From learning about AI to application for marketing (webinar recording)

This is the second session in a 4-part webinar series exploring the transformative potential of AI in the Newcomer-serving sector, with a special focus on how AI can enhance our services and operations. In this session, Dan Kershaw, ED, Furniture Bank, talks about becoming AI Augmented - the Furniture Bank's Integration Journey.

This webinar, hosted by Peel Newcomer Strategy Group with support of Digital Transformation committee of the Executive Council of Peel-Halton Settlement Partnerships, is designed to share knowledge and discuss the potential impact of Artificial Intelligence.

The Furniture Bank has shared resources about how they're implementing AI:

Why Furniture Bank Has a Responsible AI Manifesto
"Furniture Bank’s commitment to ethical AI is rooted in our desire to leverage technology in a manner that aligns with our values and enhances the lives of those we serve. Our AI guidelines serve as a moral compass, ensuring that our use of AI is transparent, human-centered, and accountable."

How AI Art Works?
How Furniture Bank created AI photos of poverty without harming those they seek to help

Using AI to avoid "poverty porn"
Instead of trying to capture the realities of furniture poverty through a traditional camera, Furniture Bank used Artificial Intelligence (AI) to create their images.

How we created AI photos of poverty?
How Furniture Bank continues their digital transformation by applying AI tools for a bigger impact on their Annual Campaign

Machine-Generated Transcript

What follows is an AI-generated transcript of Dan's presentation using Otter.ai. It may contain errors and odd sentence breaks and is not a substitute for watching the video.

Marco Campana 0:00
Welcome everybody, so I'll be your moderator for today's session. My first duty is to introduce Dan, but I'm also going to be handling the Q and A portion. So I'm throwing a link into the chat for slido.com where you can post your questions as Dan is presenting and during the Q and A at the end. Prefer if you could post them there. If you do post them in the chat, I'll add them to slido because it also allows you to vote up other people's questions, so we know kind of what are some of the priority or more interesting questions, or repetitive questions, even that you might have. So Dan is the head of the furniture bank, where he has been, and I'll let him do a more fulsome introduction of himself in the furniture bank. But he's here because I've been following the work that he's been doing at the furniture bank around AI. They are one of the nonprofit pioneers in Canada who are looking at, how can we use AI as a critical tool, right? So instead of being intimidated or overwhelmed with with AI, what could it do for us? How could it help us in our work? How could it provide efficiencies? How could we harness this to make our work better or easier? They've gone as far as using it in a marketing campaign, which I'm sure he'll dive deeply into today, but also created a responsible AI manifesto, and that word responsible is an essential part I think, of what we want to focus on today is that, how do we use AI and really technology more broadly, ethically, responsibly, in a client centric way, in a way that works for us, as workers, as organizations, as community leaders and for the communities we serve. So I'm going to hand it over to Dan, and without further ado, Dan, welcome. Thank you so much for joining us, and I look forward to learning from you.

