engineers and technicians in the next five years. Yet, barely 1,500 engineers and 1,000 chip technicians join the chip industry each year. The excitement over AI is one reason why the US Bureau of Labor Statistics estimates that the US will need an additional 1 mn workers in STEM in 2030 than it did in 2020, without even counting the workers who retire before then. Hence, US corporations are lobbying hard for more skilled migrants from India. Almost 80% of global chip manufacturing has shifted to South Korea and Taiwan in the last 25 years. However, even South Korea will have a shortage of 56,000 workers by 2031. Both South Korea and Taiwan have falling populations. Students enrolling in higher education have been declining in numbers every year since 2012. What does this mean for India? Gradual wave: The AI revolution will take much longer than expected. It will not be a tidal wave crushing employment in a thousand sectors. It will be far more gradual, given intrinsic difficulties and the coming skilled manpower shortage. Manpower-hungry: This will affect all Indian plants in the semiconductor sector. Like their compatriots in the US, South Korea and Taiwan, Indian companies are going to have a tough time hiring and training enough skilled manpower. Importing technicians from South Korea and Taiwan will not be easy, given the shortage of technicians in both countries. Low-end technologies are easily available in semiconductors, but AI is at the cutting edge. Train, train: India needs to embark on a massive training programme to develop the brainpower needed for an AI revolution. Experience in high-level chip manufacturing and operating is scarce globally. India needs partnerships with top global companies for both expertise and investment. STEM the tide: India has a huge underlying advantage in manpower. It produces an estimated 3.2 mn STEM graduates a year. At least a quarter are unemployable graduates from third-rate colleges. But no other country, save China, which is becoming globally untouchable, produces remotely as many STEM graduates. Even if one-third are unusable, that means India produces over 2 mn employable STEM graduates a year. Indian IT companies complain of a shortage of at least 1 lakh software engineers, but the shortage is of skilled techies, not rookies. Only a fraction of fresh STEM graduates will find IT jobs. The best of them, coming from IITs, often go abroad. But more than enough remain in India to be hired by the 1,500 MNCs that have set up GCCs in India. GCCs were originally called in-house centres, doing low-level wage-intensive jobs like call centres and medical transcription. They have now developed into design and R&D centres, with an emphasis on AI. Microsoft, Google, Amazon and IBM are among those using GCCs to develop AI. This means invaluable high-level training for Indian techies who can later be hired by Indian companies. No training programme devised by the government could ever match the learning-by-doing that GCCs are providing. Even better, analysts expect hundreds of additional MNCs to set up GCCs in India in the next five years. This should further expand the skilling of Indian techies. India is in a sweet spot. AI may in due course destroy jobs, but will also create jobs. Every hi-tech job is associated with several low-end jobs, so AI will be a job multiplier. There lies hope.