Ep 226: Beyond AI Ethics – Why nonprofits must focus on beneficial and responsible AI

Navigating the AI Revolution in the Nonprofit Sector

As the world becomes increasingly technological, the nonprofit sector cannot afford to be left behind. Leveraging artificial intelligence (AI) is no longer a futuristic concept but rather a necessity in the world of philanthropy. AI provides nonprofits with the ability to predict donor behaviour, streamline operations, and deliver personalized content to stakeholders - all at an unprecedented scale.

Harnessing Generative AI

Generative AI, which goes beyond predictive capabilities to generate custom content, offers an incredible opportunity for nonprofits. This technology enables organizations to deliver tailored and personalized interactions without needing significant resources or expertise. By learning and adapting to the unique aspects of each donor, generative AI allows nonprofits to harness precision and personalization, thereby enhancing relationships and fostering increased engagement.

Building Trust through Responsible AI

As with any technology, the use of AI in the nonprofit sector comes with ethical considerations. To maintain and build trust with stakeholders, organizations must demonstrate authenticity, transparency, and vulnerability in their use of AI. In essence, the implementation of AI should enhance, not replace, human-to-human interactions. This can be achieved by ensuring that AI use is explainable, accountable and transparent.

The Importance of Adapting

The digital divide in the nonprofit sector - the gap between those adopting AI tech and those who are yet to do so - is rapidly shrinking. In this light, failing to adapt to this technological shift could mean being left behind. However, adapting isn't about making a drastic leap. Instead, nonprofits can take small steps to understand and gradually incorporate AI into their operations, remembering always that AI is about enhancing human relationships rather than just about data and models.

Ready for the AI Future

The digital revolution in the nonprofit sector is well underway, and embracing AI is an imperative, not an option. The future lies in recognizing that people, at their core, are data driven and that AI is a powerful tool that can be used to build stronger, more efficient, and more impactful nonprofit organizations. To do this, we must all commit to learning something new every day and taking small, consistent steps towards a more technologically agile future.

Final Thoughts

As technology continues to advance at a breakneck pace, nonprofits must stay ahead of the curve to not only survive but thrive. The harnessing of data, predictive analytics, and generative AI is no longer a choice but an essential strategy for success in the nonprofit sector. By focusing on small, consistent steps and bearing in mind that the use of AI should enhance rather than replace human interaction, nonprofits can ensure their sustainable growth and impact in the age of AI.

Topics Covered in This Episode

1. Impact of Generative AI for Nonprofits
2. Digital Divide in Nonprofit Sector
3. Role of Trust in Nonprofits and responsible AI usage
4. Traditional Fundraising vs. generative AI
5. Future of AI in Nonprofits

Podcast Transcript

Jordan Wilson [00:00:16]:
One group that I think can really benefit from generative AI is nonprofits. But there's a lot to understand. There's a lot that you have to consider, not just on what are the best ways to use AI for nonprofits, but also how you can do it in an ethical manner and how you can do it responsibly. I have a little bit of a background there myself, but I'm no expert. But, luckily, if you're tuning in today, we have an expert that we're gonna be going over today and talking beyond AI ethics and why nonprofits must focus on beneficial and responsible AI. I'm super excited, and thank you for joining Everyday AI. If you're new here, Everyday AI is a daily livestream podcast and free daily newsletter helping everyday people like you and me learn and leverage generative AI. So if you haven't already, make sure to go to your dayai.com and sign up for that free daily newsletter.

Jordan Wilson [00:01:12]:
Each and every day, we recap, our, interview for the day, go into even more depth on the topics that we talk about, as well keep you updated on everything going on in the news, fresh finds from across the Internet, etcetera. And this is one of those, you know, normally we go live every single day at 7:30 AM Central Time. That doesn't always work for every single guest, and sometimes we do prerecorded ones. So today is prerecorded, but we are debuting this live. And don't worry. I'll be in there in the comments. So still get your questions in. Maybe we'll just happen to answer them anyways, on our conversation.

Jordan Wilson [00:01:43]:
But if not, I'll make sure to answer what I can, and maybe we'll tap into our guests for the rest. So speaking of that, I'm very excited. Let's go ahead and bring on to the show. There we go. Got him. Nathan Chapelle, who is the chief AI officer at DonorSearch AI. Nathan, thank you for joining the Everyday AI Show.

