Rachel: There we go.
To provide opportunities for grant professionals like yourselves and nonprofit leaders to learn from each other. The goal in this grant’s roundtable series is to bring you all into the conversation about a topic related to our fields getting support from each other, answering questions, and sharing knowledge together.
For those of you who are new to Instrumentl, we're the most loved all-in-one grants platform for grant prospecting, tracking, and management. We currently helped more than 3,000 nonprofits and grant consultants save time in finding and applying for more grants.
To start us off, today, I'm going to quickly introduce the team in Instrumentl who's putting this event together. My name is Rachel and I'm one of the newest team members at Instrumentl. My job is to help develop educational events, like this one, to share with our Instrumentl community. I have a decade plus of experience in the informal education field, specifically in museums, and I'm based in Los Angeles, California.
Hi, everybody.
I'm also very excited to be joined by my colleague, Ryan. Ryan will be supporting behind the scenes in the chat box. So, feel free to ask questions as we go through the presentation. He'll help out as much as possible with those links you might need and answers to questions. It's also Ryan's birthday today. Happy birthday, Ryan. So make sure to give him some birthday love in the chat. And thanks for doing this event on your birthday, Ryan.
Some quick FYIs. So as we get started, our event will be about an hour long. This is a panel discussion with a live Q&A. It will be recorded, and the slides will be shared with everybody afterwards. So keep your eyes peeled for that follow-up email. And try and remember that if you ever have things you want to review afterwards, you can always look back on those recordings and slides that'll be sent to you.
Thank you so much for spending time being with us today. So make sure you're taking care of your personal needs while we're together, bring a snack, take a bio break as needed. But I want to make sure everybody's back here at around actually like 11:35 or so, my time. I'm sorry, 12:35 my time, that’s Pacific Time. That's when we're going to start our live Q&A and share a little more about how to get those freebies and enter our raffle.
To make sure you can follow along with our speakers, I have enabled closed captions. These are automated closed captions that you can toggle on in Zoom. You can use that Zoom taskbar down at the bottom to make sure your closed captions are visible.
We have some very fun freebies for those who stick it out today and submit their feedback form. So if you share your feedback with us at the end of the program, you're going to get a copy of our newest resource on AI, 5 Tips for Utilizing AI in Grant Proposals & Reports. And thanks to our partner, Krista, for an awesome opportunity. Everyone will get a chance to learn the latest generative AI strategies and techniques to save time and win more grants with Krista’s grant writing with AI simplified training. She will be teaching the most streamlined and simplified way to implement AI in your grant writing process in this free event. So everybody who submits their feedback form will have access to this freebie.
Plus, if you are a lucky winner of our raffle today, Margit has generously offered a special raffle prize access to her self-paced online program, All About Grant Writing, which is a $997 value, and includes small group coaching on AI with Margit herself. Thanks partners for sharing these awesome opportunities for our attendees.
So as we get rolling here, just some reminders of how you can participate, this is your time to learn. So take screenshots of slides, jot down whatever is really resonating with you. And let's get inspired by this learning community today. Use the chat box to share ideas with each other. And I promise those emails or Slack messages can wait. We want to try and be present with one another during this program today. We'll gather all your questions for our closing Q&A. So, add questions in the chat with those three hashtags at the front. That'll help Ryan organize those so I can moderate them at the end of the program. And as a kind of reminder, we're going to stay on mute for the duration of our program just to make sure we're creating the best environment for learning.
One last thing before I intro our special guests, we're trying something a little bit new where I've created these event resource web pages for every event that we put on. So, I know the chat box can get a little unwieldy sometimes. And there's a lot of great info in there that we want to look back on after the event. So this is going to be a repository for you to look back on and make sure that you can review all those links that have been dropped in the chat, any nuggets of wisdom that your peers have shared. And we'll be able to share that. Ryan just dropped the link in the chat. So, I encourage you to go ahead and save that or bookmark that page so you can check it out after the event. I'll add in all of our little nuggets of wisdom that gets shared.
Awesome, Constance. I'm so glad to hear that.
Okay. Finally, getting to the point where I get to introduce our awesome speakers for today. Starting with Margit. Margit Brazda Poirier is the owner and founder of Grants4Good LLC. She created the company in 2009 to help nonprofit organizations and businesses find and get grant funding. She is one of the 400 or so nationally certified grant professionals, and one of only 30 approved trainers in the US on this topic. Margit and her team have written and received millions in grants for nonprofits and businesses. And she has the unique perspective of understanding both grant seeking and grant making, given her experience leading a prominent New York foundation. She's spoken at national and international conferences of the association of funding professionals, grant professional association, and others. Welcome Margit.
Margit: Thank you, Rachel. It's great to be here. I would normally be joining you from my hometown of Rochester, New York. But it is February break week for kids. So, I'm in Colorado. We're skiing. Well, not today. But we are. We are here skiing for the week. So I saw some folks from Colorado and I saw some folks from Rochester and everywhere in between. So, welcome everybody.
Rachel: Yay. Thanks for joining us on your vacation. That is so kind of you and we're so glad to have you.