Dan Kershaw 1:31
Thank you. So I thought I would spend today trying to convince 100% of you to lean in and start playing with AI, because when every place I've been using generative AI since the summer of 2022 which makes me supposedly an expert, it just makes me earlier. So I wanted to share some slides to pull you through, sort of my personal journey, but we're applying it in furniture bank, and then help bring it into your own lives, and some easy ways to really repeat what I did and what my staff do, because we can all learn how to use AI in a way that fits our particular needs, in our particular roles at any point in the day, all day, Every day. So I've been with furniture bank. It's my 11th year, and I'll talk about what furniture bank is and the like. But we've always had technology as an enabler for us to be able to do more, to reach more to support so I want to dive into that. In theory, I'm going to successfully share my screen, and I'm happy to answer any question about AI. It's such it's AI is like talking about food. It is such a vast topic, and there's a lot of misconceptions out there. So, and I use the term being AI augmented. You'll hear all sorts of things out there about AI, replacing and doing all these things, but I really look at for our sector, the use of AI is very much about using it in a way that is not just responsible but beneficial for the work that we all do. Um, so without further ado, here are some factoids you may or may not know. These are recent stats from 2024 and it more or less says all of you. I'm going to be binary and some a bit controversial. None of us are using AI in a meaningful way. Two thirds of us say it's a lack of familiarity that we feel stops us from using it. 40% say nobody was educated in AI. 80% lack an AI usage policy, and yet we're all seeing an increase in need. I know my organization seeing a decrease in resources. So that's the reality that we all live in, and I know we can talk about the challenges that we face, but my point of view is we as organizations and individuals can do business or nonprofit, or however you want to define yourself, as usual, or you can join me and Marco and a few others in this weird way where maybe you triple your impact or quadruple it. In my case, over the last two years, I've been journaling how much extra free time I've created using AI. I'm finding every week I'm getting about 45 minutes back to do other things now that could be just read a book, or could be do more stewardship, engage. Management, whatever you want to do, but you multiply 45 minutes a week by 52 that is an immense amount of time applying a tool that, for the most part, is free, and sitting on your phone or your computer and you can use today and now. And I want you. I'm assuming many of you recognize what this is, the infamous scientific calculator. Now, if we were in person, I would say, show of hands. Who had one of these? And many of us did. And the follow on question is, how many of you used? You know, there were lots of scientific calculators. It had lots of capabilities. It required humans to work, and it gave us answers. Many were wrong, many were right, because if you did, don't use it properly, it's it was feared at the beginning. You know, we had protests in the streets about calculators coming into our world, most of us stick with the black buttons. One plus one equals I was one of the few people who said, What are these magical buttons? Because in a show of hands, some people were like, Oh, I really, really like the log button. And this is AI, is that if you're just going to stick with the black buttons, you're going to think, AI can't help you. So when I look at AI, it has all the similarities that I saw as a scientific calendar, many, many different models, an immense number of capabilities well beyond what a calculator used to do. It can see, it can hear, it can talk it like. The number of ways that you'll be able to apply it in your uses are all possible. It requires humans to make it work properly. It will also give you correct information. It will also give you wrong information. It's feared at the beginning, and I suspect it's going to be very accepted by the end. So I personally believe that every single one of you can start using AI today. I believe you can become very familiar with generative AI. You may not be Marco and radically changing the infrastructure of your organization with AI, but you can certainly be comfortable with it and know where it will help you, personally, in your career, or even at home. It's smart enough today. I can take a picture of my fridge and say, What can I make? And it will actually tell you. It will give you recipes. Again, if you're a chef, you'll have to determine if it's really good or not. Most people say they don't have a policy. I'm going to give you a free policy to start with. Already exists, and then you customize it for your organization. And I really do believe that you can start the journey of figuring out where you want to augment your own organization. That's what I believe can occur today, if you are curious, and if you will start the journey of learning, because it's not hard to learn. So I call it homeschooling. I'm going to point you to some free things to do or a book to buy, and then you just start, and you're starting the journey that I started two years ago. I'm no different. I just started earlier, people say it's going to be expensive. It's not all the tools I personally use on my day to day. For me as an executive, cost less than a coffee a day, and in most cases, they're free. So I want to start with why furniture banks, why I believe from the organization I'm accountable for is AI important. So here's what furniture banks issue is that is an AI generated image, by the way, because we all know housing is incredibly hard to find. We also know it doesn't magically become furnished and equipped and ready to welcome a home. And it's one thing to help one family at a time. It's another thing when there's 10s of 1000s of people who need furnishing. So I'm going to walk you through three truths. Every year, North America destroys 12 million tons of furniture and Home Goods every single year in landfill and incinerators, including things that are sitting in boxes ready to go. Furniture is the third most expensive purchase that all of us will make in our lives, preceded by a house and a car. Currently, when we do the math, if you were to walk into retail and have to furnish an apartment, you're looking at over $8,000 of free cash. You need to actually equip an apartment from nothing to fully equipped. Anybody know any social programs that have that type of money to furnish and equip a house? No, I. Uh, what our research is showing is that we have millions of people who have housing but do not have furnishings. It is a form of hidden homelessness. Furniture poverty, squalor is another good word for it. So furniture bank is an organization. We're based in Toronto. We service predominantly City of Toronto, but notionally anywhere. But we're very focused on that transitional, social, affordable rental housing, trying to make sure that when families are moving, that they're able to get those goods. So we have this term called homing. Housing is empty. With the goods and a bit of transportation, families can create homes. In our case, the cycle time is we're going out to Jessica's home to remove her sofa and dining room table. Kept bringing it back to the local furniture bank in Etobicoke, families are coming in, curating their home based on the community and corporate donations that we've received, and then we're delivering it back out to them the next day. So the cycle time for Jessica's donation of impact is 72 hours here in Toronto, the time we see with our clients is very short. When the family comes in and they're selecting they're there for about an hour, and their delivery team is delivering for 45 minutes. So we were a moment in time,