Nathan Chappell [00:02:04]:
Yeah. Thanks, Jordan. It's great to be here. I've been looking forward to this.

Jordan Wilson [00:02:07]:
Oh, me too. I love like, I can talk nonprofits all day. More on that in a minute. But tell us tell us a little bit about, you know, who you are and what you do at DonorSearch AI.

Nathan Chappell [00:02:16]:
Yeah. Yeah. Well, I mean, I've probably had a, kind of long windy career, like, in the nonprofit sector that a lot of people never intended to be in that space. And I was a technologist out of undergrad, started, one of the first, well, the first dotcom to sell skis on the Internet. I just early adopter. I was focused on big data storage and and that type of thing. Got into nonprofit on accident, Thought I'd be there for a few years, and I found myself there for 20 years. Always a little bit of a fish out of water thinking, like, why is this sector so unique in the in their, like, resistance to adopting technology, and why are they so slow? And, back in 2017, you know, frankly, I was just kinda dismayed at the lack of innovation in the sector.

Nathan Chappell [00:02:56]:
So I, you know, sat down and and ended up creating the first algorithm that predicted gratitude, that we spent about a year and a half, spent about a 1,000,000 and a half dollars to build. It worked really well. It's been totally disruptive to the sector, and and that's what I do. So my team at DonorSearch AI, probably the the I mean, I I'm very biased in this, but, we have an incredible data science team that now does this with, I mean, 100. I mean, we have about 10,000 nonprofit clients and about, over a 100 of them where we build and operationalize custom machine learning models, for. So that and now, of course, I spent a lot of time, both in the area of predictive and generative together and kind of what that means. And also on the responsible AI side, like you said in the introduction, which is definitely more my passion project, that tends to be more my nonprofit work right now. It's really on the the responsible AI side.

Jordan Wilson [00:03:47]:
Okay. I have so many questions now, and I'm I'm like, if you're listening on the podcast, I'm geeking out smiling because, yeah, I I myself spent, you know, about nine and a half years, working at a nonprofit. So, you know, what, Nathan's talking about there, I feel it. Right? But, you you know, Nathan, I'm I'm curious. So you've you've worked with, you know, you said, more than 10,000, nonprofit clients. You know, when you're talking AI to to the average, you know, nonprofit worker Yeah. Is it is it a lot? Because at least in my experience, I feel there's 2 types of nonprofits. I think there's, like, 2% that are, you you know, so into to data and, you you know, donor data and technology.

Jordan Wilson [00:04:26]:
And then I feel everyone else can kind of be a dinosaur, unfortunately, and it's hard. I mean, what's your experience been working for so long with so many nonprofits when it comes to adapting to new technologies like artificial intelligence?

Nathan Chappell [00:04:38]:
Yeah. I mean, it and, you know, to level the playing field, there's about 1.71 800,000 nonprofits in the US alone, around 10,000,000 in the world. And so, there is absolutely a digital divide. You know, there are those that have, you know, in in this case, we'll say, like, have, you know, a view on, like, big data and data storage, data collection, and now AI, and those that don't have. And there's a stark difference. So, you know, we a vast majority of our clients buy data from us and that they've done that for 15 years. Those are the ones that at least see the value in big data. And and at you know, for the most part, throughout my career, we always bought a lot of data.

Nathan Chappell [00:05:14]:
Didn't really know what to do with it. It wasn't really until AI, you know, really, truly more on the predictive side, like machine learning, deep learning allowed us to take all that enriched data, the data that we had as a nonprofit plus the data that we could buy, and then start, you know, making sense of it. That's still pretty new. I mean, there are a lot of reasons why we could spend a whole you know, we could spend an hour talking about why nonprofits are slow to adopt, but we don't have time for that. But I think what's more important is that the nonprofit sector is is challenged right now. Less and less people are giving to charity. I wrote a book about it called The Generosity Crisis in November of, whatever last year was or the year before last year. Time flies.

Nathan Chappell [00:05:52]:
But the, the reality is that, are less and less people like you and I are giving to charity because we're essentially disconnected. And so we see, like, the advent of AI and how AI can now essentially allow nonprofits to become more efficient, allow them to spend more time human to human. Like, this is a breakthrough and an advent of time that nonprofits have always been waiting for. And so, I know you you focus a lot on Gen AI. And and, you know, from a Gen AI perspective, like, I started 20 years ago, like, when you were a nonprofit. Can you imagine what life would have been like if you had tools like, you know, perplexity or cloud or chat gbt at your disposal? Like right? Like crushed it. Like you would have crushed it. Like, and so now we're in this time where like nonprofits can level up and, and compete at a level they've never been able to, yet they're still so slow to adopt.

