098 – Use AI Ethically In Your Non-Profit with Joanne Toller

Episode Transcription

David Pisarek: Imagine a world where artificial intelligence boosts your fundraising efforts and strengthens donor trust. That sounds intriguing, right? Stay tuned as we explore this fascinating topic with our expert guest, Joanne Toller, Cause Specialist.

Welcome to the Non-Profit Digital Success Podcast. I'm your host, as usual; I'm David. In this episode, we're going to be talking all things AI in fundraising with Joanne Toller.

Joanne is a pioneer in fundraising and non-profit marketing since 1994. She empowers non-profit professionals with advanced digital skills. With three decades of experience, she transforms organizations and careers as a dedicated coach and trainer. Contact Joanne for a consult and discover her tailored programs at nonprofit-training.org.

Joanne, welcome back to the show. It's great to have you back as a guest.

Joanne Toller: I'm super excited, especially given the topic that we are talking about.

David Pisarek: Yeah, I think the biggest thing I have to say about AI is that there are a lot of people who are really worried and concerned about AI. "AI is coming for my job. AI is going to take over. I don't know what to do with AI. How do I leverage it?" My go-to line is, "AI isn't going to take your job, but somebody who knows AI and can leverage it will." What are your thoughts on that?

Joanne Toller: Yeah, no, I think that AI will help you do your job better. I have an example:

There is a fellow that I'm working with, and he's amazing one-on-one. He's very charismatic. He's passionate about his cause. They're doing great things, but sit him down in front of a computer, and his brain just goes... He can't do it. He cannot write to save his life, but yet he can speak it.

When we were doing his AI policy, I asked him, "Well, how and why do you use AI, and how does it benefit you?" It's all about how he can sit down and take all of that stuff out of his head and then get AI to help him write it. But then he takes it back and goes through it and gives it that human touch and personalizes it and how much it's helped him to upskill.

Writing is just not an area that he's good at. Or maybe you're a social media person, and you can't do graphic design. Well, AI can help you with creativity. It can help you upskill in a lot of ways.

David Pisarek: Yeah, I jokingly say, and even directly to my team, "AI ChatGPT, it's our best employee." It's so versatile and can do so many things, but it doesn't do anything 100%. I feel like it gets us maybe 70% of the way there. Then we need to apply our brains, our opinions, our thoughts, our language, our intonation, and our culture into it because it doesn't really get that type of stuff out of the box.

Joanne Toller: No. That goes right down to what AI is—it's built to simulate human intelligence and not replace it. Taking that into account, what do stakeholders think of AI, and are they afraid of it? Because there are definitely two camps.

There's the pro-AI and the "absolutely not" in the non-profit space. There are those two camps. I like to educate them on "What is it?" and "How it's everywhere?" It's built into so many of the programs we already use. We're talking about weak AI. We're not talking about super-intelligent AI—the "killer robot" AI. We're talking about some programming and some algorithms that work to help simulate human intelligence. We're not at the killer robot page yet, so that's not what we're using in our non-profit work.

David Pisarek: AI has been around for quite some time. Myself, I love to play with technology. As soon as I learned about AI, I was like, "Oh, okay, cool. What can I make this thing do or not do?" But anybody that... Let's see if this activates any devices I have in my office here. But if you have a Google device, you say, "Hey Google," or "Hey Siri," or Alexa or whatever—that's the crude AI versus the super-intelligent AI that we're referring to.

Actually, I have a story. The other day, I was looking to put an email together for something. Every so often, I'll hop into AI and I'll be like, "All right, I need you to act as my VP of marketing. I need an email to this person," blah blah blah. I gave it the whole prompt, and what it output was something very mediocre. It didn't have a nice language flow to it, things like that. I put back into the prompt, "This sounds too robotic. There's way too much jargon. Rewrite it as if a human had actually written this." What it output after that was actually pretty good. I made some edits to it, but it was pretty great just using it like that.

