*This piece was co-written with the manager of Classy’s data insights team, Ben Cipollini.
Although the past 100 years have seen the most dramatic technological upheavals to life than in all of human history, the next 100 years is set to pave the way for a multi-generational leap forward.
From HBO’s new series Westworld to articles and advertisements all over the web, you’ve likely encountered Silicon Valley’s hottest new buzzword: artificial intelligence (AI).
As the field develops, it seems nearly every industry is trying to understand how any current and future developments might affect them. Many questions abound:
- Will it replace human jobs?
- Will it enhance our products and services?
- Can we use it to be more competitive?
- Can we use it to save more lives?
- Will it do more harm than good?
While it’s certainly an exciting time to witness such developments and ask such questions, it’s now more important than ever to understand what artificial intelligence actually means, and how it’s different from other industry terms (such as machine learning, analytics, and data science).
When we have a firm grasp of these concepts, we can more confidently think through any implications for each industry—social impact included. Equipped with a basic understanding, you can then determine the difference between truly valuable conversation and a buzzword that’s being thrown around.
In this piece, we introduce the concept of AI as it relates to other industry terms, consider recent developments in the nonprofit industry, and suggest unique ways it may be leveraged in the future.
The phrase AI likely conjures up some futuristic pop culture references: robots, androids, smart homes, operating systems gone mad….
While the true definition of artificial intelligence is closely tied to topics historically reserved for science fiction novelists, current developments being made in AI (and their business applications) are a bit more down to earth.
- Analytics are a set of statistical techniques that could be done on a calculator, that help summarize and understand trends in data.
- Computer science refers to the branch of science that “deals with the theory of computation or the design of computers.”
- Machine learning is a branch of computer science. It focuses on programming techniques that allow a machine to find hidden regularities in data to solve various types of problems, such as classification (e.g. is there a face within an image?), prediction (exemplified by the Netflix prize: based on movies I have rated, what other movies would you predict that I will like?), or even action selection (e.g. what chess move makes me most likely to win?).
- Artificial intelligence is a branch of computer science as well, and its current approach also fits under machine learning. Most generally, it deals with “the simulation of intelligent behavior in computers” or “the capability of a machine to imitate intelligent human behavior.” Today’s AI generally uses a programming technique called neural networks which can take very raw data, and by processing and reprocessing it, solve many kinds of tasks without human intervention.
- Data science is an umbrella term that encompasses data engineering (gathering, cleaning, and matching data), data analytics, machine learning, and data visualization.
In most cases, current artificial intelligence is just an advanced machine learning technique; these algorithms can classify objects in images more reliably than humans, understand (or predict) what you’re saying (e.g. Apple’s Siri, Amazon’s Echo technologies), and beat humans at extremely challenging games from Go to poker to a slew of games on the old Atari.
New systems aim to move beyond these developments: to generate images from words, translate conversations across languages, and even complete automated question answering.
The most desired and difficult goal—which is under active research within and outside of academia—is to use these same techniques to train a single system that can solve many unrelated tasks or, ultimately, any task. This is known as artificial general intelligence (you might have seen Elon Musk or Stephen Hawking in the news talking about this phrase). This refers to the state of intelligence in a machine when it can perform anything a human could intellectually. It’s also commonly referred to as strong AI or full AI.
While we aren’t yet clear if the latest innovations in neural networks will lead to artificial general intelligence, machine learning and artificial intelligence have uses for nonprofits, and serious implications for the social sector.
For example, imagine if you could better understand and even predict the future actions of your donors.
And imagine if you could use machine learning to improve your programs and have a larger impact.
Nonprofits, AI, the Future, and Now
As donors demand even more transparency from the causes they support, AI will help nonprofits to process and deliver more data insights related to their programs and impact messaging. This level of transparency will become the new norm, and nonprofits will be expected to operate at this standard.
And while AI will undoubtedly affect job opportunities for certain populations (think self-driving cars), it will also strengthen what nonprofits are capable of in terms of social good and create new jobs that were previously unimaginable.
In fact, in a survey of 1,896 experts completed by Pew Research Center, while 48 percent foresaw a future where AI would displace more jobs than it created by 2025, 52 percent believed humans would adapt to create new jobs and markets. The study commentary adds, “just as we’ve been doing since the Industrial Revolution.”
Many nonprofits are already witnessing this change and the powerful implications of AI when it comes to their impact. Below are some examples of organizations leading the way in applying machine learning to social causes like animal welfare, health, and human rights.
PAWS, an organization dedicated to combating poaching, is using modeling and machine learning to give park rangers the information they need to predict poachers’ actions and stop them.
Microsoft is also working to use artificial intelligence to identify patterns in land use from high-resolution images that allow nonprofits to more “effectively deploy conservation efforts for the greatest impact.”
AI even recently saved a woman’s life. IBM’s Watson (the name for their AI system) identified a woman’s rare form of leukemia. In order to do so, Watson looked at her genetic information and compared it to 20 million oncology studies. In 10 minutes. By analyzing tons of data at a speed that’s just not humanly possible, Watson allowed the medical team to effectively treat the patient in time.
Nonprofits and banks are using artificial intelligence to help put a stop to human trafficking. Machine learning allows the banks to quickly review large amounts of data for suspicious behavior, and with more accuracy than manual sifting.
DataKind is also working to leverage data to improve human lives. For example, they’re partnering with Microsoft to use data science to reduce traffic-related deaths in large cities like New York City and Seattle.
Tech and Nonprofit Partnerships Paving the Way
To ensure advancements in AI continue to impact the social impact industry, a recent coalition has formed to help facilitate shared discoveries and best practices. Tech giants like Google, IBM, Microsoft, Facebook, and Amazon have created a Partnership on AI to Benefit People and Society.
Working with other companies like Salesforce and eBay, they’re partnering with nonprofits, such as Human Rights Watch and UNICEF, to further the use of AI in this space. For example, UNICEF is applying elements of machine learning to private sector data through this partnership in order to create models that assist emergency response efforts.
And Facebook, specifically, is also working with a handful of nonprofits such as the American Red Cross. By providing platform data for free, Facebook can provide disaster relief teams with valuable information. They can use tracking maps to identify the movement of people, where to dedicate resources, where people are marked “safe,” and evacuation routes.
Another coalition to have your eyes on is that of the Gates Foundation, the #GivingTuesday team, WePay, and DataKind. Funded by the Gates Foundation, this group conducted a study with top data scientists around crowdfunding to make days like #GivingTuesday even more successful.
Cliches aside, it’s an exciting time to be alive and work in the social impact industry. Never before has technology played such a crucial role. Never before has the for-profit and nonprofit industries joined forces in such intentional ways to drive change.
To stay up to date on developments in technology and the social impact industry, be sure to subscribe to our blog below.