Nathan Chappell [00:06:40]:
Only about 5% of nonprofits are actually fully deploying AI right now. So a lot of work to do. We made a lot of progress, but know, still definitely a lot of work to do.

Jordan Wilson [00:06:50]:
So one thing that I wanted to pull out of that response there, Nathan, is, you know, how there is this, you know, like like the book behind you, you know, for the those joining on the, on the live stream, you know, there is a generosity crisis. Right? So I think that a lot of people have this thought of of generative AI or AI as, you know, people think, oh, it's robots. It's, you know, Terminator. It's Skynet, etcetera. But what I'm hearing from you is that nonprofits can actually use AI, to make things more personal and and to strengthen relationships. And it's not necessarily a a very impersonal, you you know, tool or technology. Can you talk a little bit more about that and how maybe nonprofits might want to change how they're viewing, artificial intelligence to actually be something that can bolster relationships and and drive positive change for this, you know, generosity crisis we may be facing.

Nathan Chappell [00:07:46]:
Yeah. No. I mean, such a great question. Lots to unpack there. I mean, the the reality is that for, you know, about 40 years now, nonprofits have been essentially using data largely based on like how wealthy people are. And they've been taking that data and prioritizing this idea of like, hey, you're a good donor or you're a good prospect because you're wealthy and I'm gonna spend more time with you. The reality is that wealth and generosity have very little to do with each other. Our very first data science project showed that wealth alone is about less than 10% predictive of whether or not you're likely to make a gift.

Nathan Chappell [00:08:15]:
So no matter how wealthy you are, there's only a 49.6 percent chance in America that you're gonna give to any charity. Yet most nonprofits have just thought, oh, if I could, you know, get a dollar from everyone on the planet, you know, we'd we'd be rich. So first is, I think, accepting this idea that not everyone is a good prospect. So predictive AI was really that area that we could really narrow in that people that had the deepest connection to our organization were best prospects. That makes sense. So you're measuring connection. And then my connection today is different than my connection tomorrow, in 2 weeks, in 2 months, in 2 years using AI to to really harvest that connection. So the predictive side has been around for a while and has been actually pretty helpful at streamlining and identifying not all people are created equal.

Nathan Chappell [00:08:58]:
Some people care about you more. And those that are raising their hand will show up in data. Then what gets really exciting is advent of Gen AI to basically take those predictions and then take action on them. Right? So, like, predictive AI is about predicting things and making sense of all that data, but generative is about creating things. So you match. And I think this is where our sector is gonna go, where it's, like, really understanding, like, just creating more for the sake of creating is not gonna help anyone. But creating more and and specific to an individual, where now we're talking about, like, this idea of precision philanthropy where it's like precision medicine. Like, every person is unique to, like, how a a therapy will help them or hurt them based on their DNA.

Nathan Chappell [00:09:39]:
Precision philanthropy is the same idea that there's no such thing as a good donor or bad prospect, that all people have a varying degree of connection to your organization. And then the tools and and that you use to essentially activate and spend more time with those individuals is really gonna be, you know, what make will make a big difference. And I'll say all that to say, AI is not a silver bullet. It's not a magic wand. It should never replace a human, But what it should do is it should create a tremendous amount of efficiency where people in nonprofit who are the hardest working, you know, people as you know, having been in that sector, like it's a passion, not a job. You know, it's, it's a, it's a way of lay life, not a career. How do I use AI so that I can spend more time human to human? And that's true with AI in any domain, whether it's in, you know, clinicians, you know, wanting to use AI so they could spend more time with their patients or whatever the use case might be. That's where I see things get really exciting for the nonprofit sector.

Jordan Wilson [00:10:37]:
You know, something that you mentioned there is, you know, kind of this intersection of predictive AI and generative AI. So so, Nathan, maybe for those, people in the nonprofit sector out there and they're hearing this and you you know, you hear all these buzzwords. Right? And and for those people that maybe aren't, you know, lifelong technologists or have a background in machine learning like you do, can you briefly explain, you you know, kind of what that means, you know, bringing predictive AI and generative AI together and specifically, you know, how, you know, nonprofit people that maybe aren't even super technical can take advantage of that intersection.