Joanne Toller: Oh yeah, prompt engineering. There is a whole thing I do on that as well. But I really do want to address the stakeholder perception thing because I know there's going to be a lot of non-profit professionals whose executive director, board, or donors don't get it, and they don't understand it.

The more we can educate them on AI and then also address their concerns around AI, the better chance we have of being able to implement some of this. Another big one that I always want to make sure to get out there is data protection and privacy.

We are not inputting sensitive information into these models. I mean, there are some specifically built for that that are very protected and secure. Those are professional platforms, not ChatGPT. We're not sticking donor data or donor information into ChatGPT and making sure that we have those policies and that everyone is using them.

Because I got to say this too, everybody's using it. You might be an executive director and think, "Oh no, we don't use AI." Somebody in your organization is using AI. They're using it. You want to make sure that they're well-trained on privacy, human oversight, and all of these things that we're talking about.

David Pisarek: For sure. There's security of information and privacy protection and all that type of stuff.

If you want to run AI on batches of data, anonymize the data. There's stuff that you can do. It's like, "These are the 80 donations we got. Here are the values, and here are the dates." But you don't necessarily need to have people's names, email addresses, or anything like that.

Then you could ask it to maybe run some reports. You can do stuff like that in spreadsheets pretty easily. But it's fun to play with in that way.

Outside of security and data protection, are there any other concerns that non-profits face around AI with regard to fundraising?

Joanne Toller: Yeah, misinformation, depending on how the AI was trained. But look at the internet. There's so much misinformation out there.

I used to do some blog writing for a client. We had to have stats in there, and I would always, always double-check those stats because half the time, they weren't correct. I don't want to be just "ChatGPT, make this" and then put it out there into the world when it could be wrong.

A lot of the concerns that are coming out right now, too, are using AI. I could, in theory, build a, let's say, Kevin Costner. I could make an avatar of him. I could clone his voice, and then I could have him do a commercial for us. Well, that's not Kevin Costner doing a commercial as a spokesperson for us. That's AI. I would never, ever, ever do that thing.

David Pisarek: There's an ethical boundary there for sure.

Joanne Toller: Ethical, yes.

David Pisarek: AI is a learning model. It doesn't create anything new. What it does is it takes information that it has learned and gathered to create something that feels like it's new, but it's based on things that it learned.

As you were talking about stats, I remember that I wanted some stats for something. I was putting together a blog article, and this was the early days of ChatGPT. This was probably about two years ago. I said, "I need some stats around this. What would be an interesting case study?" It made up an organization, and it made up these stats, and I was like, "Hold on a second." I googled the organization, and there's not a single result on Google.

If there's not a single result on Google, you know there's something a little bit fishy. However, one of the things in terms of prompts is asking it to cite the source. But just asking it to cite the source isn't good enough. You need to go to the source, and you need to check it to make sure that it's valid.

ChatGPT or whatever tool you're using—you still need to fact-check, you still need to double-check, make sure it's accurate, that it fits, that it's not taking something out of context as well because that's really easy to do.

Joanne Toller: Yeah. Whenever I work with organizations and help them build their AI policy, which is actually super easy—you don't need to hire me to do it. It is super easy.

There are certain things we go through, and human oversight and fact-checking are big components. Within it, we have an introduction that talks about how and why we're using AI so that you can address those stakeholders right away. Then we go into how we educate and train our staff and then our policies around data protection, our policies around human oversight, and then any commitment to ethical use—how we're going to stay abreast of what's going on.

And there are some things happening with different governments and how they're looking at it now. I know you put your podcasts on YouTube. I have YouTube. YouTube actually asks you now, "Is there any AI in your work?"

David Pisarek: Yeah, I like to think of AI as a helpful brainstorming tool and functionality to get the juices flowing.

If we were going to put a podcast episode together, what are some topics that I could talk about? What are some things that we can maybe dive into? There are all kinds of things like that. How do you see the role of AI in fundraising?

