The year 2020 has challenged governments and deepened socio-economic inequalities more than ever before. Telecommuting has allowed some businesses to survive by enabling skilled workers to work from home. However, low-skilled workers have no opportunities along these lines. As a result, 2 billion low-skilled workers from the informal sector globally have been rendered helpless due to Covid-19. Due to low digital literacy, lack of computers or reliable internet connectivity, and ultimately the nature of their occupations — many low-skilled workers have not been able to make transition to remote online work.
What is the future of work for those who have lost their jobs due to Covid-19? How can they access opportunities within the digital economy? Artificial intelligence (AI) may provide an unlikely solution.
As AI applications grow, they are creating new jobs not only in high-skilled domains like coding but also in low-skilled ones due to its ever-increasing demand for foundational human input. Essentially, human workers are needed to prepare data for AI, annotate the datasets used to train AI models, monitor the performance of these models, and correct inaccurate predictions. Data labeling has thus emerged as a massive industry in its own right.
Due to low digital literacy, lack of computers or reliable internet connectivity, and ultimately the nature of their occupations — many low-skilled workers have not been able to make transition to remote online work.
Humans in the Loop is a social enterprise working to provide vulnerable people, including refugees, asylum-seekers and conflict-affected people across the Balkans and the Middle East, with easy and accessible job opportunities. Founded in 2017 by Iva Gumnishka, Humans in the Loop has established an exemplary model of social and economic inclusion of the low-skilled through ‘impact sourcing’. Workers receive training and remote work tasks which they can perform from home on a flexible schedule. The training programme is designed to build basic computer literacy, upskill workers gradually and does not require prior knowledge of the English language or computer skills. The official job title is “professional human in the loop,” someone who specialises in helping teach AI to emulate human thinking.
Talking about her inspiration, Gumnishka says that the company is conscious of the thin line between empowering workers through accessible job opportunities versus creating an environment for low-paid invisible microwork: “We have checks and balances in place that ensure that our workforce is remunerated in a dignified way and it has clear pathways for professional development in the future.” <1>
In India, one of the first proponents of impact sourcing, iMerit, has provided employment to more than 2500 underprivileged individuals since 2012 in their centres in Ranchi, Shillong, Vizag and Kolkata. iMerit employees do data labeling for drones in the agriculture sector, for medical imaging and for e-commerce companies.
Lack of off-site resources is the first obstacle the companies had to combat as Covid-19 pushed people to stay at home.
iMerit CEO Radha Ramaswami Basu says, “Advances in AI technology are sometimes seen as a threat to jobs, but iMerit has demonstrated that the opposite is actually true. Humans are a much-needed component of the AI pipeline and, in fact, it is safe to say, without human involvement, AI is not deployable.” <2>
While enterprises like iMerit, HITL etc. provide a short-term solution to the skills crisis, the road to employability of the low-skilled has its own challenges. Lack of off-site resources is the first obstacle the companies had to combat as Covid-19 pushed people to stay at home. Talking about the preparation against Covid-19 to keep the company up and running, Basu said: “The biggest shift involved changing our delivery process from 100% on premise to 100% work from home. This change involved iMerit working very closely with our partner-clients around the prioritisation of jobs and issues related to data security. Now, that it is in place, the company is in a better position to deal with any obstacles that may come its way.”
Here, the Bulgarian model had to make smaller adjustments as most workers worked from home already: “We were among the lucky companies which were not affected by the Covid-19 crisis. In fact, it proved the resilience of our model even more, since our workers were already working remotely. All of our platforms and tools were already enabled for remote work which has been at the core of our model since early 2019,” Gumnishka shared.
Most people working for labeling jobs remotely usually view it as an additional source of money, rather than their primary employment, unless there are regular projects coming their way.
However, like any other tech-centered solution, this too comes with its own negatives. At an individual and organisational level, lack of regularity in work and pay is a major barrier for those seeking make a living on data labeling. The pay is irregular with people being paid on a project-to-project basis. At an international level, there are sharp divides in basic digital literacy, access to technologies and related infrastructure, between the developed and developing world. According to the International Telecommunication Union, the percentage of internet users is much higher in developed countries than in developing countries where four out of five persons are still not able to avail of the benefits of being online. In the Asia Pacific region Japan and South Korea account for 70% of broadband users, while accounting for around 4% of the region’s population. At the same time, living in a developed nation does not guarantee equal access to internet either: broadband access in the US costs USD 68 per month and with few competitors in the market, people are not left with much choice. Moreover, most people working for labeling jobs remotely usually view it as an additional source of money, rather than their primary employment, unless there are regular projects coming their way. Therefore, the opportunity though promising in the short term is essentially limited in its long term impact and therefore its inclusivity.
Developing economies with large informal sectors can take advantage of the ‘impact sourcing’ model to create a digital economy where the low skilled are essential.
“As we are thinking about what people refer to as ‘the new normal,’ we must shape it in a way which is inclusive for lower-skilled workers and gives them access to the resources that enables their resilience in difficult times and gives them access better opportunities in the future.” Decrease in employees in one sector can cause demand for employees in another: this is the opportunity AI will create. If leveraged the right way, the growing prospects of the AI industry can empower the under-skilled. Covid-19 is changing the world: we must ensure that this change is for the better.
<1> Iva Gumnishka, interview by Namrata Yadav, 15 June 2020.
<2> Radha Basu, interview by Namrata Yadav, 8 June 2020.
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.