Our next speaker for today is Dr. Krista Kurlinkus, founder of Grant Writing Made Easy and CEO of Krista Kurlinkus LLC. She has an impressive background as a PhD in English rhetoric, composition, and literacy studies. Her eight years in grant writing, consulting and education have established her as a nationally recognized expert in the field of grant writing. She is also a GPA-approved grant writing training trainer. Dr. Kurlinkus has written winning multimillion dollar federal grants, global development grants, state grants, multiple six-figure research grants in corporate, private, and community foundation grants for over 100 nonprofit government and academic clients.
Welcome, Krista. Thank you so much for being here.
Rachel: Rachel, thank you so much for having me. And I am now in Roscoe, Illinois, since we're talking locations. I moved here six months ago from Norman, Oklahoma. So if anybody is in the Chicago area and wants to meet up, let me know in the chat. So, just thank you to everybody for being here and taking the time out on your Friday.
Rachel: Thanks, Krista.
I'm going to start us off with a quick grounding definition just to make sure we're all on the same page on what we're talking about today. So, what is AI? AI or artificial intelligence is going to be the simulation of human intelligence processes by machines, mostly computer systems. So artificial intelligence is this machine's ability to perform cognitive functions that we generally associate with humans, right, such as perceiving, reasoning, learning, and interacting. There are many different forms of AI from your Roomba that is maybe using AI to map your home. I have one that moves around my living room every couple of hours, to your smart home devices like Alexa, to even self-driving cars, right? But what many folks are talking about these days are natural language processing chatbots. That includes ChatGPT, or Google Bard, now known as Gemini. And these are also known as large language models or LLMs. So, that's what we're going to be focusing on specifically today.
And before I launch into my first question for our panel here, I'm curious to get a sense of who is using some of these tools in the room. So, I'm going to launch another quick poll. It's going to pop up in front of you in just a second. What, if any, AI tools or platforms are you currently using right now? You can select all that apply. And these include ChatGPT, Google Bard, Microsoft Copilot, other. And if you have an other, if you don't mind sharing it in the chat, I love to see what folks are using. Or none of these. I'm new to this. And that's okay, too.
I'll give folks about 10 to 15 more seconds to answer, and I'm seeing some folks answer in the chat as well. Grantable. Oh, Cloud AI? Very cool. Catherine. Otter AI? Yeah, Constance, I've used that one as well. Grammarly? Yep. Laurie, I see that. Fireflies AI? I'm curious about that one.
All right. I'm going to go ahead and close our poll and I'll share the result so everyone can see what we're looking at here. It looks like predominantly folks are familiar with ChatGPT. We've got some other platforms being used in there. And then otherwise, people are pretty new to this. So, this will be a nice opportunity to dive into some of the specificity of these platforms. And hopefully, everyone will walk away learning a little bit more about what's possible for them in this world.
Okay. So, we're starting with what is kind of the impact on our field. We're going to talk a little bit about the ways that AI is impacting grant seeking. And we're going to start on kind of these real-world impact questions. So for both of you, I'm going to open this question up and start with Margit. I'm curious kind of from your point of view, how is AI impacting the field of grant seeking?
Margit: Well, it is creating a massive buzz. I don't think I've seen so much activity about talking about a particular topic since AI, generative AI came on the scene last year. So, it's definitely affecting us in that regard. I think it's also really highlighting what we do and don't do because this question will probably come later, is AI coming for our job? But, really, when you think about our profession and what we do, it's so much more than writing, it's so much more than AI can do. So, I feel like it's opened up some really great discussions about grant professionals, as opposed to grant writers and our role with organizations in the world.
Rachel: Krista, do you want to add your thoughts to that? How have you seen it impacting our field?
Krista: Certainly. Well, I think it's an opportunity to democratize grant seeking. That has been from day one from what I’ve said. If we can get enough people on board and train on how to use it, then it could make grant seeking more accessible and effective for nonprofits with fewer resources in terms of both money and time. So, smaller nonprofits who cannot afford to hire a full-time grant writer or development director or a grant writing consultant, this is their opportunity to figure out, “Okay, how can we still have a grant seeking program without spending tens of thousands of dollars on it?”
The other thing I want to say is, I think grant makers are only starting to reckon with AI. I haven't seen any of them include restrictions on using AI on anything, probably because it would be extremely difficult to enforce and it's just complicated to deal with. I've also spoken with nonprofits who are submitting multiple grant proposals every day because they're using AI. I'm not advocating for that because I think there just aren't that many grants that one nonprofit would be qualified for. So, they're doing this kind of scattershot method because they know, “Oh, I can quickly and easily produce grant proposals.”
But what I want people to realize is that's not going to get you the results that you're looking for. Generative AI like ChatGPT like Gemini, they're not designed to work that way. So, you still have to. Despite this huge innovation, you still have to have a specific skillset beyond even just the being a grant professional part. It's prompt engineering. So, you've got to combine now two different skill sets like Margit was talking about with being a grant professional, and also prompt engineering to effectively create that kind of shortcut for your grant writing.
So, I think we're all still trying to wrap our heads around it. And it's developing so quickly, new happens literally every single week that we're always trying to stay on top of it. So, I'm hoping that what we have to offer today is going to help you focus in, “Okay, what are the most important things I need to know and keep paying attention to as things progress in order to use this in that kind of democratizing way?”
Rachel: Yeah, I hadn't heard about that kind of accessibility of grant seeking and how this can be a tool to amplify that. So, that's a really interesting perspective.