Dan Kershaw 11:34
and it's we know we're doing good, because we get the anecdotal feedback, but it's when you meet a mother who, 10 years ago, got a home from furniture bank. And she tells me, this is a personal story, that because of having a furnished home, she was able to keep her family together. Because of keeping her family together, here she was in a grad program debating other grad students on the importance of furniture. So our posture is, if it's a million people in Canada and there's 12 million tons of furniture being destroyed, how can we help more people better you? Lots of photos, and it is family of one, family of 13 every there is no group that is not affected by this type of poverty. So our posture, and it used to be, how can technology at furniture bank get more furniture? But now it's how do we use AI to really augment the work that we do to help more for get more furnishings into more homes in Toronto, in the GTA and across Canada. What I mean by Canada is because we have a lot of technology in play, we've actually built a model that other charities can plug into our infrastructure and essentially spin up their own furniture bank and allow us to help them with that collection of donations, and let them work with the agencies and the families to distribute those things. So if that is of interest. I know Jessica and I are having sidebars about Peel region, but if anybody thinks boy burnishing homes is real pain, but boy, wouldn't it be great if we can have a program at scale that has some funding that goes with it, we have a methodology that it might fit, and would love to talk to you anyway. I digress. So our journey, you can all do it. The only thing that can stop you is if your board or CEO Ed says no, and then I would be worried, because AI is here. It is not going away. It was like arguing gravity away. It is here, and it is changing our world as we speak. Today is the very worst AI we will all use today. Tomorrow will be better. So for me, in 2022 it was that time of the year, it was like annual campaign fundraising. How do we do it now, in the type of work we do, how do you do you show ethically, the squalor and pain of a child sleeping in a bed of clothes? That's what they call poverty porn. That's exploitive. So we've never been able to actually show what it looks like. And yet we had pages upon pages upon pages of stories. And in July of 2022, there's this thing called mid journey, which would take words and imagine what those words meant. And we came up with. Uh, this campaign called the picture isn't real. The reality is and we took the stories and the words that the clients gave us and gave their words power and let AI visualize, well, what does it look like when your child's sleeping on the floor? And we ran our campaign around that. And for us, it was It started as the fundraising How do you get in front of very busy donors who are receiving solicitations from 88,000 charities across Canada? And for us, it was also a way of bringing the stark reality of the type of poverty that's behind closed doors. So we did actually created 40 different views images. Imagine a imagine pictures, but driven from the words of the clients that we support. If you're interested, there's a QR code that will take you to the full gallery. And what that led to was US busting open a mindset I've always had. If anybody's put up with listening to me, they already know I don't think like a traditional charity Executive Director, whenever I hear they say or we've always done it this way. That was always Dan would go the other way. That's me. But when I looked at what AI was starting to create and create the potential for, I started to see AI being able to really augment the way our sector works. So, you know, I have that desert over on the left hand side of that image because, you know, it's for the amount of money you need to spend to use AI, even in the most basic but profound ways, is peanuts. It's not a lot of money. You can make it bigger, and I'll talk about that, and I know Marco talked about it last week, but you can AI, augment everything you do in your organization in sensible ways that help you help more. We created a manifesto. Now let's be clear, the manifesto is a policy. I created it for my staff, for them to have a working guide, and I'm using this metaphor of a jungle. It is a jungle. You can quickly get lost in it because of how fast it's changing all across the world. You have to intentionally want to follow it, or stick to a guide and have a group people to work with. So it is a working guide is intended to give my staff empowerment, but also accountability for how they'll do experiments. We want to be very ethical and transparent when we're using it, and all of these things are we've also taken policies that were already out there and adjusted it for what is a furniture bank and how do we need a we will be using AI and a lot of it, like any policy is giving the ground rules that if somebody asks, How are you you can answer most of you don't have an AI policy. You have dei policies, you have harassment policies, you have list all the other policies you have you need today, an AI policy because like it, or lump it there are, there's shadow AI. Use, what is shadow? Ai? Use that you mean it's your co workers who are bringing their own AI to work because they've already figured out it's a huge productivity tool and are using it at work without management actually articulating how they want to use so there is a a risk by not having a policy and engaging in this conversation to those of you who are EVs or in places of are accountable for that. We I have it as a web page. It's a manifesto. We evolve it. I'm a marketer by training. Manifestos are far more interesting than policies, but it, a lot of it is a guide that you know, we're evolving it, and we continue to experiment. That's important you're experimenting, because the use case that each of you will have that AI, improves your personal journey or your organization's journey, are all going to be unique and different. Unlike a scientific calculator, this is where it's different. All of you could put one plus one into a calculator. You will all get two back. But in generative AI, there are so many factors that will be very personal for each of you. So we started to really embrace curiosity. What if we could triple our impact? What if we got new superpowers? What do I mean? Superpowers? What. To create photo parking, the ethical issues of target photographing children and families in squalor, the cost, complexity we could never have done, and yet now I could give our clients power in their own words, and have photos to have a conversation about this type of poverty, a cut costs. We all have expenses. They aren't always people. We can eliminate and automate the most tedious parts of our jobs so we can do higher value things all day, every day. And there's always a conversation about AI replacing work. AI is going to disrupt jobs. It isn't necessarily going to replace them. But on my perspective is we talk about decent work in our sector. We never really talk about making the work more decent in the Decent Work movement, and using AI in beneficial and responsible ways, gives you the opportunity to make the work that your teams are doing better, more creative, more gratifying, because we all have the not fun parts of work, and AI can help augment those things in places. So that's our mindset when we look at it