Joanne Toller: Well, speaking of YouTube, I'm working on a video right now, and it kills me that I'm doing this (laughs) for grant writing using AI, but again, it's just a little piece because sometimes they ask you questions in the grant proposal where your brain just shuts off. "What are they saying here? I don't understand."

Using ChatGPT to help flesh all of that out. I use it—it is instrumental for grant writing or for grant research. I'm a huge fan of theirs. They're coming out with an AI grant writing feature.

I use so many of these platforms quite a bit, like GiveButter. They're coming out with AI. So many of them have AI built right into it now that I see a lot more use for personalized communications, but especially for predictive analytics.

Now we have a tool for those of us like me who are math impaired. I really struggle with math. I take it, put it in, and then have it help me analyze the data while making sure that I take out any personal information. Because I always talk about how important data-driven decisions are.

David Pisarek: Absolutely. In my agency, we've got KPIs and scorecards and metrics, and we've got spreadsheets with all kinds of data and stuff. I wanted to merge them and pull data from here and pull data from there. I was like, "Okay, what can I get this AI to help me with to save me a full whack of time?" It probably took about two hours of prompting to get the right formula. But essentially, it gave me a 50-line formula for Google Sheets that brought in all the pieces that I needed.

Now I have a scorecard that's being dynamically generated with the data from multiple sources. We can look at this, and I look at it actually quite often to go, "Okay, this is real-time insight into what things are going on in my agency, in the business. How is it going with employees, with our clients, etc.?" Imagine if you could do that on the fundraising side with big donors that you're cultivating to help get reports automated out of systems.

It's not that the AI is continually running against the data, but that you've got formulas to bring data in, MySQL queries, different kinds of scripts, or whatever to formulate to help you with your decision-making.

Joanne Toller: Yeah, there's a program that I've recently started to use. It's going to take me a while, but it's Mind Studio where I can really begin to create my own so that I have my data sets, and then I have queries. It's like a mind map but with AI. It's so cool.

I tend to completely tech out on these things because I've used it for formula and code. I've used it to write code. I don't know that the average non-profit would, but the cost savings that it can provide.

I just look at things like branding. You're rebranding your non-profit, and you need a logo—there's AI that can help you do that in vector files so that it's a professional logo. You need to write all your new website copy, or you need to build a website—there are platforms that help you do that.

There's pretty much anything that you can think of. There is a platform, usually with AI behind it, that's available to you.

David Pisarek: The caveat to that—I've played around with some of the AI web page mockup stuff and the logo generators—they all tend to produce, at least what I've seen so far, they all look the same. If you do want to stand out, you've got to be really explicit with your prompt or just hire the people who will get it done for you.

AI is really great for the quick things, but there's a limit to it, at least right now. Maybe by the time this episode airs, things will be a little different. The landscape is changing so quickly.

Joanne Toller: I've seen some good logo generation. The ones that I've seen lacking are the video, the text-to-video, and image-to-video. I'm waiting for Soar to come out and see how that is, but any of the ones I've used haven't been that great.

It takes a lot of prompting and a lot of credits to get it to do what I want it to do.

David Pisarek: For sure.

You mentioned it briefly around policies. What type of policies and regulations should non-profits be thinking of? You mentioned it's easy to do, but what should they think about when they're even putting it together as well?

Joanne Toller: I like to start with how and why we're using AI, what are the benefits to it, and get clear on that. Then from there, you're going to detail how you handle privacy, which we should all have a privacy policy anyway for our donors, so how you're going to address that.

Then another section for that human oversight and your commitment to transparency. I've got a little workbook on it that I can share that makes it a lot easier, including an example.

I'd be more than happy to share that because, at this stage in the game, there aren't really any government policies around it.

Now, there is a bill in play right now, or there is some stuff that the government of Canada is looking at—I think it's Bill C-27, the Artificial Intelligence and Data Act. They're looking, and I believe it's with a committee right now, to set up policies around misuse or malicious use of AI and then maybe some framework around the development of AI.