I'm curious to hear from Margit here. Uin was wondering, “What does AI do best versus what a grant writer does best when it comes to grant writing?” Can you share a little bit of your perspective on that?
Margit: Yeah, I'd be glad to. Well, AI, it's important to understand a little bit about how the different programs work. Let's just take ChatGPT, the most common. And I noticed close to 70%, in your poll had used chat GPT. But very few had used Gemini, which until recently was called Google Bard, but it's on Google's platform. So you have to understand that ChatGPT is really working on predicting what words come next based on billions of parameters of training data. So we know it doesn't inherently have a brain, but it is working on a predictive model based on probability. So I was a math geek, which is why I love probability and statistics and everything else that went along with it. So, I'm fascinated by how these work. And let me just contrast, make a very important contrast. And that is Gemini, or Google Bard, works differently somewhat differently from ChatGPT. It will help you access real-time data by searching Google. Whereas, ChatGPT is not accessing real-time data. So, they both have their strengths.
We could go into a lot more detail about the differences. But what I'm finding is that in order to answer this question of what does a grant professional do best when it comes to grant writing versus AI, you've got to understand what's out there and how it works. So, what I recommend, really, anyone on this webinar we’re doing is I've got -- well, I have a free sheet that helps you develop some AI prompts. You may also want to dive deeper and get into prompt engineering. I took a prompt engineering course, through Coursera. I can share that link, Rachel, in the resource sheet you mentioned. I'll do that afterwards.
But you want to get used to asking the right questions because AI is very limited by what it is you're actually asking and how you ask it. And that's why prompt engineering is so great, or these free kind of cheat sheets or prompt sheets that we have. So I think, really, what a grant professional does best is help with the planning of the project, thinking through logistics, also having the personal relationships. I know when I consult with clients when I do my coaching, we talk about the types of foundations that are out there, which government grants might be best for them. And we really talk strategy about how to approach funders before applying and how to follow up. And so, it goes so much more beyond the writing. But it's really project planning that matters. And that's where I think grant professionals have a massive role.
Having said that, if you enter the right prompts, AI can help you do that too. I'll summarize this answer. The question is, I feel like AI is really just limited by your own creativity and willingness to try out different prompts to see what works and what doesn't.
Rachel: Yeah. It's really learning how to use the tool right and use it correctly.
Margit: It’s a practice.
Rachel: Yeah, yeah.
Margit: It’s just practice. And I talked with -- I was on a recent -- I've taught several webinars on this topic, and somebody said, “You know what, Margit? I just don't have time to learn one more thing. I'm already so busy.” And I get it. As nonprofit professionals, we're already so busy. And so it's good to think about this as not having to invest a lot of time and learning something new, but rather getting almost -- getting instantaneous ROI return on your investment. You spent five minutes on ChatGPT asking for a letter of inquiry or a few other things, and you'll have already saved hours of time.
Rachel: Yeah, that's great. And we'll dive a little more into those time saving techniques in a couple of slides here. Krista, Saranjit was wondering, to what extent would you rely on AI for your grants? How reliable actually is it?
Krista: Well, I think this goes back to what Margot said, which is you have to know the AI that you're working with, and what its capabilities and what its deficiencies might be. So, for example, when ChatGPT first came out, 3 and 3.5 could not access the internet. But now if you have ChatGPT, plus, you subscribe for 20 bucks a month that accesses the internet, it pulls in external research and integrates it. So it's really important to realize like what -- you have a thorough knowledge of the software or the AI that you're working with and understand how it's working. Because if a year ago, I had relied on ChatGPT for giving me information about news, or things that were relevant, like most recent data on something for a grant, it would have been wrong. And it would have been hallucinating or making guesses based on data. But now, it can access the internet and it can pull that information in and I can, with one click, click on it and say, “Okay, here is the URL that it pulled that from. Let me go confirm that this is accurate, that it's relevant, that is a credible source, and maybe even go in and look at some other sources that it's cited.”
So, things like that are really important. So, you've got to understand how it works. And I really do now lean on ChatGPT a lot. And to me the writing quality is closer to what I would expect from myself and the way I can train it and the way I can create my own GPT and use that over and over so I don't have to feed it the same instructions. To me, that is really setting it apart. But now, I've got to learn what Gemini can do because it just came out with its new version. And now we have to see what limits are there and what's not there. So if that's -- and back to what Margit said, if that's like an added task to you to where you're like, “Oh, now I've got to do this one other thing,” there's some upfront investment in learning how -- what limitations are with each software with each AI.
But once you have that, it comes so much quicker. And like Margit said, you can do an LOI in 30 seconds if you already have the information together.
Rachel: Yeah. Yeah, I think the custom GPT building and some of the kind of newer features of more advanced ChatGPT versions make a huge difference. And I don't know if folks are familiar with that even as an option. But if you are subscribing to GPT 4.0. Right? You can end up, yeah, creating your own GPT isn't even uploading information up there like your previous writing. I saw someone mentioned that in the chat to kind of give it more of your voice. So, yeah, there's a lot of things, I think, to make it a little more reliable and make it feel a little more like your own.