Dan Kershaw 21:21
easy places you can all start with images, if you need to illustrate, if you need marketing, if you need social media, AI can help you today. You use Canva. It's there. In fact, AI is jammed in all of the tools you use today. Doesn't mean it's helpful, but they're there opportunities for video today, I can take text, turn it into a video, and have those videos translated and over to 200 languages. The old way, go shoot a video, then go translate them all. We would never afford. We would never do because operationally, it would be too complex. Now you can all text can be translated into words. Text written, audio and video, all available, all ready to do in words. This is in furniture bank's case. We don't have dedicated fundraisers. We have a grant writer who supports me, and then we work through the grants and AI augments, working through we all, if any of us, who are fundraisers on the call, we take the grants, we take what they care about, we take their questions, and then we apply the full This is what furniture bank is in excruciating detail and map what they want to what we are. And that gets us a good first draft that we can then fine tune and refine and then submit. And that allows us to do the fun parts of fundraising and spend less time on the process of fundraising, you can all start, and should start with what's easy, what's low cost, what's low complexity in ai ai augmenting both your own role, but also your own organization. So for furniture bank, now we've just implemented this is where you're starting investing. We have an AI augmented phone system to help with we run a social enterprise model. So we compete with Junkers who are happy to contribute to that 12 million tons of destroyed furniture. So we have an AI augmented phone system that helps simplify coaching. It's following sentiment. It's it anyway, I can take a whole webinar on how it's helping. One of the things I like is I can say, what are customers asking on these calls, and I can actually get an answer, and then the marketer in me says, well, then I can create that training material out on the website and video, so those questions don't have to occur. We're looking at training videos. So the we run us that support service model, that where we support other charities running their own furniture banks, like Matthew house in Ottawa, Redwood Park communities and Barrie oyep in Winnipeg, we provide training videos that historically were very time consuming. Now we can blend screenshots with AI and actually create robust training videos that for a fraction of the effort and energy, time being a big currency. So your own journey is equally possible. All you have to do as of today is do the following. Read. This is my prescription. Pretend I'm a pharmacist. I want you to read one book. This is where that'll cost you. Then I want you to listen to a few podcasts and. Go take some classes. One of them, I've got a discount code for the other one's free. And most importantly, use it everywhere as your constant companion and learn how it can help you personally. It's not like Google search. You can't just put in an answer. It is very much like a a university intern, like a grad level, really smart, but really inexperienced. And if you they are your teammate, your co pilot, your wing person, however you want to frame them. If you have it up and you try different things, you're going to find very quickly. This is helpful. This is not but you have to, as you'll read in the book, you really need 10 hours of just using it before you'll very quickly identify where it is of most impact for you in your organization. So co intelligence, remember, AI, this generative AI is the first invention that is happening all at once everywhere, every other invention in time, it started somewhere in spread. But everything that's happening in AI that's affecting us today in Peel Region is also occurring in Jakarta. It's also occurring in London, England and Johannesburg and everywhere else. So this is an invention and an innovation that is spreading all at once around the world, and it's very hard to internalize. What does that mean to me, as an employee, as a parent, and our role in society? This is the only book I've come across that is a accessible My rule of thumb is my mom will put up with me and read one of my books or watch one of my videos, and she'll tell me quite quickly if it's too nerdy or geeky. My mom read this and she thanked me for it. So this is an accessible book that will help give you some constructs on what is happening around AI. It's not about necessarily in work. It's about AI in its totality, and it just came out in April, so it's still current and accurate. Podcasts, they're free. These are the ones I personally listen to. If you're going to do just two the AI daily brief, it's 15 minutes a day, and that at least gives you the headlines of what's going on in this weird, massive world. Just so you feel that you have a sense on trends and directions and things like that. You want to get a little deeper everyday AI. The only podcast that's really talking about how AI is being applied in the nonprofit space that I've come across is fundraising AI, and they will bring on different guests to talk about different things. Just listen to them and soak it up. That's all you need to do. Free Training. October 1, fundraising.ai. Is putting on a free summit all day. They are bringing the who's who of all the AI experts that are it's fundraising with the dash of nonprofit. Again, most AI conferences are more about business. They aren't really interested in responsible and beneficial AI. They're about profitable AI. So this is embedded very much aligned with the issues and themes that we care about. So that is free all day. There's a fellow Tim that I he and I have met over the over the last couple years because we have a similar journey. He has created an AI for any one course $99 he's getting me an affiliate code, and I'm passing on the discount onto that. So I'll, I'll get that added. It's a one hour boot camp that's intended for all of you that if you've not started, it's just getting you, your feet wet, your toes in the pool and start using it, because it is just that accessible. Fundraising.ai. Has a free nonprofit AI policy, you could literally copy, paste, put your name in, and your least got a version one. Like all policies, they change, but you at least have one that becomes a conversation point with management, but it is perfectly fine and generally fundraising.ai. Is always providing podcast conferences and other opportunities for learning how to put. AI to work for use, if you're looking at bringing AI into the organization. I didn't put it on here, but it very much starts with whoever is in charge. I'll pick on Jessica. If the Nelly doesn't believe in AI, then AI goes underground. But if Denali says we United Way greater toronto will become a AI augmented organization, then there's a whole cascade of things that you can do. Look at why you would want to augment your organization with AI. There are certain things that AI will naturally lend itself to in your own organization, and it will be different than your next door neighbor, most conference, any panel that you get on, generally as Microsoft or Salesforce or Blackboard, or You have consultants and platforms and softwares. Who their Why is, well, they will say, beneficial, help the community. At the end of it, their Why is, sell something. So be mindful of who you're hearing from. What is their why for talking to you about AI,

