Continuing the conversation on the extent of AI’s permeation into the fabric of our society, this second blog in the series is focused on how AI influences management decision making in organizations in general and in particular, AI’s influence in management funding decisions at private donor organizations in the aid/humanitarian sector. Generally, AI has transformed the way organizations operate, and more importantly, in which organizations make decisions. The focus is on private donor organizations because they are devoid of other pressing factors that usually dictate stances of many organizations – the bottom line. Private donor organizations fund a number of programs, projects and tools, that permeates all aspects of society. Therefore, their funding decisions are critical to the discourse on AI’s permeation into our lives. But do we know if AI has made a dent into the private donor organizations world yet and if so, how has it influenced their funding decision process?
For private donor organizations (PDOs), embrace of AI’s role in management funding decision-making has gradually been surging. Currently, extant literature show that PDOs fall into three main spectrums with regard to AI in general management funding processes in the course of their humanitarian work.
- First, those who have fully embraced AI and have automated most of their funding processes. Some researchers [1a, 1b] cheer this position for full embrace of AI and posit that AI’s promise of fast, accurate, repeatable, and low-cost decisions, with quality approaching human-like intelligence, could be an important driver to and full automation of decision-making and for even incrementally more accurate, and significant advantages over time.
- Yet, there are PDOs, including notable large ones who are at the extreme end of that spectrum with cautious embrace of AI. They too have their advocates [2a, 2b] who have cheered their concerns about the possibility of a total abdication of the human intelligence (HI) factor to AI in humanitarian service. As one researcher puts it, “human cognitive tasks can seemingly be automated by AI, but we risk a loss in predictability and explainability when doing so” [2a].
- However, the remaining majority of PDOs use some basic form of AI in funding related decision-making. Mostly, AI is used for up-to-date information, analyzing data trends, coordinating data delivery and developing data consistency [3].
The new aid agenda and the influence of AI on funding decisions may affect donor trends for years to come [4]. As this new aid agenda now moves the spotlight of attention from the quality of recipients’ sectoral policies to channels of ‘dialogue’ through which PDOs can have: (a) inputs into recipient decisions; (b) provide a wide range of technical and administrative support; and (c) urge policy choices that are more evidence-based than has hitherto been the case [4], it is not surprising that there are competing speculates of the influence of AI in funding decisions at PDOs.
In spite of the argument for or against the influence of AI in funding decisions, these competing speculates agree on one thing – that certain bottlenecks must be overcome for the potential of AI to be fully realized, whether used by PDOs or any corporate entity in any sector of society. Review of evidence suggest that four major bottlenecks need to be addressed in order for AI’s benefits to be maximized and appropriately leveraged in any management decision making and particularly in funding decisions that support social good such as those made by PDOs.
These bottlenecks are (a) bias and fairness, (b) privacy, (c) security and safe use and (d) ‘explainability’ which is the ability to identify the feature or data set that leads to a particular decision or prediction [5] . Concerns about the lack of adequate, quality, and integrated knowledge-based data repositories to feed AI’s algorithms that would ensure unbiased, secure, and explainable output to support PDOs evidence-based funding decisions [6]. A recent examination of a facial recognition software showed an error rate of 0.8 percent for light-skinned men and an error rate of 34.7 percent for dark-skinned women [5]. Other concerns have been raised regarding privacy, ownership and security of data collected by and from AI tools. Think about it this way: who owns the data AI collects about? Who owns your image collected by AI’s facial recognition tools?
Consider a facial recognition tool being developed by Kimetrica [7] a social enterprise and funded by the United Nations Children’s Fund – UNICEF [8], to detect malnutrition in children around the world. According to Kimetrica, their facial recognition tool can help diagnose malnutrition quickly and efficiently particularly in the last miles of communities where medical help is lacking either due to conflicts or lack of accessibility, where it is usually difficult for care to reach children in need. Noting that the 2017 United Nations’ report on Progress towards the Sustainable Development Goals [9] estimates that 155 million children around the world under the age of five are deemed stunted in their growth as a result of poor or chronic malnutrition, some see this as an effective tool to support the UN’s Sustainable Development Goal to end hunger, achieve food security and improve nutrition.
Yet others have raised concerns about risks due to the massive data collection and storage of thousands of images of children who are most vulnerable. Another concern is with the ownership of such data. Who owns this collection of facial images of vulnerable, malnourished children? Is it the funder (UNICEF), tool developer (Kimetrica)? What about the governments of the countries where these images are taken and whose citizens are the children or the families of the children? Do the children themselves even have a right to their own image?
All things considered, AI’s influence in organizational management funding decision and in particular, donor organizations are inevitable as AI continues to charge forward, permeating every aspect of our lives including the humanitarian sector. So how can donors, in making their funding decisions, advance development and humanitarian goals through the projects they decide to commit funds to, while ensuring processes and systems remain inclusive and trustworthy of society?
What do you think?
This is part 2 of a four-series post on AI and the ongoing dialogue on AI’s current and future place in our society in relationship to human decision-making . Check out my previous post on AI’s influence in our personal lives: When AI Goes Out To Lunch Who Pays?
___________________
References
[1a] Metcalf, L., Askay, D.A., Rosenberg, L.B. (2019). Keeping humans in the loop: Pooling knowledge through artificial swarm intelligence to improve business decision making. California Management Review, 61(4), 84-109. doi:10.1177/0008125619862256
[1b] Agrawal, A., Gans, J.S., & Goldfarb, A. (2019). Exploring the impact of artificial Intelligence: Prediction versus judgment. Information Economics and Policy 47, 1-6. doi:10.1016/j.infoecopol.2019.05.001
[2a] Bolander, T. (2019). What do we lose when machines take the decisions? Journal of Management and Governance, 23(4), 849-867. doi: 10.1007/s10997-019-09493-x
[2b] Danielson, P. (2009). Can robots have a conscience? Journal of Nature, 457(7229), 540–540. doi: https://doi.org/10.1038/457540a
[3] Mackworth, A. (2005). The Coevolution of AI and AAAI. AI Magazine 26, 51–52
[4] Killick, T. (2004). Politics, evidence and the new aid agenda. Development Policy Review, 22(1), 5-29. doi: 10. 1111/j.1467-8659.2004.00235.x
[5] McKinsey Global Institute (2018). Retrieved February 24, 2020 from https://www.mckinsey.com/featured-insights/artificial-intelligence/applying-artificial-intelligence-for-social-good
[6] Martinez, M. (2011). Personalizable knowledge integration (Doctoral dissertation). Retrieved from ProQuest Dissertations & Theses Global. (904143213)
[7] Kimetrica. Retrieved February 25, 2020 from https://kimetrica.com/our-projects/
[8] United Nations Children’s Fund. Retrieved February 25, 2020 from https://www.unicef.org/innovation/
[9] United Nations. (2017) Progress Towards Sustainable Development Goals. Retrieved February 25, 2020 from https://www.un.org/ga/search/view_doc.asp?symbol=E/2017/66&Lang=E