I want to talk a little bit about navigating some of these concerns that we have around AI. I think a lot of us feel my senses positively about the benefits and the ways we can use AI for good. But there's also, of course, some things that we should be aware of when we're using these platforms. And so, I wanted to start with this question here. And this is for Krista. Wayne was wondering, “Can I be sure that using AI is not plagiarizing? What are the ethical implications of AI?” And Jason and Mary Beth also had similar questions here. So, I know we've had a few folks asked this question.
Krista: Yeah, I love this question. So, this kind of goes back to what you were talking about, Rachel, is having that foundational understanding of how AI works, right? So it is not saying, “Okay, Google, give me a grant that looks like this,” and pulling it and putting it on your screen. It is a large language model that is trained on 1.8 trillion parameters. ChatGPT 4.0 is. So, it was fed a ton of information. And then the open AI trainers, that's the company that created ChatGPT are going in and testing it.
So, for example, you could put a picture of a tree in and then ask it what color is the wood of the tree? So most of us will be able to identify that, right? So the team would say, “Okay, whatever the AI tells us, is it correct or incorrect?” And it teaches it. If it says green, and it's not green, it's gray or brown, then it says, “No, it's gray.” So, it learns that way.
So when it's feeding you information, it's never just pulling something at found on the internet and plopping it on your screen. It is taking information that it has actually learned from a huge data set in all these parameters, and then providing a human-like answer. So with that being said, I want to say, if you then copy and paste someone else's text into ChatGPT or give AI a website and say, “Mimic this website in terms of style and voice,” to me, that is what is like plagiarizing.
So when I work with AI, I always train it to sound like me by feeding it my own original writing, telling it to mimic my voice and its response. Because I do believe that the closest thing to plagiarism is saying, “Sound like this person. Sound like this website. Sound like this text.” So, I think there are ethical ways to deal with that. And that is by getting it to learn how you want it to sound based on your writing.
Rachel: Yeah, I think we don't think about the fact that we can train it with our own, right, like our own writing, our own things that we've already generated. And it's just kind of enhancing or using, yeah, our voice to continue to learn and share out.
Kind of a similar question, but maybe different take here. Yvonne was curious, “Margit, how do you address bias and maybe confidentiality with information that you're using in an AI or in LLM?”
Margit: Yeah, this is a really important question. And I'm going to address the confidentiality first and specifically your data privacy because I've been looking at the chat coming through and people are concerned about data privacy and what they enter into the chat bar, whether it's ChatGPT, or Gemini, or Microsoft Copilot. They're wondering, “What's happening to this information I'm putting in here? Is it training other data,” et cetera.
So just to keep it simple, there's one key rule that I always, always follow. When I'm using AI and I suggest it to everyone I meet. And that is, do not put anything in that chat bar that you wouldn't feel comfortable sharing with your co-worker. Okay? So, let's say you and I are working together on a project to offer after school programs to youth in the foster care system. It's okay to talk about the mission, the youth you serve. It's okay to talk about your location, the project. But you don't want to share any private information about the youth you serve, or let's say you're looking for specific donors. You don't want to put any kind of financial, confidential financial information in there.
So instead, you can use substitutions, like instead of naming a foundation, you put an XYZ as your substitution. And then when you generate your content, you put the actual foundation name back in there. So really, I would simply be careful. I love using AI. But you can't a hundred percent trust it. And so, definitely, think about what you're putting in there. But it would have to be something you're comfortable sharing with someone.
Addressing bias is really a very different kind of question. And there's been a lot of experimental research on bias in earlier versions of ChatGPT, and even current versions. And, yes, there's definitely gender bias, racial bias, marginalized groups. And the reason being is these platforms are developed by humans, which inherently have bias whether we realize it or not. And so it is trained on, again, billions of parameter of data that was also at some point developed and entered by humans. So, it inherently has bias. That's why.
And I know we haven't said it yet. But this is really useful as a starting point AI. So in other words, whatever content you get, you always have to read and edit. So, yes, it will definitely save you time starting documents, getting drafts. But, of course, you're going to be reading through and making sure. As Chris has said, it's going to sound like your voice or that it's accurate. You're going to check the references, make sure they are real, that they're not hallucinations. And for me, very near and dear to my heart is that it is using language in a way that I want it to so that it isn't further marginalizing people with disabilities, black, and brown people, LGBQT, and other populations that are already have had enough of that. So definitely, definitely, you want to be careful with that.
Rachel: Yeah. I think editing, making sure you're not just blindly copy-pasting anything that's coming out of your AI platform of choice is a really great note to highlight here. Krista, Trey was wondering kind of where do you see are the biggest limits of AI? If you can kind of synthesize that into one answer.
Krista: So, again, I think it goes back to what you're using. But the limits of AI are what they are today, and they could change tomorrow. But so far, what I've seen is, it's not able to, at least in what we have access to, hold a ton of information and design complex programs, multiple programs within a nonprofit. When you think about as a grant writer, all the knowledge that you hold about whether you're working with different clients, or just the nonprofit you're on staff for, it's just this huge ecosystem of information that you have to know and understand to be able to effectively write the grant. So, I think that is a major limitation of AI for grant writers.
But I think with each iteration and each new release, we get closer to maybe not having that limitation anymore. I’m not trying to scare anybody. But, for example, I'll talk a little bit about custom GPTs. So if you have that GPT plus subscription, you can make your own CustomGPT. And so, I have one that's called grant writer that my students can access. And they can use it and it has all the knowledge base of all of my coursework built into it. And it has custom instructions so that it will answer you in a way that will get you a good grant written.