Dan Kershaw 31:20
go find your fellow AI curious co workers. They're out there, but they're also probably afraid to mention, like, I'm AI curious, but we at furnish bank, we created a small group. I know there's like three or four of us that are more on the dabble experiment early. And I have some co workers that when you figured it out and it just works then, then I'll join it, and that'll be the same with everyone like pivot tables in Excel, there's a handful of people who are like, it's the best thing in Excel ever, and others who are like, Oh, I'm afraid to touch that button. It's okay. Find those who are curious and start the conversation, and then start very small, low risk experiments and share that with your coworkers. I wish I could capture the look of wonder and delight that we all have. I am sure Marco had it. There was a moment in time where Marco was like, Oh my gosh, I've giggled sometimes by the magic that is possible when you put it to use in your own environment. And that's part of it. Just share and discuss and then look for the bigger opportunities. Because if you go and read what we all remember this during the blur, there was this company called moderna. They did some vaccines and stuff. They've gone all in. They're at the other end. They've made the whole organization all AI. And if you want to see, you know the case studies out there to say, case study AI moderna, they've gone to the other end of the journey, which is everything's AI to give them the ability to create better vaccines faster. That's their why. So there is a huge continuum. Just like food, there's Kraft dinner, and then there's nine course meals. It's all food, and AI has that same complexity, and a lot of it's for you to just start your own personal journey and then start to apply within your own organization. And I will go to questions, because this is where the fun begins. I Yeah,

