Nethope Logo

Key takeaways from the AI discussions at the NetHope Global Summit 2019

Last month we had the opportunity to host several discussions about Artificial Intelligence (AI) and Machine Learning (ML) at the NetHope Global Summit 2019, held October 21-25 in Puerto Rico.

October 30, 2019

Above: AI discussions during the NetHope Global Summit held October 21-25, 2019 in Puerto Rico

By Leila Toplic, Lead for Emerging Technologies at NetHope, and Steve Hellen, ICT4D Director at Catholic Relief Services


Last month we had the opportunity to host several discussions about Artificial Intelligence (AI) and Machine Learning (ML) at the NetHope Global Summit 2019, held October 21-25 in Puerto Rico.

Here are our takeaways on four key questions:

 width=Where do we see the potential of AI to deliver value and what will it take to realize that value?
  1. AI/ML apply to all 17 Sustainable Development Goals.

    Based on the early practical implementations in our sector, AI/ML can help us make decisions and act faster in emergencies, reach more people (e.g., refugees) with services and information they need, give us insights to predict and possibly prevent the worst from happening (e.g., disease outbreaks, famines, poaching of wildlife).

  2. While it’s early days for AI in social impact space, we're starting to see the potential for AI to support every aspect of our work – from field programs (health, education, poverty alleviation, conservancy) to powering digital transformation within the organizations by helping us improve processes, create efficiencies, increase reach and effectiveness.
  3. Doing good better with AI will be manifested in both improvements to existing programs and processes (e.g., automating tasks in data management) and in the creation of new solutions like TESSA chatbot or PAWS.

     width=What are some of the challenges that humanitarian and development organizations are facing today when it comes to AI?

    • Capacity & Resources – AI is new for our sector and most nonprofits lack the capacity and expertise to evaluate, develop, procure and use AI-enabled solutions. There’s also a gap in resources: finding a donor who wants to spend money on new solutions and pilots; finding technical experts to help with initial model development; and AI/ML tools and services that don’t require specialized expertise.
    • Data - Data is a crucial component of any AI solution. Machine Learning algorithms learn from data so it is important to have sufficient amounts of high-quality data for the problem you want to solve. Getting "sufficient amounts of high-quality data” is a massive challenge for our sector, with issues across the whole journey of getting the data ready for AI – from data collection and cleansing to data management.
    • Sustainability – Many of the early practical implementations have come out of cross-sector collaborations: NGOs working together with tech companies and/or academic institutions who have AI/ML technical expertise and can help with the initial model development. Transitioning from the pilot phase where model is developed by a partner outside of the sector to scaling and sustaining where NGOs own re-training of the model and refreshing the data can be challenging and require new resources and processes.
    • Ethics - The challenge facing NGOs today is to maximize the positive benefits of AI, while preserving human rights and privacy, and protecting people from any negative impact of AI. Ethical considerations need to be embedded across all touch-points - the teams we hire, the data we use, how we frame the problem, how we develop and implement the solution, and whether we use powerful AI capabilities that could cause harm.

     width=How is NetHope responding to the needs of the nonprofit sector and shaping the adoption and use of AI?

    Following the working sessions at the 2018 NetHope Summit, NetHope's Emerging Technologies Initiative was set up as a sector-wide approach to evaluating and integrating artificial intelligence (AI), blockchain, and other emerging technologies in humanitarian, development, and conservation work - and a collaboration mechanism for NGOs to partner with academic institutions, tech companies, and others in the AI ecosystem.

    With the Emerging Technologies initiative, we’re focusing on increasing NGO’s internal expertise and capacity to evaluate, develop, procure, and use emerging technologies to further their impact.  These capabilities help organizations make informed decisions, do their work better, anticipate issues that might arise from technologies like AI, and ensure that people in need are aware of the systems that affect them and their communities.

    We believe that nonprofits have a responsibility to understand AI well enough to know what questions they should be asking when evaluating the need for AI in their work and its effects on the outcomes.

    The AI Working Group is part of the Emerging Technologies Initiative and focuses on capacity building, programs, and toolkits and standards related to AI.

    At the 2019 Summit, we focused on four needs that our NGO members have related to AI:

    • Need to advance the understanding of the potential risks of applying and using AI and how to implement AI responsibly and ethically. In the Ethical, Responsible AI working session, we examined a practical example from our sector that highlights risks, issues, and path to achieving ethical AI, and began gathering input on principles, criteria, and questions that will help inform a practical guide for ethical AI international development practitioners.
    • Need to get started without having specialized expertise in AI and ML. There are a number of off-the-shelf AI /ML tools and services from companies like Amazon, Google, Microsoft, and Salesforce that NGOs can start exploring and experimenting.
    • Need to learn from practical implementations of AI in our sector - both field programs and internal operations – and reuse & scale what works. Plan International and The Carter Center, NetHope members, shared how they are using AI in their work, including lessons learned across all stages of the process - from how to frame a problem for AI and get your data in order, to how to resource teams and have the right processes in place.
    • Need to evaluate AI/ML for the specific problem statements that each of our members are focused on with the support from tech experts from Amazon, Google, Microsoft, Salesforce, and USAID, and using a framework of 32 questions which was developed by NetHope’s AI Working Group in collaboration with partners.

    What can nonprofits do today?

    • Learn about AI/ML so you can make informed decisions - IF and how to use AI in your work.
    • Start with the problem you want to solve, not with technology.
    • Get your data in order. AI requires large amounts of high quality data.
    • Learn from other AI implementations in the nonprofit sector. Reuse what works.
    • Try something small first.
    • Understand ethical, responsible development & use of AI.
    • Partner for expertise & resources. Collaborate to achieve a greater impact.
    • If you are a NetHope member, join NetHope’s AI Working Group


    Twitter @NetHope_orgFacebook NetHopeorgYouTube iconLinkedIn Company NetHopeInstagram nethope_org