But you can also do that. Let's say you're a consultant. You could build one for each client, and then it will have that knowledge base for each client. You can put their materials in there. Now, again, that goes back to privacy concerns, data concerns, of course. The thing with GPT and the thing that I so far like better than Gemini is that you can opt out of it using your data for training. And within your GPTs, each different one, you can keep it out of that as well.
So for me, I think that the limitations of AI are only today. And I think that there's just such a big future for it. And we just need to keep paying attention and keep updating ourselves on it. And then maybe one day, we won't really have to write grants at all. We'll just be doing grant management. And I think there's even the capacity out there for it to analyze an organization's data. And then grant maker’s award data and saying, “Here are all the ones that you should apply for,” like even beyond Instrumentl goes through right now, which is enough. But that's one thing that I'm looking forward to in the future.
Rachel: Yeah, yeah, the limitations are only what's today. I think that's a really interesting kind of perspective on this.
Our last question in this section is from Terry. Margit, Terry is saying that they have not broached the subject of using AI with their leadership team yet. How would you recommend they start that conversation?
Margit: Yeah, I think that's a great question. Because when we did the poll, initially, a lot of people were very new to AI. And so maybe you need approval or confidence to approach people about this. And really, there's no harm in just starting with a free version. I know we've talked about paid versions of ChatGPT. But start with a free version because there's really no investment except your time. And then make sure you get some -- we have prompt sheets at Grants4Good that we can offer you for free. And just start working with some of those prompt sheets. Because soon enough, you'll get results that you can easily share with your coworkers, with staff, with supervisors that will convince them that this is something you really should be exploring.
And frankly, it's kind of like the internet. If we had resisted starting with the internet back in 1993, where would we be now? So, it's just going to be important for all of us as professionals to be familiar, at least right now on a basic level, with how to use AI because things will get more complex over time.
Rachel: Yeah, that's a great point.
I'm going to take us into this next section. I want to emphasize there's different ways we can be using AI to enhance our grant writing. It may not replace all of the things that we may be needing to write, right? But we've got some good tips that can help us think of ways to save time and enhance your writing. So, I'm going to start with this one from Venetia. Krista, Venetia was wondering, can AI manage the grant writing process from start to finish? What are some of the most efficient ways to use AI for grant writing?
Krista: Well, I like that she used the word “manage,” because I don't think it's quite there. I think it can assist the grant writing process from start to finish. So, I like to think of ChatGPT as an any AI that you use as your writing assistant. So, imagine you hire a grant writing assistant. What would you ask them to do? So, you might ask them to analyze an RFP or NOFO when you get like an 80 Page RFP and you’re like, “Oh, great. Now I'm going to read through this.” You can upload it to ChatGPT. There's a little doc thing on ChatGPT Plus, or you can upload it to Gemini, and ask it specific questions.
I have experienced some resistance lately with it asking it to summarize it. It will be like, “No, I'm not allowed to do that.” But if you just keep asking it in different ways or asking it specific things like, “Here, can you tell me when the grant is due? What are the requirements? What are the reporting requirements? What are the eligibility restrictions?” And just keep a paper, a document on file that has the list of things you always ask. And then copy and paste it in there, and it will give you all the information quickly.
So, I use it for that. I use it for program design ideas, to draft outlines, to even do budget drafts that people have no idea what to do for a budget to do research and incorporate research into statements of need and program design, and then even to revise and edit. So somebody said in here word counts and character counts. I often use it to reduce for length even though it's not able to meet a specific word or character count. You can say reduce the length of this by half and it will come something close to doing that.
Rachel: Yeah, that's a great tip. Like, “Hey, I just need to cut this down a bit. Can you make this into 150 words, or whatever?” That's been something I've definitely used.
Margit, Chase was wondering, beyond ChatGPT, what else is out there that you're using? I know we've mentioned a couple of other platforms. What do you find has been most effective for your grant seeking?
Margit: Well, there are literally thousands of companies starting up all the time that have developed their own GPTs, per se, which is GPT generative, pre-trained transformer. So, there's a little trivia question for you.
So, there are so many of those. And I think the key is not to get overwhelmed with all the products out there, but focus on one or two and get very comfortable with using those. Find out what they can do for you, instead of trying to tackle 5, 6, or 10 different platforms. So for example, if you're doing a lot of work in graphic design and developing ads, why not try Canva? Canva has its own generative AI version now with the plus, which I use. Or in our field, if you're focusing on writing, I have to say, I really like what ChatGPT generates. I think it's very good in the writing, possibly better than Gemini, at least as of today, this minute.
I also use Gemini a lot for things like researching funders and funding searches. So, I like the real-time data and I like what it generates for me. So I just -- I really mostly focus on ChatGPT and Gemini. But I don't put on blinders to the other things out there. It's just a matter of what my current usage is.
Rachel: Yeah, that makes sense. And feeling like you can be an expert in at least one or two, right? And then having the flexibility.
Margit: Yeah.
Rachel: Yeah, to try.
Margit: Yeah, invest in time for one or two at first.
Rachel: Yeah, that absolutely makes sense. Okay. And my closing question for both of you, before we transition into our next section here, what is one tip, or maybe two -- again, aside on one, that each of us should practice to get the best content out of AI? And maybe Krista if you want to start us off and then I'll popcorn it back to Margit.
