How nonprofits should (and shouldn’t) be using tech

Businesses have long leaned on cutting-edge technology to maximize profits, while the nonprofit sector has traditionally been slower to incorporate innovations such as AI. But if we want to tackle the world’s most pressing social problems, that must change, says philanthropist and MacArthur Genius grant recipient Jim Fruchterman in a new book.  

In Technology for Good: How Nonprofit Leaders Are Using Software and Data to Solve Our Most Pressing Social Problems, out September 2nd, Fruchterman highlights social good organizations that are using technology to solve real-world problems—homelessness, mental illness, climate change, child abuse, and more. Fruchterman is a tech-for-good leader and the founder of the nonprofit Benetech, which created Bookshare, an online library for people who are blind or visually impaired. He spoke with Fast Company about some of the ways technology is being used to make the world a better place and what he hopes readers glean from his book. 

It’s refreshing to talk to a business leader about how to use technology for social good instead of how to use it to drive profits.  

[Laughs.] I call it moving from money to meaning.  

What are the challenges that nonprofits and social-good organizations face when it comes to technology?  

One is a lack of money. Funding is tight. And while tech is often cost-effective, if you have a hard time coming up with the money to buy the tech, it’s hard to use. There are also often low levels of tech capacity among the staff. People are used to using the telephone as opposed to going on a Zoom call. The social sector also prioritizes different things than the for-profit sector. It’s not just about efficiency. People are still pretty important in the social change sector. Saying, “Hey, you can get rid of a bunch of people” (by implementing a new technology) may not be the best sales pitch for a charity that is trying to help people.  

What are some social problems that technology could help solve?  

I spotlight in my book TalkingPoints, which helps teachers communicate with kids’ parents who don’t speak English. If you can get the parents more engaged, kids are a lot more successful in school. That’s a great example of a technology that fills a need for immigrant parents. 

Community Solutions’ Built for Zero initiative is trying to end homelessness. For years, we treated the symptoms: Let’s build temporary housing, get people food and clothing. They’re asking, “Can we say that everyone who was homeless three months ago is now housed?” The key tech innovation is a by-name list keeping track of everyone across community places that these people go into. When shelters say, “Our beds have been 80% used this month,” that measures output, but it doesn’t say anything about whether we’re solving the homelessness problem. 

So much of what the social sector does is move information around—well, that’s what information technology is for.  

What would you say to a nonprofit leader who feels overwhelmed by or unqualified to make decisions around technology? 

Find people in your field who are ahead of you on the technology journey and learn from them. Talk to your peers. If they’re saying, “We’re writing a third more grants with the same amount of staff” using ChatGPT or Claude, then that’s worth paying attention to, because it’s not their business to sell you things.  

In your book, you highlight some bad ideas in tech-for-good efforts. Which do you see repeated the most?  

The cult of the custom. It’s the idea that “my nonprofit is such a unique snowflake that I need custom software built to solve my organization’s problem.” And businesses stopped writing custom software 20 years ago because no golf course, no restaurant, no dentist needs to be writing software to run their company. When you write your own software, you’re the only customer. It means that every bug that needs to be fixed, you’re the only one paying for it. You should look for a product that can be adapted to your needs. Also: I see lots of people building an app that no one will download. Or people following whatever the latest fad is—five or eight years ago, that was blockchain. That didn’t work out. Three to five years ago, it was the metaverse. That didn’t work out. Right now, it’s generative AI. 

I’m glad you brought up AI. What’s your take on where AI should and shouldn’t be used in social impact work?  

I think you shouldn’t replace human empathy and understanding with AI that doesn’t understand what it’s saying and have any empathy whatsoever. People in the nonprofit sector turn to human beings to help them. The best applications of AI in social good are around making the people on the frontlines of social change more effective. Let’s say I’m trying to automate a mental health counselor. Do I want to replace the counselor with a chatbot? Right now, it’s not a great idea. But if we can instead cut their amount of data entry time or paperwork time in half, then that’s time they can spend with another person who needs their help.