How Artificial Intelligence Can Benefit Nonprofits

Canadian nonprofits have always faced challenges, funding shortages chief among them. Cuts to federal and provincial government spending delivered a blow to the sector in the 90s, and further difficulties arose during the financial crisis of 2008. Lasting impacts include the demise of a large number of charities and insufficient manpower at others.

 

Many nonprofits that continue to thrive do so thanks to the creative problem-solving skills of leaders, employees, and volunteers. Of course, coming up with creative solutions is only half the battle, and the easy half at that. Difficulties with organization, communication, and workflow can arise at any stage of a project. Nonprofits are particularly vulnerable to these problems, as the time and talent they can devote to planning are limited.

 

Luckily, the widespread adoption of cost-effective technologies – including social and wireless communication channels, data aggregation and management software, and others – have begun to level the playing field. Digital platforms and tools can go a long way in improving organizational productivity; those struggling with scarce resources may see the most substantial benefits.

 

 

Workplace Inefficiency and the Potential of Artificial Intelligence

 

One type of technology – artificial intelligence (AI) – shows a lot of promise with regards to saving time and improving productivity. AI is also rousing a lot of polarizing debate. Thought leaders like Stephen Hawking, Bill Gates and Elon Musk have voiced concerns regarding its potential dangers. Dystopian editorials describe mass job-loss at the hands of machines that can outperform human beings. Are these concerns really warranted? It’s probably too early to say. We can, however, assess current technologies with AI capabilities, which help users carry out tasks informed by their own knowledge and experience in a more efficient way.

 

A recent Wired article explores the areas of our daily lives that are already affected by AI. Examples include “email spam filtering, Google search predictions, and voice-recognition”. For a growing number of people, virtual assistants – software agents capable of performing a variety of tasks, from issuing personal reminders to posting on social media accounts – can be added to this list. What these technologies have in common is “machine-learning algorithms that enable them to react and respond in real time”. Their ability to “learn” and “respond” enables these technologies to perform human tasks that require low cognitive functioning. Used in the right combinations, AI-enabled tools can make a large overall contribution to the productivity of a workplace.

 

Ironically, the data that has made these technologies possible has also been a huge source of inefficiency. If it weren’t for the widespread use of email (especially in the office, where its an obligatory form of communication), we wouldn’t need spam filters. Consider a statistic from the 2011 book, “Overload! How Too Much Information is Hazardous to your Organization”. According to the book’s author, Jonathan Spira, it takes five minutes to get back on track after a 30-second interruption. When one considers the volume of online junkmail most people receive, the value of these filters becomes apparent.

 

The data inundation described in “Overload” may have a greater-than-average impact on the nonprofit sector. Spira states that “58 percent of government workers spend half the workday filing, deleting or sorting information.” Many employees and volunteers at nonprofits carry out the same tasks; however, due to limited resources and pressing community need, these individuals often face unusually-heavy workloads. These workloads will only increase as the use of big data – very large data sets that can be analyzed for useful patterns – becomes an essential part of many organizations.

 

Jacob Harold, president and CEO of GuideStar (a provider of nonprofit information), describes the potential insights that can be gleaned from analyzing data. In an an article for Skoll World Forum, Harold asserts that using big data could help nonprofits “track their activities, engage with their stakeholders and understand their context.” Unfortunately, many organizations don’t have the money to invest in the “cutting-edge information systems” devoted to these functions. For this reason, and due to other limitations, Harold suggests a commitment to medium data, which he defines as “structured information about who you are, what you’re trying to do, and what’s happening”.

 

Only time will tell which approach nonprofits adopt, but one thing is clear: the data they share – be it big or medium – will need to be filtered, organized, and assessed. The addition of AI capabilities to existent automated monitoring technologies could offer an affordable solution.

 

 

AI in Practice: Media-Monitoring

 

In the nonprofit sector, one already-common application for technologies that deal with data is media monitoring. During a recent series of interviews with nonprofit professionals, I found that tracking the news was a major priority for most. Whether they’re following the social impacts of community initiatives or monitoring relevant public opinion, nonprofits are investing resources (human and, when possible, financial) into staying informed of media activity. This is especially true of organizations that are called upon to respond to news stories and events, as well those involved in educating the public.

 

Like big data, online media is accumulating more rapidly than ever. Unfortunately, as the production of online content by both mainstream and amateur journalists increases, the amount of time professionals can spend analyzing content decreases. There exists a clear need for software that can do more than simply recognize words. Imagine a program that, instead of simply calling your attention to an article that contains a word you’re tracking, reads the article for you, then sums it up in a single sentence. Imagine how much time you could save if this program learned your preferences, then told you how you’re likely to feel about each article that matches your search criteria. Enter artificial intelligence.

 

To illustrate some of the benefits of AI, I used Gnowit’s media-monitoring platform to carry out a keyword search inspired by the nonprofit sector. As there are many Canadian nonprofits devoted to issues affecting women, I chose subjects that matter to these groups, typing “domestic violence” and “status of women” into the system’s search fields. While examining the results, I paid special attention to insights provided by AI-enabled features.

 

According to Gnowit, 309 articles containing my keywords had appeared in the media during the previous two-week period. Unsurprisingly, 88% of the articles uncovered were judged to be negative in tone (there were many results related to domestic violence). Those perceived as positive described “big advances” and “big improvements”, two of the key phrases extracted from the overall pool of results. The formation of these analytics was dependent upon the system’s assessment of each individual piece of content. A summary for one of the articles appears below.

Single Article Snapshot

 

As you can see, the system’s AI component judged the sentiment of the article, identified the sentence within it that best sums up its content, and extracted the core topics it contains. These assessments – carried out for each of the 309 articles found – occurred in an instant, at the click of a button. I also had the option of changing the sentiment attached to each article in order to train the system to understand what I consider to be positive, negative, and neutral.

 

 

The Future of AI

 

Much like technologies that help analysts understand big data, AI-enabled monitoring software can provide insights into high volumes of dense, long-form content. This is noteworthy because, though social media is starting to hold more weight in many fields, traditional media – which generally publishes long-form content – has retained its widespread distribution and authority. Given its capacity to deal with text of varying lengths and levels of thematic complexity, AI offers the most promising solution to the problems associated with monitoring traditional media.

 

But AI is more than just the future of media monitoring. Applications for technologies that can extract, learn from, and act on patterns in vast quantities of data will grow in the coming years. In addition to increasing workplace productivity – which can help organizations do more with limited manpower – AI can help nonprofits find and understand trends in human behaviour. The true benefits of these insights will occur within the communities these organizations serve.

 

 

The Offer:

At Gnowit, we admire those who devote themselves to making Canada (and the world) a better place. We want to help nonprofits achieve their goals. For this reason, we are offering 6 months of free use of our media-monitoring platform to the first 20 organizations that contact us at support@gnowit.com. We hope those who take advantage of this opportunity will gain a great deal of insight through the experience.

Feature image by Saad Faruque

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply