Study on GenAI’s Impact on Labor Market Stability in India

Experts from various fields discussed the impact of Generative AI (GenAI) on the Indian workforce at the Digital Futures Lab on June 21 and 22, 2024. Speakers discussed the impact of GenAI on various employment sectors, including writing, research, unorganized work, and agriculture. During the discussion, participants also spoke about the need to consider the impact of GenAI on productivity, efficiency, and job replacement, calling for greater accountability of AI companies for the possibility of their products being used for illegal purposes.

Use GenAI to increase productivity:

Speakers discussed the impact of GenAI on editing and curation work, particularly in the area of ​​data management. One speaker spoke of how trainees were required to perform simple tasks that, while useful in practice, did not fully utilize human potential. Now AI models such as Chat-GPT have demonstrated the ability to perform such tasks more efficiently, eliminating human error and reducing training requirements. This shift encourages companies to use AI for tasks such as coding, data scraping, and other deterministic processes.

But in doing so, GenAI also raises the bar for entry-level intellectual work, the speaker said. Companies are increasingly emphasizing AI skills in their hiring processes, although there is some confusion about what constitutes sufficient AI training. The impact on employment is unclear, as the number of internships has not changed noticeably. But there is a growing skills gap, most clearly seen with companies like TCS saying they will train people to use GenAI.

Another speaker claimed that focusing on AI-generated content has led to a 30% increase in productivity. The speaker also mentioned that some countries, such as the Republic of Korea, are working on a robot tax to prevent humans from being overhauled in favor of generative AI. The discussion also highlighted the need for clearer definitions of AI skills to prevent fraudulent claims of expertise based solely on basic prompting skills.

Increasing productivity will impact working hours:

Referring to the discussion on delegating secondary tasks to GenAI, one of the speakers stated that the reduction of work should also take into account its impact on time.

“If we don’t shorten the hours while also reducing work, it will lead to unemployment,” the speaker said, emphasizing that the global South should step up its demands for shorter working hours and how that time should be invested. Speakers suggested that AI should work with union leaders who are not as tech-oriented.

GenAI redirects workforce towards grassroots work:

Another speaker said that in rural areas, AI is enabling women working in their native languages ​​to do business and create new opportunities. Partnerships are emerging with universities to match student credits with public sector job opportunities, indicating growing interest in AI among both students and consumers.

Another observation was that students who could not boast of AI skills on their CVs were turning to NGOs, diverting a significant number of students from advanced universities to work in NGOs with salaries of 3-4 lakhs. This trend suggests that AI is changing the job market, creating new niches for educated workers in the non-profit sector, while potentially displacing traditional roles in larger corporations.

One speaker spoke about how GenAI has the potential to positively impact employability and career prospects in India. The speaker gave an example of how AI tools can help women return to work after giving birth. Another example was the beauty industry, where workers typically earn between INR 60 and 150 crore. Using GenAI, women could increase their productivity by better understanding skin tones, hair types, and product suitability. Similarly, AI can increase their ability to customize services, potentially leading to improved earnings and work efficiency.

GenAI can help overcome training and skills gaps:

One speaker spoke about how GenAI can help with training and skill development and overcome barriers like language gaps. They gave the example of AI Whisper, which helps people with no experience with writing or technologies like Google Translate and voice-to-text overcome language and literacy issues.

The speaker continued by talking about GenAI’s potential to break down silos in training systems, suggesting the need to consider and budget for these technologies in educational planning. However, the use of AI in training raises important questions about human development, pay structures, productivity rates, and efficiency. The discussion also raised concerns about how overreliance on AI could be problematic in developing countries like India, potentially crowding out competition or exacerbating existing inequalities.

Moreover, the pressure on women to increase their productivity through AI-assisted remote work may reinforce patriarchal structures rather than alleviate them.

What impact will GenAI have on women?

While some speakers talked about how GenAI can ease the burden on some women, another speaker pointed out that jobs in sectors like textiles, apparel, and food (packaging) are currently being automated. Most of these jobs are held by women. In this regard, the speaker asked others to consider the impact of technological advancements on women, especially considering that most women in India do not own electronic devices and have to share them with their children.

“Are we prepared to discuss the impact of AI on women? How will AI be embraced by women? How do we give them access to this discussion?” the speaker asked.

Why Farmers Reject Agritech?

During the discussion, speakers talked about the use of latest technologies in sectors like agriculture. One of the speakers said that farmers actively refuse to use technology because they have been historically cheated many times. The speaker said that while Agritech tells farmers what they already know, it does not provide them with a guarantee of alleviating uncertainty. Farmers are still wary of digitalization measures due to lack of transparency, unfair prices etc. and the inability of technology to increase the overall income. Hence, the speaker said that there is a need to look at GenAI in this context.

Speakers demand accountability from AI companies:

Referring to ride-sharing platforms like Uber, which initially claimed to be merely aggregators, one speaker said AI companies could face similar scrutiny and legal challenges. Just as Uber was ultimately held liable for its drivers’ insurance and safety, there is a growing belief that AI companies should be held liable for the content their technology enables or produces. The speaker gave the example of how a startup person working on an AI tool predicted that the tool would likely be used for illegal purposes in the case of pornography, but could not be held liable for that.

The speaker argued that social media platforms are currently facing increased regulation and that a similar trend could emerge for AI companies. A key proposed principle is that these companies should implement robust discriminators or filters to prevent the spread of problematic content.

The algorithm should be considered in the context of artificial intelligence:

Another speaker noted that the navigation algorithm sometimes plays a bigger role in allocating work than the AI. Therefore, there is a need to audit these algorithms, even if the AI ​​is not in the system, because the algorithm is used as a lever of control.

Read also: