Speculative Friction: Unintended Impacts of Generative AI in India

About Speculative Friction

Advances in generative AI (GenAI) are poised to transform society in unimaginable ways. Popular narratives in India highlight how it will unleash a new era of innovation and entrepreneurship, bridge India’s linguistic and digital divide, and improve public service delivery. With a large digitally connected population, plenty of technology talent, and robust digital infrastructure, AI is believed to have the potential to transform the lives of all Indians. Numerous new initiatives and partnerships are being forged across government, industry and civil society to help develop India-specific use cases that can realise this vision for the country.

However, the use of Gen AI also poses many harms and risks. Many of these can not be foreseen during development and deployment. We need new methods and approaches to anticipate and mitigate these harms.

GenAI models are known to produce sexist, racist and other forms of biased content. If these tools become widely integrated into our social infrastructure, we risk reproducing patterns of historical and structural discrimination.

Learning from the past years of technological innovation, we can safely assume that benefits and harms will not be evenly distributed across society. For example, communities from low-resource languages and/or on the wrong side of the digital divide risk being excluded from emergent gains in GenAI.

Moreover, if GenAI lives up to expectations and becomes the infrastructure across all areas of social life, the impact is likely to extend far beyond specific use cases. For example, companies are already developing personal AI agents that can do basic tasks, such as booking a doctor's appointment, and deeply personal tasks, such as finding a marriage partner.

How applications like these transform human agency, relationships, and norms cannot be captured by a benefit-and-harm framework - change is often subtle, and its impact can only be known in the aggregate in the long term. By that time, it is often too late to course-correct.

The Speculative Friction project aims to fill a critical gap in the current discourse surrounding GenAI in India, which predominantly focuses on the benefits and potential use cases.

Utilising participatory foresight methods and design fiction, the project will develop a set of illustrated stories that provide incisive commentary on the social transformations that GenAI is likely to herald in India over the coming years. The stories aim to reflect the unintended impacts of GenAI, especially in critical social sectors and interpersonal settings.