Don't quote me on that. It's a massive read. I'm just waiting until it gets a little bit closer to the end, and then I'll go through it, so there's that.

David Pisarek: Or you can take it and put it into AI and tell it to give you the 15-bullet summary.

Joanne Toller: Yeah, that's very true. Get AI to answer that one for me.

I know that a bunch of the big players in the AI world got together on protecting children around AI. That was pretty neat to see. That was an initiative with All Tech is Human. There were a bunch of different big players who got together on that one.

There are things happening. There's a big fundraising conference on AI coming up as well. Just within that policy, detailing how you're going to keep abreast of everything.

David Pisarek: Yeah, I think it's important to have somebody in the team wear the hat and keep up to date with the latest trends, insights, and tools.

I think there's some website out there. I don't remember the exact name. We'll have it in the show notes. I think it's called "There's an AI for that" or something like that, where you can go and put in, "I need an app to create something," or whatever.

Sure enough, there are thousands and thousands of apps, but they're all primarily built off the same back-end engine. It's really about how it's being leveraged and is there something out there that you can use? Really, it comes down to playing with it. Is this something that you can actually do something with on a monthly basis, on a quarterly basis, or on a daily basis? Make it part of the workflow and your process so that you can achieve more and do more efficiently.

Joanne Toller: Yeah, very true.

There's a fellow I follow on YouTube, Matt Wolfe. I love him. He has all of the AI news, so it really helps. I've found some really great programs from him as well. It really keeps me abreast of what's going on so that I can help share that information.

David Pisarek: That's awesome. Do you have any favourite AI tools that are your go-to?

Joanne Toller: Okay, well, my go-to is ChatGPT. I know there are so many others, but that's just my go-to.

Then DALL-E—I know there's Midjourney and a few others. That's my go-to.

Suno is probably my favourite. It generates AI music. I run a Tooltip Tuesday, so I went in and had it create a Tooltip Tuesday theme song for me. It's so much fun. It makes these theme songs, and they're unique. Then I did another one just for me. I have a theme song now.

Eleven Labs is another one. I have a book on strategic planning, and I've been meaning to do the audio version. Rather than sitting in front of my microphone for hours, I cloned my voice and got it to do the audio version in minutes. It's crazy what AI can do. If anything is generative, I love to play with it.

David Pisarek: It's a little bit scary how some of these things are actually working. The fact that I've heard some scary stories in the media about bullying in schools where people get somebody's photo and then use AI to generate fake images and stuff like that. But there's a lot of good that it can be used for. Keep that in mind as you're playing with this and exploring it—that there are good uses for this stuff. It doesn't always have to be for nefarious reasons or anything like that.

Think about headshots of your board of directors that you want on the website. I mean, it'd be great if you could get them all together and then hire a photographer to come and shoot them, and then they all look the same. But quite often, you'll end up with photos—one's in somebody's backyard in the summer, and one is on the mountain in the winter.

Being able to use AI to clean them up and make them all look similar—would be a really easy use as well. You can do that.

Joanne Toller: That would be a great use. I mean, with permission—always ethical AI. I think that a lot of these policies—I mean, God bless Canada for being forward-thinking.

One of the things I was reading about was how they are going to police this. So if there is a child that does that—that goes out and bullies somebody else with fake AI—then there can be repercussions. It will be built into the criminal code somehow because I think that's very important.

David Pisarek: Absolutely. On a happier note, what do you see as a future trend in the use of AI in fundraising and non-profits?

Joanne Toller: Well, I just flash back to 1994. I started fundraising in 1994, and I remember the big thing back then was websites. All these non-profits think, "It's just a fad. The internet's just a fad." There were some who were scrambling, "How are we going to do this?" and how expensive it was for them to put together their websites.