Krista: I think I've given probably my best tips so far. So, I'm going to give a fun tip because -- I just learned this yesterday from OpenAI as head of developer relations. He was doing a podcast. And he said, “If you put smiley faces in your prompt’s emojis, you'll get better responses.
Rachel: That is fascinating. Did he say why? What is it about that that prompts the AI to get better responses?
Krista: He didn't say. But I assume it's some kind of emotional component even though it doesn't have emotions. It's meant to be like a human and it wants to know you're having a friendly conversation.
Rachel: That is so funny. Okay, I love that fun fact. That's a great one, Krista.
Margit, do you have a tip you want to share?
Margit: That is really funny what Krista shared. I'm going to try that now. Because when I first started using ChatGPT, I was treating it like a person. I said, “Please generate a one page.” And I would say “please” and “thank you.” And then after a while, I'm like, “Well, I guess I don't have to do that. It's a robot.” But then I heard, “Well, it is better if you use good manner.” So, I don't know if there's evidence for that. But I think it's a pretty -- it's really fun to do, anyway. And this way you keep your manners when you're working with people.
Okay, here's my top tip. So in order to make this work for you, after today's webinar, get out a piece of paper or get in a Word document, whatever you use, and write down. Don't even overthink it. Just write down a whole bunch of things that you don't want to spend your time on. Okay? So, let's say you've got a whole bunch of thank you letters to send out to donors and you don't feel like doing it. Or you don't want to revamp your annual appeal letter, if that's something you do. Just write down a whole bunch of those tasks. And let play around with AI to let it do the work for you. And that's what's going to make you think, “Wow, this is worthwhile.”
Rachel: Yeah, that absolutely makes sense. If you really break down all those things, right, that you could have this tool --
Margit: At least a few of those can be taken right off the plate so you can then spend your time on things that require the more personal touch, the things you love or are passionate about, or so you can be in Colorado right now with your family skiing.
Rachel:Now that's the hack that I need to –
Margit: Yes. We all need that little R&R, right?
Rachel: I love that. Well, thank you, Margit and Krista, for answering. These were all pre-submitted questions by our attendees. We are going to move into a live Q&A.
But before we do that, I'm really excited. I feel like I'm like Oprah or Jimmy Kimmel, or something right now. I'm bringing on a very special guest, surprise special guest that's going to be talking a little bit about some hot off the press behind the scenes info about a new feature we're working on at Instrumentl. So, I'm going to go ahead and invite Angela to come on up. And I'm going to remove my spotlight so we can all see her.
Angela has spent a decade trying to win and award grants. She built to the first version of Instrumentl in her Berkeley basement apartment to make those jobs easier. And today, she leaves new product initiatives at Instrumentl and like our new grant writing AI tool. So I'm going to stop the screen share so Angela can share a little bit more about this tool, and then we're going to transition back into our live Q&A.
Angela: Great. Thank you, Rachel. And hi, everyone. Looking back now, that first version of Instrumentl really pales in comparison to the Instrumentl we know today. Thanks to the incredible people on our team, like Rachel, Amelie, Will, Tracy, and the 40 other folks that make up the Instrumentl force today.
I'm really happy to be here with all of you AI and grant enthusiasts. I feel like I'm in really good company here. And I think you know what discussion really brought up were some amazing good points. There are major drawbacks and risks with using generative AI for grant writing. I'm sure some of you have received job cover letters, or you've received emails, or maybe you've reviewed grant applications that are written by AI. And they can be pretty bad. They can sound like they're written by a clueless robot. And that's not effective for grants.
Instead of using AI to write creatively, AI can also be used to save us time on the less creative, less fun tasks. So imagine, instead of you spending your time doing something like visiting the funder’s website, gaining access to their grant application, then copy and pasting all of their application questions into a format you can use in Google Docs or Microsoft doc, we can have AI do that for you in an instant. AI, give me all the application requirements for the Walmart Foundation Food Safety grant and put them into a Microsoft doc. Done. AI can do that for you.
Or instead of you retyping the same answers to the same or similar questions from funders over and over again, “AI, how did I answer this question last time?” AI can resurface your most compelling and related writing at the time you need it. So, what's possible and most effective is not to have AI write a creative proposal narrative on your behalf. Plus, that's kind of fun for you, right? But AI can help you save time in those less creative, less fun jobs, too.
And Rachel, if you wouldn't mind sharing the next slide?
Rachel: Yeah, absolutely. Let me get that back up for you.
Angela: So today, I'm really excited to announce that we built AI to do those tasks I just mentioned. It's a new AI grant writing tool built into Instrumentl that helps you submit high-quality grant proposals faster. So no more copy and pasting and reformatting funder application questions. We collected all of those for you saving you time. And no more rooting through your documents to find the last time you answered this question. Or to find your best mission statement in order to find your best proposal narrative. We resurfaced your most compelling and related content when you need it right inside the proposal where you're working, saving you time.
And if you want to train AI on your content only, as our speakers mentioned today, you can do that on Instrumentl. Instrumental’s AI is trained on your content and only the content that you select. So this is the first time I'm announcing our new AI grant writing tool to anyone publicly, which does make me a little nervous. But I thought today was the right time. Because if there's anyone I want to share this news with first, it's grant writing AI enthusiast like all of you. And we really want you to be involved.