Marco Campana 33:44
awesome. Thank you very much, Dan. So I'm going to repost the we've got one question, but I want to, I want to editorialize a little bit before we dive in, if that's okay. And one is to say that when, when Dan showed the the scientific calculator and you started positioning yourself, was I this person who focused kind of just on those low buttons or on on everything? And I can say to you that I was never. I had friends who use those scientific calculators and used all of those functions, and I did not technology intimidated me in high school. I had friends who were coding. We're talking about the 1980s here. People, people were coding in early languages. It was a it was a complete black hole for me, and what I proved to you is that you can learn this. So if you're wondering, I didn't use that scientific calculator in terms of those rows. I wasn't one of those people. You can become one of those people, in part, because the technology has become easier than ever before, right? Dan,

Dan Kershaw 34:41
yeah, because building on that. Thank you. Marco, so back then it was we had to push push buttons and hope we remembered how it worked. What we're finding today, this is research that's a week old, is that those people who are in fine arts who can actually communicate and articulate what. They need. They're the experts. Because if you give AI, you know, please write me a haiku about refugee settlement and Peel Region, it will. It'll be a bad poem, but if you gave it a more fulsome conversation about all of the AI will do it so your interface isn't memorizing manuals. Your interface is being clear on what you need. And that's why I'm that's I guess, why I love getting on these panels is you can all start it, literally today, for free, and you start your journey, and you'll have your aha moment, and you'll have your giggle moment, you'll have your OMG moment when you look at what will this do to if you have kids? So this is like the invention of electricity. It is a disruptive invention. We can't change that. It's here, but we can understand how it will affect us personally, in our families and in our organizations, and lean in and have ownership of there's a conversation about AI has bias. I flip it around to say, yes, it does, and if our sector gets behind applying our bias to the models is not a good thing, but if we step back from Ai, businesses will happily train their models, and the bias will only get worse. So if we as a sector don't engage with AI, we will be left behind. That's one of my main is if we all get involved and I'll start highlighting how it should change the way of the work happens. Organizations will pay attention.