Nathan Chappell [00:11:14]:
Yeah. No. That's a great point. I actually, did a poll recently of, like, how if someone asked you, could you explain the differences between predictive and generative AI and, you know, at least fumble your way through it. So it's you know, we're we're getting there. You know? So I think what resonates for most people, you know, this idea, especially in the nonprofit sector, are questions like how much will someone give or when will they give or, or if they will give. So all those things were predictive AI, which, you know, predicts things. So it's about precision.

Nathan Chappell [00:11:43]:
It's about creating precision based on lots of data. Those questions are not new questions. In fact, like, I I I started talking about this back in 2017, and I stumbled upon this quote by Aristotle. Like, this is, like, 23 100 years ago. Aristotle said, you know, the idea of giving away money is easy, but how much to give and when to give and how to give it in what way, that's hard. Right? Those are essentially barriers to generosity that have existed, like, since the beginning of, like, you know, Western civilization. But the idea of, so like predicting who might do those things or how much is one thing, but the idea of like, what's the best way to reach out to them in a way that's a personalized to them based on their own unique characteristics. That's generative.

Nathan Chappell [00:12:25]:
So generative, again, creates things where predictive AI predicts things. If you mat if you match those things together, now you have precision and personalization at scale. I and, you know, I think one of the best examples that I've seen recently is like Carvana. Like Carvana is like, as a company, you know, struggling a bit, like the stock market's got not not been super great. They jumped into this, you know, predictive gen area because they're like, okay. We have lots of people who bought cars. Those people have stories about their cars. And I don't know if you've seen this example.

Nathan Chappell [00:12:54]:
So if others haven't, it's a really great example. They produce, like, I think, 1,300,000 custom videos. Basically, it's a picture of the car that I bought, like, 2 years ago, and it's a car talking to me. So everything is generated through, you know, Gen AI, but they use precision just to determine, okay, not every customer is gonna get one of these videos. Only certain customers are gonna get these videos. But they were able to, like, offer that precision at scale and in a really profound way. And I see that that blend of precision and personalization supporting the nonprofit sector. So, well, like being able to say like, you are important to us because of these reasons, and this is a journey we've been on together and, you know, our journey has just begun.

Nathan Chappell [00:13:35]:
Like how about we, you know, change the world together and be able to do that at scale on the fly, on you know, in real time would would be just absolutely amazing. So we're already seeing these technologies emerge, you know, really, you know, prolifically in the, in the private sector. So now in the nonprofit sector, which lags, it's like they become more affordable and they, you know, become scalable where nonprofits now, you know, for many years didn't have access to machine learning. Now they do. They didn't access to generative AI. Now everyone has access to generative AI. So it's, again, I'll leave with the optimistic side of, like, it's a really cool time to be alive. Nonprofits really need to, like, double down to get out of their comfort zone and learn something new every day about AI and probably starts with listening to your podcast every day.

Nathan Chappell [00:14:18]:
If you do that, you're probably, like, 95% ahead of the, you know, the rest of the the crowd.

Jordan Wilson [00:14:24]:
Yeah. And one thing again, like, when you keep saying these things, right, I'm like, man, what a time to be alive. Right? Like like, if you're a nonprofit right now, like, because I I remember personally, you know, I I would spend, you know, 10, 15, 20 hours, on these on on on these projects that now take 10, 15, 20 seconds. Right? Yeah. So so even this example, right, of of Carvana, you know, kind of operating with this, you know, precision and personalization at scale. What do you think is is maybe some examples? You know, what is that that example for the average nonprofit that they can use, you know, precision and personalization at scale because it's it's available now. Right? Like, you have the data. The data's always been out there for nonprofits, but the data has usually led to a lot of manual work.

Jordan Wilson [00:15:14]:

Nathan Chappell [00:15:14]:

Jordan Wilson [00:15:15]:
So what is that kind of sweet spot between precision and personalization that your average nonprofit can go take advantage of right now.

Nathan Chappell [00:15:22]:
Yeah. And you can start small. I mean, it's all about baby steps. Right? Not everyone's gonna, you know, build and operationalize custom machine learning models and then, you know, build out the curve on a type thing. But the reality is that, you know, almost any nonprofit can use they they have some basic information. So say, like, a university or a k twelve school or whatever be like, hey. This person went here around these the this year. You know? This is the year they graduate or this is the the the degree they got.