So, Rachel, if you can go to the next slide. Thank you.
If you are already an Instrumentl subscriber, we invite you to get early access to our AI grant writing tool by clicking on the link that Ryan shared in the chat. Just share your email, your name, and a few grants you're thinking of applying to in the next few months. And then I'll personally reach out to you via email, answer any of your questions, and give you a demo of the product for you personally.
So the steps would be become an Instrumentl subscriber, tell your account executive or customer success manager that you want access to our new AI grant writing tool. We would really love to have you. So thanks, everybody, for your ear. Get access to that new tool I mentioned via the link that Ryan shared in the chat. And let me hear your questions in the chat if you got them. Thanks, Rachel. Thanks, everybody. Back to you.
Rachel: Yay. Thank you for joining us. This is so fun. I love that we got to do this special kind of a highlight with a new tool. I'm going to add back our speakers here. Awesome.
If you don't have an Instrumentl account yet, you can make one right now. This link will be available in the follow-up resources as well. But you can go in really quickly and fill out your organization information. You'll get a chance to explore Instrumentl. And then when this AI submission tool comes out, you'll be able to access that as well. So I encourage you to scope it out. We're really excited. And for those who are current users, we look forward to having you help us beta test this new feature.
Okay. I'm going to move us in to our live Q&A. We've gotten through so much shade. I'm actually going to stop my screen share so I can just have us chatting the three of us. And I'm going to start by jumping into some of our questions from our audience members. So Sarah was wondering, how can I create more unique content with AI? And Chelsea had a related question, which was, is it generating the same answers for everyone using it? Can someone kind of chime in and share what their thoughts are on that?
Margit: I gotta share something funny about that. It will not create the same answers every time. I taught a webinar last month about using AI and I shared some prompts. And somebody said, “Margit, I used your prompt exactly to the letter and it got a different result.” I said, “That's because you used it on a different day, or even within five minutes, you can put in the same prompt and get a different result.”
So, there's a good thing about that is that if you don't really like the first result you got, you can ask it to reword it. You can be specific about how it should change the result. Or you just check in another time and you'll see something else. So, no, it will not look the same every time, which is also why one of the concerns I've heard about, will grant applications all look the same? No, they won’t for that same reason.
Rachel: Yeah, that's a good reminder. Right? We all are putting in our unique take on even every prompt.
Alicia was wondering -- and I think, Krista, you said. You might have thoughts on this one. Have you had any conversations with grant makers? From the grant maker perspective, does a grant writer using AI make your proposal less personal or feel like a cheat or a shortcut?
Krista: I haven't heard anything like that from grant makers. But I'm sure that it exists. If you are not being a really great prompt engineer, if you're just giving it very little, you're going to get very little out. I will say that I talked to a grants administrator at a university and an R1 university who was inundated with all of these grants from faculty wanting her to help them with them. And she was like, “They're all so bad,” because they're just putting the questions in ChatGPT and getting answers in its bare minimum. They don't know how to do the prompts. They're not giving it context. They're not giving it the right instructions. And so if you put in a sixth grade level question, you get out a sixth grade level answer.
So, that's what I think people will notice, if it's very basic. And that's why doing things like asking it to revise, giving specific guidance, sometimes I'll point to specific sentences and say, “I like what you did in the sentence. And here's why. I want you to revise the rest of it to sound more like that.” Or I'll feed it my own writing and say, “Mimic my voice and style,” and that helps to be a strong writer to begin with because you know how to instruct it, and how to evaluate what it creates.
Rachel: Right, that prompt engineering, which is kind of a term we've been dropping a couple times today. It’s so valuable. And the more you practice that, right, the better your responses are going to be.
I have kind of related question from Nan that I thought was interesting for us to explore. Doesn't AI contradict the personal connection that we are constantly told is crucial to getting the grant? What are your thoughts on that?
Margit: I think it actually helps free up some of our time to making that personal connection. And one thing I do with AI is I experimented with social media posts. And they did not sound like me at all. And they contain 10 or 20 emojis, way more than I would ever use. They had words like unlock and unleashed, which I don't use in my language.
So I think in terms of personalizing and not losing that authenticity of who you are and how you communicate, it's really important to, again, use it as a starting point AI. But be sure you get your own authentic voice in there. I've even used it to coach my clients on conversations with funders on what points to bring up based on that funder’s mission and past projects that they funded. But of course, they're not going to be reading those points verbatim. But it does help them develop a closer personal connection. So, yeah, it can be used to enhance it.
Rachel: Yeah, absolutely. I love that. Some of those words that come up and you're like, “Oh, that's not me at all.”
Margit: Yeah.
Rachel: Krista, do you want to add too?
Krista: Totally agree. And I love that you said that about unleash and unlock.
Margit: Right.
Krista: …something like that. I'm like, unleash and unlock is cheesy. Please don't be cheesy.
Margit: Yeah.
Rachel: That makes me think of -- there's a question from Garrett that was -- how does one humanize the responses GPT produces? Or I think like some variant on this question would be like, how do you get it to not say unlock or unleash? And you can literally ask it like, don't be cheesy. Try and sound more blah, blah, blah, right? What would you suggest for kind of that question there?
