Opinion: Done right, AI in global development offers vast opportunity

Artificial intelligence is revolutionizing industries, and its potential in global development is remarkable. Beyond being a buzzword, AI is a dynamic tool that can foster innovation, advance equity, and shape the future of global problem-solving. However, it is crucial to understand both the benefits and potential pitfalls of AI to maximize its value and use it correctly. The value of AI in global development In a world that is constantly changing and overwhelmed by information that surpasses human comprehension, AI serves as a force multiplier and a valuable ally. It can analyze massive datasets, identify patterns, and provide insights in a fraction of the time and cost that humans require. The World Food Programme's SKAI initiative is a compelling example of AI's potential. SKAI has been trained by using satellite imagery from three past disaster events: the 2010 Haiti earthquake, the 2017 Santa Rosa wildfire, and the Syrian conflict in 2016. In each of those training cases, SKAI achieved greater than 80% accuracy in identifying damaged buildings, speeding response, and prioritization. AI's capabilities, such as digitally twinning supply chains, personalized disaster modeling and response, and customer feedback modeling, are transforming disaster management strategies. AI's potential impact in global development extends beyond immediate crisis response: • In war-torn regions like Ukraine, AI could play a crucial role in modeling, planning, and executing reconstruction initiatives. • In agriculture, AI can scan crops for pests and assess hydration levels, potentially increasing yields significantly. Initiatives such as HungerMap LIVE uses artificial intelligence to supply what have been dubbed "nowcasts" — virtually real-time, granular estimates of the food security situation. • In health care, AI is invaluable, facilitating early health diagnostics, organizing health data, and enabling telehealth services, particularly in underserved areas of Africa and Asia. • For monitoring and evaluation, AI is transforming access to traditionally paper-bound knowledge, such as project reports and M&E information, making it more accessible. • AI also reduces bureaucratic tasks day or day or in emergencies. For example, AI document processing helped automate the ingestion and processing of passport and visa applications for NGO’s during the Afghan evacuations. AI can assist in various administrative tasks, including requests for proposals and solicitation creation, speech writing, report compilation, and social media management. By reducing the workload for global development professionals and NGO staff, AI enables them to focus on more critical tasks. Crucially, the emergence of AI is not about replacing jobs but augmenting human capabilities. AI allows local organizations to access best practices, industry norms, and refine their proposals, making them more competitive globally. Language models like Bard and ChatGPT are the new “ground-level” of public information access. When combined with a development organization's own internal "AI Sandboxes," AI enables development organizations to unlock the wealth of knowledge within their walls, experts, and history, creating internal super brains to sort and help supply their best development collective intelligence. Embracing AI: A moral imperative? Development professionals are often swimming in a global ocean of research, pilots, tests, and trials, while simultaneously existing in a bandwidth desert. The World Bank studied the demand and use of its policy reports, and found that 87% of them go uncited while 31% of them are never even downloaded. Optimizing how we can be permeable to all of this knowledge, while accessing it in the most efficient way is a challenge. It is a moral imperative to harness the best information and evidence when applying public or entrusted resources, and that’s where AI has potential to improve the effectiveness of development work. AI can serve as an ultra-efficient “intern,” gathering and organizing information, although this assistance doesn’t absolve the responsibility of development professionals to scrutinize, refine, fact-check, and determine how to deploy this information as a component of their final output. A human-in-the-loop approach is crucial, where AI provides options and insights, and humans then take informed, experienced decisions. An example where AI complements highly trained and specialized experts can be seen in the United States Patent Office, where AI assists patent officers in research and gathering information, saving time for evaluating the merits of patents. Similarly, as development professionals, we also adjudicate resources and our roles by making evidence-driven decisions. Potential pitfalls: bias and misuse AI's power is undeniable, but it also carries risks, such as bias. AI reflects the biases present in the data it is given to learn from, as well as the directions it is given on how to organize that information, amplifying existing biases in our practices, norms, and historical perspectives. Identifying and addressing these biases proactively is essential. However, it is crucial to recognize that bias is often a by-product of experiences and expertise, shaping best practices and standards. Nevertheless, continuous vigilance is necessary to ensure that AI aligns with ethical principles. Moreover, AI can produce diverse and varied results based on individual interactions, similar to different interpretations of a book report. This variability is a feature, not a bug, nurturing a more rich and nuanced application of AI that, when paired with unique perspectives, can lead to diverse and innovative outcomes. In my volunteer work providing free AI training to development organizations and leaders, I still see too many leaders forming their opinion of AI based on the last three headlines they have read rather than by actually using the tools. It’s common to see unique social media snapshots of AI foibles passed around instead of teams and leaders taking the time to explore the tools, pressure test them, discuss them, and thoughtfully consider their impact and how to use them appropriately. While this is not unusual when new technologies emerge, AI is moving too fast to ignore, and a critical pitfall is development leaders assuming that AI either has no relevance to them or dismissing it simply as dangerous. The future of AI in global development AI offers enormous potential for global development, from disaster management to health care diagnostics and beyond. However, as we embrace this technology, it is crucial to understand its benefits, limitations, and the need for ethical oversight. By harnessing AI responsibly, we can reshape the global development landscape and make substantial strides toward a more equitable and sustainable world. Navigating the impact of technology, including AI, is a challenge that requires understanding and intentional direction of its potential. It can be seen as a threat or as a tool and ally to navigate the exponential age we find ourselves in. Author’s note: Any opinions expressed are mine in my personal capacity.