AI and Job losses: White House Releases Policy Document

INDIA STILL TO WAKE UP TO AI IMPACT 

Latest analysis based on World Bank data, suggests that 77% and 69% of jobs China and India at risk compared to 47% in the US

In all likelihood most of us who use smartphones and messaging apps and virtual assistants like Google Allo, Apple Siri, Facebook Messenger, WeChat, or more recently Amazon Echo, which is yet to be launched in India, have already experienced the beauty of Artificial Intelligence. All of fun and easy to play around with, what however AI apps portend for countries like India is quite chilling. 

More than the scare of robots and intelligent assembly lines taking away factory jobs, it is white collar tech jobs that employ the young, and contact center jobs that employ the young and the not so highly educated, that are under threat.

Latest Update

Automation has started showings its impact on Indian IT jobs sooner than expected. Infosys has “released” 8000-9000 employees in the past one year as lower end jobs have been automated, says the Economic Times quoting the company’s HR head Krishnamurthy Shankar. Infosys has been releasing 2000 people every quarter. These employees are now being trained for other higher end jobs, he said.

In 2016 AI emerged from the shadows of academic and tech institutions’ labs into the world of startups and entrepreneurial space looking for opportunities for actual implementations and business models truly kicking off what the industry has started calling the Fourth Industrial Revolution. The inflexion between AI and natural language processing driven by big data and deep learning capability of machines that enable 90% accuracy in machine parsing and better understanding of the spoken word or typed queries by machines, is driving most of this entrepreneurial activity heralding greater automation of most industries.

At last count there were upwards of 2277 startups working in AI at a total investment upwards of $14 billion. India accounted for some 170 of these companies with an investment of close to $300 million, According to consulting firm Zinnov.

Apart from their use in what experts are calling ‘conversational commerce’, in 2016 BOTS also went mainstream in India across industry sectors including banking, outsourced services and, lo and behold, journalism and writing. ICICI Bank started experimenting with BOTS claiming to be the first bank in India to start using AI for tasks like generating customer IDs, updating addresses and mobile numbers, resolving ATM-related queries, etc. Over 200 business processes at the bank are already carried out by BOTS resulting in a 60% improvement in response time, the management claims. Likewise HDFC Bank too had announced it plans to introduce an AI driven robot at one of its office on an experimental basis.

Increased productivity and cost reduction is the motivation for these implementations. But the downside, that experts have been warning of for several years, is now quite obvious – job losses. While AI led automation will be a global phenomenon, the hardest hit will be the people in the emerging markets like India, China and Philippines which have benefited the most from the technology-led services job boom. Where the West is concerned the impact will be felt in the coming years more in the automotive sector, particularly transportation and allied fields, where autonomous vehicles threaten to take over.

“Advances in AI will prove to be disruptive, resulting in new opportunities for collaboration between humans and machines, as well as a loss of traditional jobs such as legal analysts, financial and sports reporters, online marketers, anesthesiologists, diagnosticians, and financial analysts. Similarly, a large number of call centers previously offshored to developing countries could be affected by increasingly sophisticated natural language processing systems that can substitute for human workers”, the World Bank said in its 2016 World Development Report. Instances of such job losses due to introduction of smart virtual assistants are already visible.

Building on their seminal work in 2013, Oxford Martin School’s Carl Benedikt Frey and Michael A. Osborne’s recently estimated that while 47% of the US workforce is at risk of automation, alarmingly, their estimates based on latest World Bank data, suggest that 77% and 69% of jobs China and India are respectively, are under threat. Frey and Osborne highlight that the impact of AI will be more accentuated on task oriented jobs rather than occupations implying workers employed in jobs that involve repetitive activities will be at risk on account of automation.

Interestingly, the American government, notwithstanding its preoccupation with the changeover to the Trump administration appears to be more seized of this issue. In report issued late last month by President Obama’s executive office, the outgoing administration set broad policy prescriptions on the government’s response to the disruption from AI.

The White House advocates three specific strategies to educate and prepare new workers to enter the workforce, cushion workers who lose jobs, to keep them attached to the labour force and to combat inequality. As such the report goes on to provide three recommendations:

• Invest in and develop AI for its many benefits. 
• Educate and train Americas for jobs for the future. 
• Aid workers in the transition and empower workers to ensure broadly shared growth. 

While the 55 page report does accept that it is possible for the economy to generate high levels of employment with more advanced automation and higher levels of productivity than there are today, it nevertheless adds that Federal policies will need to play a role in helping Americans to navigate transitions in labor market demand caused by changes in technology over time. While the first three strategies deal with capacity building within the economy to deal with the opportunities and demands from AI in terms of investing and skilling for it, the last is perhaps what reflects the deeper concern within the US administration on the impact of AI on jobs.

The prescription highlights modernizing and strengthening the social safety net for displaced workers to ensure they can still make ends meet, retrain and potentially transition careers. This basically calls for strengthening critical supports such as unemployment insurance, Medicaid, Supplemental Nutrition Assistance Program (SNAP), and Temporary Assistance for Needy Families (TANF), and putting in place new programs such as wage insurance and emergency aid for families in crisis among other things. It also recommends strengthening the 

Unemployment Insurance Program and Medicaid assistance to unemployed. “ If, however, the economic impact of AI is relatively intense or comes on relatively quickly, and if the number of jobs affected approaches the estimates of Frey and Osborne, then the unemployment insurance system may need to be significantly upgraded to match the magnitude of economic disruption and ensure that displaced workers do not leave the labor force,” it stresses. 

While Western governments definitely seem to be bothered by this, the moot question is whether India is taking note? If so, then there is no visible indication either by the official establishment or even the industry which at best has paid lip service only and actually highlighted the opportunities from AI for itself. It is time India took notice of the impending downside of AI.

The Indian IT-BPM industry in India, estimated at nearly $146 billion in FY 2015, currently provides direct employment to of 3.7 million and an indirect employment up to 10 million people. Apex industry body Nasscom projects this to reach around USD 350-400 billion by 2025. AI will generate $70+ billion in productivity-led gains for global IT-BPM service providers particularly in the areas of consulting, engineering services, application development, management and testing, BPM and infrastructure services, a Nasscom-Boston Consulting Group study says. AI could also open up $400+ billion of net new addressable market; of which up to $100 billion will be realisable by 2020, it adds. The study is quite silent on how many jobs AI would generate in the Indian economy. But many are clear AI will replace more jobs than it will create.