Krista: Look up some adjectives for writing style or voice and then say, “Okay, what is an adjective that fits what I'm looking for?” And instruct it to be like that. And even again, given an example.
Margit: Sometimes I'll just say, “Make this post more professional sounding,” or just different tones, humor, maybe depending on what you're aiming for. But even then, I'm very, very vigilant about making sure that it's still reflecting how I would communicate.
Rachel: Yeah. Yeah. I mean, you can even ask it to kind of take on a persona, right? You've talked a little bit about feeding it your examples of writing, or if you want to say, “Hey, you're an expert copywriter that has knowledge about social psychology,” whatever it might be, right? Making sure it has the context, rather than just putting in a question and expecting an awesome answer. It takes a little more work than that.
Margit: I think, too, as a rule of thumb, anything that you're sending out to people who know and trust you, it really should come from you. I mean, when I write my emails, they're personalized. That's not ChatGPT doing the work. You know what I mean? Whereas if it's a grant application, a letter of inquiry that can sound more business professional, that's fine. I'm not going to be -- I'm not talking to someone so much one on one from more on the authentic voice. So –
Rachel: Yeah.
Margit: Yeah. I think it's important not to lose yourself in this stuff, but use it as a really helpful tool, which it is. Yeah.
Rachel: We have time for one more question. And I think we've referenced some of these resources already that we'll be sharing after the event. But Trey was wondering, are there any AI templates or kind of some some prompts that we can use to create documents? For example, Trey was mentioning, I would fill in data and then AI could generate something like a newsletter, for example.
Krista: I have prompts that I do. I have two courses, write your grant in one hour using AI and intro to grant writing with AI where I have prompts there and I update them all the time. But I also gave a link to Rachel to my seven-day learn to write grants with AI. And it gives you prompts and tells you how to use them. So, it's an email sequence and seven days long. It's free if you want to sign up for that.
Margit: And I think between the two of us, Krista and I have it covered. I also have prompt sheets. You can contact me for those. I teach webinars. There's one on my events site right now that's actually on how to research funder. So, a little different topic that we've talked about today. But if you find that you're at this dead end of always contacting the same funders, then you can take my on demand webinar that's on the grants4good.com events page.
And I can't wait to see who wins my All About Grant Writing course. But if you aren't the person who wins that self-paced online program, you can -- I'm running a promotion where I'm doing small group coaching sessions. So, we would be working together in small groups for the next three months on using AI in this. So, I hope I get to work with many of you after this.
Rachel: I love that. And I love that our community is coming together to continue to learn about this tool. I think this is a topic that will not just be today, right? As Krista mentioned, this is something that is constantly evolving and will impact us, I think, for many years to come. So, I'm encouraged by this conversation. It makes me want to have more opportunities to learn together. I thank everybody for joining us and for asking great questions.
If there are some things we didn't get to today, I'll make sure to save those questions in the chat. And I'll put those responses up on our event resource page so you can look back there. Thank you so much to Margit and Krista for your time. I really had a lovely chat with you both today.
Margit: Thank you for the invitation. This is a wonderful topic to talk about. And I'm so excited with all the great engagement we've had in this and the fantastic turnout.
Krista: Same. Thank you.
Rachel: Thank you so much. So, I've got that feedback link on the slide. Here, you can scan the QR code, or Ryan's going to be dropping that feedback link in the chat. It'll also come in your follow-up email if you miss it. So, no problem at all. And this is where you can sign up to get some pretty awesome stuff for today.
As a reminder, when you submit your feedback form, you'll get access to our freebies. These five tips for utilizing AI and grant proposals and reports, and access to Krista's course grant writing with AI simplified, which is going to be a free training. And you'll also get an entry into our raffle. Margit’s All About Grant Writing course is our raffle prize for today. So just by entering in your feedback, you'll be getting an entry into our raffle.
I'm also excited to share the upcoming events that we have planned this month. You can hang out with me next Thursday, or I'm going to be talking about some of the ways that you can save at least 15 hours per week with smarter grant tracking strategies. That's going to be on February 29th at 2 pm. And then the following Wednesday, we're bringing in Amanda Day and Kimberly Hays de Muga of HayDay Consulting for a fun and informative session on decision making while prospecting. So, that event is going to be at 1pm. On March 6, that's Eastern Time.
Our events tend to fill up really quick. So, I do encourage you to sign up for these right now while it's fresh on your mind so you don't miss out. And the link to register will be in the chat. I see Ryan just dropped that in there.
As a reminder, we have this event Resources page. Make sure to check back here for all those chat box links and poll responses. Anything that was mentioned today, I'll make sure to add back in. And if you don't see something and you'd really like to, there's a feedback form at the bottom of this page where you can let me know if I'm missing anything.
That's all for now. Thank you so much for showing up today. We greatly appreciate you. I'll hang around for a second and answer any additional questions. But I'm going to invite our speakers to say goodbye. And anyone who needs to scoot, thank you so much for joining us. I want to play a little --
Margit: Thanks for putting on these programs, Rachel, Angela, and Ryan. We appreciate it. It's a great public service.
Rachel: So glad to hear it. Thanks for being a part of it.
Krista: Thank you.
Rachel: Bye, everyone.
Margit: Bye-bye.
Rachel: Bye.
%201.png)