Marco Campana 36:52
Yeah. So just for folks to know, we're mostly focusing on slido, I'm very happy to do the interaction. If you can raise your hand, though, so I know that you want to speak, then I can call on you so we avoid kind of chaos. There's got 130 odd people here, so the reaction button at the bottom of your screen and then raise hand, and that will let me know you want to speak. But Hanadi, clearly you want to speak, so go ahead and I'm just going to go close the window because I don't know if you can hear it, but I've got construction on my street, so Hanadi, please go ahead, unmute yourself and ask, and I will be right back. Oh, thank

Speaker 1 37:21
you so much, Dan for the presentation. It's very, very amazing. There are lots of fears with using AI, and I noticed really that AI will replace human being will. And this is really something that we keep talking about in management meeting, and we don't see the other side of what AI can bring. But is, is it going to replace human being? But it's still our AI is AI is always being controlled by human being, developed by human being, but we are going to see lots of jobs are really dropping out. And this is the fears of many people, really, of starting using AI in in a way,

Dan Kershaw 38:18
it's a good fear, but let's, let's break it down a bit. So there was a time where this thing called Excel arrived and accountants were afraid they would disappear. And I can run through lots of examples where a new technology arrives and there's a fear that jobs will disappear. The reality is, is that jobs will change. That is a certainty. I have a personal belief that it's those organizations, those individuals, who lean into, how do we integrate AI into our work? They're the ones who are going to help shape how do we want work to work in our organization, there are companies who their mindset is, Boy, wouldn't it be great to have no employees. But what we're seeing in AI output is you need humans to get you need a human in the loop as the throwaway line, you need a human to marketers as a function are likely going to be one of the most affected groups in there, so you may not need as many marketers, or you may need a different type of marketer, but you'll still Need marketers. There are certain tasks that AI does really, really well, translation. I know some of you have social enterprises doing translation as a social enterprise, the headwinds are against you, but that's not because it's out to get you, but it's like if you were a facts maker. And the internet arrived. It's unfortunate, but today we could, you know, go look at I can add a little plugin to my website, and it can translate into a 220 different languages that I as furniture bank would never, ever contemplate translating the old way, because it would be too cost prohibitive. So I think it's it's too strong a word to say AI is going to replace us. Will companies use AI to eliminate jobs? Yes, there will be examples. Will there be examples of AI augmenting and creating jobs, and that's why I brought up the conversation of decent work, is I think AI can provide more room for your teams to do higher value activities and less of the commodity tasks. But it's something to do to be mindful when you start making decisions within your own organization about how you use AI, that's that's the choice of the organization. But I, I wouldn't agree with that. It's all AI will replace us. That's not going to occur, but it's happening. You have to accept it and work within it. I think there's that acceptance piece, because it is AI is here it is. If you have word who has Microsoft, who has Google, almost every it piece you have has AI in it may not be great AI, but it's there, and it's only getting better, and it's only getting more pervasive, so you may as well lean in today to understand it and figure out how it applies to your personal career path, as well as how you could best use it for your organization.

Marco Campana 41:56
Thanks. Dan Priyanka, before I jump to you, I'm going to go to slido because folks have posted a couple of questions there, and I want to respect that this is somewhat related Dan, and I imagine you'll have a similar response. But the question is, what will happen to our human brain eventually, as we become dependent on AI and not using our own brain for solving problems or creating solutions,

Dan Kershaw 42:20
I move on to harder solutions, harder problems. We all use calculators. There's a time where, like, you can't use calculator. You have to do it by paper. So I don't I my last two years, I have AI as it's I have another screen to my rate. It's there. So for things that I know, it can do better, faster, go do it, come back, and the human in the loop, we verify it, and then we we improve upon it in the same way. You know, if we're in a restaurant, there are chefs, there are sous chefs. Like a restaurant has many people, but ultimately create one plate and AI can play that role. So if you let go of being responsible for your own material, then yeah, you will become reliant on AI to do things, and that's why I use the term augment. AI is augmenting Dan Kershaw's impact the way I need it to do it, it's also because it's it's is the most powerful. Even if you pay for it, you all break out a credit card for a buck a day, you have the most powerful technology ever. And what you do with it is up to you. For me, my mission, I'm trying to end furniture poverty. I don't want to see another child sleeping on a floor ever. That's why I use it. So if you're an artist, it's immensely powerful in helping you be creative. If you're a writer, it's immensely helpful in editing. And it'll, it'll, you can actually have it critique you like it's a coach. It can do like it's until you use it and try all the different use cases. It's not, it's not just a calculator, and I use that scientific calculator, not perfect metaphor, but until you start to see what it can do. You you there's a set of questions that are that normally are asked, that you will ask differently once you start using it,