Nathan Chappell [00:15:48]:
And, you know, or, like, they volunteered at our our food bank or whatever. They they can track, they absolutely can track some of that data that shows, like, this person cares about us. And at the very, very least, and it's super cheap to do and not many do, is, like, do a survey. Like, find out, you know, like, start out with building a Google survey to find out, like, this is what people like about us. This is what this person cares about us. We know that they're one of their top charities or, you know, they're not. Take any of that type of data, which we would consider, like, experiential data. It's like represents, like, how they care.

Nathan Chappell [00:16:22]:
Do just a little bit of segmentation around that, then use generative AI essentially to then build, say, an engagement story about your organization that's gonna speak to the reason why the person answered that survey that way anyway. And so, like, this like, that like, understanding someone from a survey response, building, an engagement strategy with a person that represents their sediment of what they were feeling in that survey response, that is, like, super scalable. Like, that is, like, every that literally costs no money. So, like, when I'm talking about the digital divide and nonprofits that don't have budget for this could do things like that right away. And not to say that, you know, they don't have to learn something new. Of course, you've gotta learn something new. You've gotta get out of your comfort zone. But start small, take baby steps, and just start doing it.

Nathan Chappell [00:17:07]:
And you're gonna figure out how to automate those types of of things, how to deploy a GPT so that you can have people engage with your nonprofit and learn more about the history and and what it means and how it impacts society. So, I mean, think about things like communication, thing, you know, from a general perspective, like analysis, creative ideas on thanking people. Like, what a novel concept. Like, people get bored of thanking people the same way. Like, ask GPT. What are some creative ways to thank someone? So, I mean, I could probably write off 20 use cases right now that would cost a nonprofit $0 that would actually make them feel and appear more modern like Starbucks who's like, hey. It's your birthday month. Like, we've missed you.

Nathan Chappell [00:17:49]:
Like, well, you could come in and see us and the coffee on us. Like, those things don't you don't have to be Starbucks to do that. Like, you literally could do that right now with, like, $0.

Jordan Wilson [00:17:59]:
Yeah. And I don't I don't know, if there's ever been a time and and I'd I'd love to get your thoughts on this, Nathan. Is has there ever been a time in in your experience, you know, working in this field, you know, you obviously have a a much, deeper and, you know, more tenured background in machine learning and AI than the average person out there. But has there ever been a a time before that you've seen, where the digital divide between nonprofits and everyone else shrinks this quickly? Because, you know, even my own personal experiences, you know, we would sometimes get, you know, support or grants, you know, from these big tech companies. And they were great by, you you know, supporting nonprofits. But then you go see, you know, how they operate. You see the technology that they're using and, you you know, from the nonprofit perspective, it's like, wow. That seems like a different world.

Jordan Wilson [00:18:48]:
No. I I I personally don't know if it's like that. You know? It seems like it's a level playing field. Can you speak to that a little bit and talk has there been a time before where the divide seemingly is shrinking that quickly?

Nathan Chappell [00:18:59]:
Well, I mean, I if, you know, because I'm old and I, you know, in high school I was learning DOS. So, like, when I, you know, in 90, you know, probably 95, 96, you know, the Internet was a great equalizer between those that have and those that didn't have. So, like, you know, the only firms that afforded compute power were those that, like, were were buying, like, big servers. Right? So, like so so that was a time where the digital divide shrank very quickly. It's like, you know, it's it's insane to think now. It's like, hey, nonprofit. Like, are you using the Internet? Well, of course, you are. So, like, so I think this is where a lot of people equate, like, generative AI to, like, the Internet because it is a great equalizer in the sense of, like, you don't have to, like, build a a big server farm, and you don't have to even pay AWS to have, like, a big or Microsoft a big cloud environment.

Nathan Chappell [00:19:44]:
You just, like, go on your Internet browser, and now you have, like, all the stuff at your disposal. But I think the the scary part with this is that nonprofits, again, being so slow that the the key difference between, like, generative AI or just AI in general compared to, like, the Internet is that AI is an exponential technology, meaning that it doesn't say the same. It it gets better. And so so progress in exponential technology is not measured in time. It's not measured by like, oh, I'm not ready for this. I'm a little scared of AI, so I'm gonna do it in a year. You're not 1 year behind. You're, like, 365 AI cycles behind.