I think this is the same thing that's happening right now—that it's here to stay. We need to accept it, embrace it, and find good ethical ways to use it. It's just here to stay. It's like the internet.

So either you're on board now as an early adopter, or you're going to be scrambling later on to try and figure it out then and catch up.

Social media fundraising was another thing. Back in 2007, when I jumped on board the social media fundraising game, it was the same thing then. Look at those early adopters now; they've got thousands and thousands of followers, and they were the ones who just started a few years ago or are scrambling to catch up.

David Pisarek: I remember hearing a story—I don't know how legitimate it is, but let's presume it's real. When telephones were introduced in the workplace, there was a really big resistance to allowing employees to have a phone on their desk. Thinking that employees are going to be sitting on the phone chatting away all day and talking with their friends or their family or whatever. It didn't happen to be the case. It's another tool that allows them to make it easier, whether you're in sales or internal or whatever, if you can pick up the phone and call somebody.

Same thing happened with computers—"Oh, they're just going to be there. They're going to be playing solitaire or Minesweeper or whatever all day." The internet came, and "Oh, they're going to be playing around and goofing off all day." No, it's an enablement tool. It's something that's going to allow us to do more faster. This is the AI revolution.

I think COVID really ramped that piece up a lot harder and a lot faster in terms of a technology revolution. We had the industrial revolution. This is the technology revolution that we're in right now.

I think it's just going to make it easier for people to connect with one another. We're recording this on Zoom. For us to have gotten together, and you're traveling right now. For us to have gotten together in person to record something with a film crew or whatever, come on. That would have been way more difficult to coordinate.

Joanne Toller: Yeah. Well, it's certainly changed my work.

So again, from 1994 to fundraising now—I mean, even this grant writing thing that I'm working on right now—back in the olden days, there weren't even platforms to do grant research on. There was nothing. Our database was a spreadsheet. It was so different from what it is today.

And now, I use AI in almost everything that I do. Even my blog articles—I use AI to generate the photos on it.

Actually, I start with an email to my subscribers and turn it into a YouTube script—boom, it's a blog—then it's a social media post. I've taken a few hours, if not a whole day's worth of work, and condensed it down so that now I'm freed up to do more with clients and take calls from non-profits.

David Pisarek: Amazing. Joanne, some fantastic insights. Love the conversation. We could probably sit and chat for the next three days straight.

Joanne Toller: We haven't even really got into prompt engineering yet!

David Pisarek: I hope that the folks who have been listening have been able to get some great advice and insights from you.

If you could challenge anybody listening to this episode with something, what would that challenge be?

Joanne Toller:

Create an AI policy. Then take that policy on your website—you should have your privacy policy and all that. Link that right there and put it in the footer of your emails.

Just make sure that you're sharing that and educating your stakeholders on, "Yes, AI is here to stay, but we're going to be responsible about it. Here's how we're going to use it, and here's how it's going to help you." If you can explain how it's going to help the mission and help you be a better non-profit, then they'll get it.

David Pisarek: If anybody wants to get in touch with you, what do they need to do?

Joanne Toller: The website is actually thecausespecialist.ca. That's the best website to get a hold of me. The other one is just a training website, or just give me a call. I'm always happy to talk to non-profits.

David Pisarek: Amazing. You mentioned a giveaway—a download for creating an AI policy. Where can they grab that?

Joanne Toller: I can send that to you, and then they can just go in there and grab it. That's probably the easiest.

David Pisarek: Joanne, thanks again so much for coming on the show again. We'll probably have you back. We'll dig into some prompt engineering and stuff around that.

Joanne Toller: Or killer robots when we get to the killer robots phase.

David Pisarek: Killer robots always remind me of a Saturday Live skit about laser cats. I remember that. We'll link to that video from the show notes page; it's on YouTube.

So for everybody listening, if you want the link to the giveaway around the non-profit AI policy, just head over to our podcast page at nonprofitdigitalsuccess.com and click on this episode for all the details.

And until next time, keep on being successful.