4 ways to bridge global inequality around emerging technology

This article is brought to you thanks to the collaboration of The European Sting with the World Economic Forum.

Author: Sreevas Sahasranamam, Professor, Adam Smith Business School, University of Glasgow

• Emerging technologies like AI could help solve global challenges and meet the UN Sustainable Development Goals.
• However, technology isn’t sufficiently concentrated in the developing economies where it is most needed to solve pressing challenges.
• Four key actions are needed to address this technology inequality.

One of the key narratives coming out of Davos 2024 was around how AI and other emerging technologies could help solve global challenges and meet the UN Sustainable Development Goals (SDGs). However, there is an inherent dichotomy within this debate: A lot of the cutting-edge research and intellectual property around these emerging technologies is concentrated in developed economies, particularly among the elite. Meanwhile, the pressing needs for the social innovation these technologies are looking to solve are in bottom-of-the-pyramid (BoP) settings whose voices are unheard in technology development.

This poses issues such as lack of access and affordability to technology, language barriers, under-representation of certain datasets in training (in the case of AI), and lack of contextualization to local settings.

The question remains of how to bridge such global inequalities around emerging technology. There are four potential mechanisms for handling this:

1. Increase exposure to emerging technologies

BoP communities have very little access and exposure to emerging technologies, let alone know how to realize value from them. So, one needs to start with the exposure challenge. This could be achieved by focusing on their integration into the global value chain of emerging technologies. A good example in this regard is Karya, a startup that is building AI datasets in vernacular Indian languages by employing rural communities as data creators. The tasks could be as simple as just reading and recording a piece of text on their mobile. The voice data created is subsequently fed into the LLMs of Microsoft, Google and others. This not only ensures improved income and livelihood, as they are paid much higher than the minimum wage, but also gives these communities both a literal and symbolic voice in the global AI narrative.

2. Build in contextualization

BoP community contexts are very different given the lack of resource availability and local social structures, so technology needs to be contextualized to these settings. For example, agriculture in India is characterized by marginal landholding, strong community bonds and low digital penetration. Supplying blockchain technology to improve traceability in supply chains for produce such as coffee and rice, TraceX adapts to such settings through the use of offline maps to allow for poor internet connectivity, Geographic Information System features that allow real-time data annotation of farm boundaries, and replicating existing social networks digitally within the value chain.

Similarly, in an AI setting, LLMs currently are predominantly English-language-optimized, so text in Hindi needs three to four times the computations for an AI chatbot to answer a question compared to the same question in English; this makes it expensive and reduces access. To solve this, startups like Sarvam AI have created Devanagari-optimized LLM that reduces the computations and thereby the costs.

3. Co-create innovation ecosystems

In the nascent phase of emerging technology development, standard operating procedures are still evolving, and there is fuzziness around outcomes. In developing economies, there is a double disadvantage that emanates from their weaker institutional environments relative to developed economies, such as weak intellectual property rights, limited resources per capita, and high degrees of information asymmetry. Recent research studying the development of biomedical devices in India finds that to navigate through challenges of emerging technology development in developing economies, there is a need to co-create innovation ecosystems, wherein the key actors leading the emerging technology development need to not only develop their own product or business, but also need to co-create a wider structure around it, collaborating on activities and building resources. This biomedical innovation ecosystem co-creation led by Sree Chitra Tirunal Institute for Medical Sciences and Technology has created biomedical devices of global repute such as low-cost heart valves, oxygenators and blood bags through frugal approaches.

4. Develop bottom-up social practices

The power structures in BoP communities oftentimes lead to hierarchical impositions of decisions and technologies. When this happens, people lack a sense of ownership of the technology and remain disengaged with it. Instead, recent research points to the need to focus on bottom-up social practices around emerging technology engagement with a thrust on dialogue and multistakeholder participation. For instance, Open Healthcare Network, created in India during COVID-19, followed such an approach wherein technology professionals, medical doctors and government administrators self-organized themselves to build digital tools that would help in handling the pandemic. There was no upfront blueprint for the tools that were needed, so the group created solutions through a process of collective dialogue, feedback and experimentation. These tools are still being used beyond the primary use case of the pandemic for intensive medical care support in remote areas as part of the 10BedICU initiative.

DISCOVER

How is the World Economic Forum helping farmers with technology?Show more

To bridge global inequalities around emerging technology, we need to democratize its development by lending an active voice to BoP communities, integrating them into the value chain, contextualizing the technology for local settings and adopting a multistakeholder, co-creation approach.