Nathan Chappell [00:20:21]:
Right? And so it's, like, this idea Harvard wrote this article on this, like, years ago, and it stuck in my brain of, like, those that failed to deploy AI may never catch up because it's that exponential technology. And so the digital divide seemingly, to your point, because you're you're looking at this objectively, is like, oh, it's like the the great equalizer. Everyone has access to it. Everything should be great. The reality is if the nonprofit sector does not get out of their comfort zone and they are like, oh, I don't trust this. I'm not gonna do it until everybody else is doing it. That digital divide will grow exponentially. And and while at the same time, there's, you know, lots of reports out there from whoever, Mackenzie and the IMF, who is saying, you know, like, AI can be, you know, can help bridge the digital divide.

Nathan Chappell [00:21:05]:
It actually can also, you know, 60% of jobs are gonna be impacted by by AI in some regard. And by the way, inequality might increase. So this is, again, I think this is a a need and opportunity for the nonprofit sector to rise to the occasion since the nonprofit sector focuses on inequality at large, that's kind of what they're built to do is that waiting and seeing is contrary to like what this what society needs right now. And so this is why, again, it's like, it's really great time to do your live, really great opportunity, but it's also like a moral imperative that the nonprofit sector rise to the occasion and be like, okay. We're in this. Like, let's let's help steer this technology into responsible ways.

Jordan Wilson [00:23:01]:
You know, my brain right now is rattling because, you know, I'm trying to process something here, Nathan, because, what you just said right there. Right? Like, my way of thinking has always been yes. You know, generative AI can be something that helps bridge the gap, but it makes sense. It makes sense what you're saying. It can bridge the gap if you're, you you know, putting through, or or or putting in the work, I guess, to actually use and understand generative AI. But if you're not, if you're falling behind, you know, that 365, you know, light years in AI Right. You know, behind, then then the divide actually becomes much, much greater.

Jordan Wilson [00:23:46]:
So how can nonprofits keep up? Right? Because I think the biggest thing is is there's always, you know, having to to win that trust of a nonprofit right before using a new technology a lot of times from personal experience. So, hey, if if you're nonprofit leader out there, don't get mad at me. But, you know, there is more of a distrust with new technology

Nathan Chappell [00:24:05]:
Oh, for sure.

Jordan Wilson [00:24:06]:
Because people say, oh, you know, I wanna protect my my donors, my volunteers, which I which I get and I understand. So how can they balance all of that?

Nathan Chappell [00:24:14]:
Yeah. And I I think some of that's warranted. Right? Because, you know, if you compare and contrast the nonprofit sector, the private sector. So let's use we could pick on Twitter. I don't know. Hopefully, they're not a sponsor of yours and this will piss them off. But okay. Okay.

Nathan Chappell [00:24:27]:
So and now it's not even Twitter anyway. So we'll just you know, at at one point years ago, Twitter created an algorithm that was, like, I don't know if you remember this, but, like, racist, ageist, Islamophobic, and ableist, and all those thing, every ish that you can imagine. And it got out, and this algorithm was predicting all these really kind of bad things. And, of course, like, their stock price goes down. But did that affect, like, Facebook stock or Google stock or Microsoft? Not at all. Right? Because that was one single player. That was fate that was, whatever we were just talking about Twitter that did those bad things. Right? And then they had to rebuild trust with their their stakeholders and their shareholders to rebuild that.

Nathan Chappell [00:25:02]:
But if the nonprofit sector did the same thing, and especially if, like, say a large nonprofit is trusted. So, like, Red Cross or somebody like that created an Aegis, Islamophobic, ableist algorithm, it would impact trust in all nonprofits. So the reality is that the nonprofit sector has nothing to provide back to individuals in terms of, like, a service or a product. So, like, if Nike, you know, did something really bad, but I still really like Nike because they fit my feet better, I'm gonna still buy Nike. But a nonprofit sector, essentially, you're exchanging money for trust. You know? So the nonprofit sector is in the trust business. So I do think it's warranted to some extent that the nonprofit sector needs to prioritize trust at yeah, at the, at the very highest level and ensure everything they do essentially does not diminish trust. And with that said, advances in technology and AI has been so fast and there are no real universal standards for what responsible AI is.

