The rise of AI is leading to remarkable changes in society. Many people are concerned about AI’s influence on job opportunities as the technology continues to advance and automate tasks once performed by humans. While there are definitely consequences to the pervasiveness of AI, this new development can actually create and foster new jobs for many. 

Data Annotation and Global Poverty

The development and accuracy of AI are heavily dependent on its training data. However, before this data is fed to AI for training, it needs to be labeled or annotated with the necessary context. This process has led to a new occupation known as data annotation or data labeling, in which individuals review raw data and label it with the context needed by the specific AI model. 

These annotations include outlining specific objects in pictures so that AIs know to pay special attention to that item or explaining the semantics of a word or phrase that could only be understood colloquially. The overall range of annotations depends on the model’s use. However, these examples showcase the wide range of responsibilities and the need for data annotators. 

Furthermore, human annotators provide nuance in their work that computers lack, which helps make AI models more accurate. This job is traditionally outsourced to countries in Africa and Asia, where populations live in impoverished communities. These data annotators work in poor conditions, with their workplaces even being referred to as “digital sweatshops.”

However, the tides might turn in favor of these communities with the right business practices and national policies. This is because AI companies want more qualified individuals to step into this role and annotate academic content. This shift in demand indicates the potential for AI to address economic poverty in areas with large populations of data annotators.

Impacts

Many large technology firms routinely outsource jobs to countries with highly skilled but undervalued workforces. Data annotation represents just one segment of the broader outsourcing trend within the Western AI industry. This fosters economic prosperity, benefiting the countries receiving foreign investment. 

On that note, data annotation is a new sector these countries can capitalize on for proper foreign investment. Many countries have skilled and knowledgeable human capital working in “lower-level” jobs, such as data annotation, due to a lack of opportunities in their countries. However, with the aforementioned shift in data annotation, these same individuals could easily qualify for “higher-level” positions, demand higher wages and advance professionally.

Sama’s Role in Africa’s AI Labor Economy

Sama is a subcontracting company hired by major technology firms to source and manage data annotation work in Africa. Through this model, the company has become a key part of the AI supply chain, connecting global tech companies with large workforces that label and process the data used to train artificial intelligence systems. Sama has helped lift more than 59,000 people out of poverty since 2008.

Its client and partner network includes companies such as Microsoft, Walmart, Getty Images and other AI-focused firms seeking large-scale human annotation services. 

Final Thoughts

Due to the rigorous nature of data annotation, the labor market is expansive. By leveraging current business practices, outsourced workers can capitalize on the shift to make this job more lucrative and even on par with other, more traditionally skilled occupations. This would lead to a holistic alleviation of poverty in local communities, as companies provide new opportunities to impoverished populations without losing the benefits of paying less for the same skill set.

Additionally, if national governments enact policies that both attract foreign investment and protect business practice standards, data annotation could become a powerful force in lowering global poverty and empowering international trade.

– Saanvi Mudpa

Saanvi Mudpa is based in Seattle, WA, United States and focuses on Technology and Solutions for The Borgen Project.

Photo: Unsplash