Nathan Chappell [00:26:02]:
So of course, like in the UK, there's some and the white house has very loose guidelines on like responsible AI. And, but so many corporations have created their own to try to instill trust. The nonprofit sector is saying, well, wait. Like, let's let's wait and see what this means until we're gonna, you know, go full bore. That's where fundraising AI, you know, stepped in. And back in November of a year ago, we created the first framework for responsible AI for the nonprofit sector. And that specifically was to highlight the need for things like transparency and and beyond ethics, but like transparency, accountability, explainability. And not to only look at is this ethical, but is it also beneficial? So, like, we might say that, like, using, you know, Instagram is not unethical, but is it beneficial? And we've seen lots of downside from that.

Nathan Chappell [00:26:53]:
And so those are the things that I think the nonprofit sector has to wrestle with. Not just things that are ethical, but are they also beneficial? And so that's kinda where we're at today. And and I think we made a lot of progress in that space. But there's, you know, there's a lot of room to grow. And I but I do think that this is a time where the nonprofit sector has to lean in, and they have to take, a strategic and measured approach to using AI to remain relevant and to address inequality as as it is also created by AI.

Jordan Wilson [00:27:21]:
So we'll definitely link to that, the framework for responsible AI in the newsletter. But, Nathan, if you could, maybe just give us, you know, a super high level overview of, you know, what nonprofits should be looking at. Right? Because, I feel those that that that quote, unquote get generative AI and and and they can work at that, you know, intersection that we were talking about earlier, you know, with with precision and personalization, they're they're gonna be, you you know, have abilities, you know, when it comes to, you know, outreach or fundraising or data collection and and what you can use the data for. Maybe new capabilities or powers that they didn't have before. So, you know, what are some of those things that that they should be focusing on once they do get it to make sure that they're still using it in a responsible way that benefits others?

Nathan Chappell [00:28:11]:
Yeah. I you know, and I don't think it's that hard. I would say when you're when you're thinking about the litmus test for a nonprofit is that, you know, is this not only beneficial for my short term goals, but also is it beneficial to our sector as a whole? So I think that's something that a nonprofit sector kinda uniquely has is this kind of protectionism of the sector as a whole, of society as a whole, and as they should. Right? Because we nonprofit, nonprofits and nonprofit employees are kinda they're grounded by this idea of, like, everything that we do and it has, you know, an impact. And so I I think for the most part, nonprofits, when they lean into this, have to ask themselves the the question. Like, is sounding authentic the same as being authentic? Right? So prioritizing authenticity, transparency, vulnerability is something that we have to go a step beyond. Whereas, like, most frameworks that exist in the world will focus on, like, robots serving humanity's best interest. And so that's like White House lingo.

Nathan Chappell [00:29:09]:
The White House lingo is like, we need to build AI that serves humanity's best interest. That literally translates to, like, robots not killing people. That's not a that I mean, of course, that's important. Right? We don't want robots going down the street. We've seen chappy, you know, and, like, bad things happen. But the reality is that nonprofit sectors the nonprofit sector needs to really focus on on AI that is going to preserve and instill trust. If there was just one question, like, if I'm using generative AI, predictive AI, and however I'm using generative AI, so whether it's communications or analysis or HR or legal support or, you know, creating images. Like, if I was to ask myself one question, does this action preserve and protect trust? Like, not rocket science at all.

Nathan Chappell [00:29:52]:
Right? Like, so that's just one question. Like, everything I do in generative AI, which is a lot. Like, I could do so much with generative AI. Does this activity preserve and protect trust? And what that means in the nonprofit sector is that when you're using AI, essentially, there should be the call to question like, how do I trust a black box? Like, I need to make sure that it to the best of my knowledge, and Jen is very hard in this case because, like, predictive AI, you can actually see mathematically, like, how a prediction is made. Like, it's actually fairly easy to see, like, prediction is made by the combination of these data points, you know, that that say this person is gonna do the same. Generative is a little harder. So I think you've gotta work with and use generative that, is taking ample activity, to make sure that they're being responsible in their building of AI. So, like, an OpenAI or a cloud or or perplexity, which, like, essentially, you know, will give you footnotes of everything that it's doing.

Nathan Chappell [00:30:46]:
I think that that's almost a requirement in the nonprofit sector where it's a nice to have in in the private sector that to really understand how the AI is actually making decisions, what it's telling you to do, and to ask the question, does this activity preserve and protect trust? Like, go from there. You know? And so, you know, demand kind of those things, transparency, accountability, explainability.

Jordan Wilson [00:31:08]:
Yeah. I think I think the explainability, especially of generative AI, is is crucial for for nonprofits. Right? Because, yeah, they have to be able to communicate it not only internally, but also with external stakeholders, you you know, donors, volunteers, etcetera. You know, what they're what they're doing. You you know, how they're using, this data that they do have. So so, Nathan, we've we've talked about so much in this episode. Right? We've talked about machine learning and donor data, predictive AI, generative AI, you know, and then how we can also you know, how nonprofits can can use this all, responsibly and use it, transparently to benefit others. So so, normally, I don't ask people, you you know, to to predict the future in their space because it's hard.

Jordan Wilson [00:31:49]:
Right? Like, we kind of talked about every day is almost like, you know, can almost feel like a a year of development. But, you know, with with someone such as yourself who spends so much time and has such a deep background, for maybe those nonprofits right now that are hearing this and they haven't fully adapted. I feel if if if if you kind of go to where we're at now, you're gonna be behind maybe. Yeah. Right? So how can they, you know, sort of skate to where the puck is going, when it comes to, you know, generative AI and taking advantage of the data and the, you know, technology that's out there? Yeah. I I mean, it's such a good question. I get this question a lot. You know,

Nathan Chappell [00:32:28]:
nonprofits or fundraisers that use AI will replace those that don't. And so I I mean, fear you know, the fear is that it's actually just not like fearmongering. Like, that's true. Like, you will not be in a place where you as an individual will compete, or as an organization will compete in the future if you're not using AI. And so now is the time you have to AI proof your career and your organization. So with that said, you know, start where you start. But to your point, like, generative is so it's a great equalizer. It's so accessible that there's no reason not to.

Nathan Chappell [00:33:00]:
So, I mean, I think about that movie, What About Bob, which is like Bill Murray and, I forget Paul, whoever it was. I forget it was, whoever it was with. Pylori was, like, this idea of, like, baby steps. Like, you just gotta take baby steps and, like, every day learn something new about AI. Try something new. Always asking the question, does this preserve and protect trust? But just start small. I think when most people don't start working in AI is because they're just afraid of starting. And, you know, the kinda common kinda philosophy in AI is that 70% of of AI has nothing to do with data or models.

Nathan Chappell [00:33:35]:
It has to do with people. So take away that idea of, like, AI is all about data and models because it's not. It's about you as an individual and the people around you and how you will either level up your your work, you know, and see you know, because you spend a lot of time in generative AI. It's like, you know, producing more work faster, you know, more accurately and with less burnout. Like, who doesn't want those things? Like, that literally is the definition of every nonprofit worker I've ever met. You know? So, like, let me do something faster with a higher, you know, quality of of work and be less burned out, like, all day long. So jump in, baby steps, learn something new every day. I always tell people, start with podcasts.

Nathan Chappell [00:34:16]:
Like, like listen to one new thing every day and, you know, and it doesn't have to be rocket science, just like get comfortable with the lingo. And then, you know, log in to chat gpt. It kinda starts there or or Cloud or Perplexity or or Copilot or whatever your flavor of choice is. And just start, you know, start asking it questions that solve, you know, some of your more immediate business needs.

Jordan Wilson [00:34:42]:
So good. And I think, Nathan, I I love the, the analogy of baby steps, but I think that you helped all of us and not just those who are in the nonprofit sector. I think you helped all of us take a little bit more than baby steps with today's conversation. So thank you so much for joining the everyday AI show. We really appreciate your insights.

Nathan Chappell [00:35:02]:
Absolutely a pleasure. Thanks so much, Jordan.

Jordan Wilson [00:35:04]:
Hey. Thank you. Thank you you for, listening, everyone. We appreciate your time. Make sure if you haven't already, go to your everydayai.com. A lot of fantastic info in there. So some of the things that we reference, we're gonna be putting in the newsletter. Make sure to check that out.

Jordan Wilson [00:35:18]:
So go to your everydayai.com. Thank you for joining us, and we'll see you back for more everyday AI. Thanks y'all.

Gain Extra Insights With Our Newsletter

Sign up for our newsletter to get more in-depth content on AI