Marco Campana 44:40
and along those lines, one question some and I think you've alluded to some of this, but what are some of the tech basics that people need in order to start using AI,

Dan Kershaw 44:52
a mobile phone with data the internet? I. That's it.

Marco Campana 45:00
I love that, because it's not about tech, is it? No

Dan Kershaw 45:03
if you can write a sentence that's clear enough, your mom, spouse or child, would understand you're good to go if you were trying to build what Marco has built with his organization, it's more than that. But my message to all of you is, if you have a browser, you can go to chatgpt right now. It's free, and you could start the podcasts are free. You can start, splurge for the Kindle 16 bucks and read the book, and it'll give you a better I use the metaphor of a sandbox. It'll give you some frame of where, how the sand all works, the sandcastles you choose to make on you. It's your choice.

Marco Campana 45:54
And I would say it's less about talking more about critical thinking skills, right? We also, we are all becoming prompt engineers. Is the latest term, right, understanding the tool enough to understand how to ask it the question, in the way that will generate some adequate response, and then being able to critically evaluate that response, right? Yes,

Dan Kershaw 46:10
I the we, I did a thought exercise, and it's that we hear prompting. And I just heard somebody explain it the following way. I'm going to try it. We've all gone to a drive through, and I'm going to pretend it's a McDonald's. When we get to the drive through, we say, I would like a hamburger, and we will get whatever McDonald's thinks a hamburger is you could also go, I would like 16 hamburgers, three of them with barbecue sauce and pickles and mustards and this. And I want this and this and this and this and this and this and this and this and this and this and this and submit. And when you get to the window, it will give you all of that. You have to verify they got it right. And if they didn't get it right, you go back and say you missed this. Redo it so it's it's less about so much prompt engineering, so much as being very precise. You You're allowed to be verbose in your instructions. So if you only give it six words. You're giving it six words of instruction. Some of my prompts are literally paragraphs. Step one. Step two. Do this. Here's a document. Read this PDF. Check this website out. It's really capable, and that's why I use the idea of, you know, like a university graduate student intern. They're smart, but if you give them horrible instructions, if you've ever worked with a grad student, they can give you some really bad results, but they're smart enough they can give you some amazing results, but it's all just speaking English, and actually, not just English, it's multilingual too.

Marco Campana 48:00
And I threw a couple links into the chat for folks that there are people who have created formulas or approaches to show, in part, the complexity of if you're Gen in generative AI, in particular, if you're trying to create a really useful response that isn't necessarily sourcing other links and things like that, but they talk about, establish the role for the AI, create the right write the action you wanted to take, provide additional context. And then and then say, what is it that you want me what? What output do I want you to see? Do I want to caption? Do I want a script? Do I want a workshop agenda, you know, things like that. And then you can continue to kind of pepper it with questions. And so those formats or those frameworks can be quite useful.

Dan Kershaw 48:43
And what I find my most common word, my prompting is rewrite and then clarifying what I want it to change and it will re roll. I use the metaphor of rolling dice sometimes, oh, that was horrible. And you'll say, rewrite and change a voice, or add this, or learn this, because there's no penalty to it. There's no and I find I actually learn a lot in reviewing what's being generated forces you to be critical of what you're reading. But that is important, that is you must verify what it creates before you can trust it.

Marco Campana 49:21
Yeah, and you've created, I put in the chat the recipe you've provided on on your site about creating AI photos, which I think, again, shows that there is a knowledge and learning curve to this, but it is. It is not about tech. It's about understanding how to use tools. And just like you're using Microsoft Office and Canva and whatever tools you might be using, this is just another tool, and I think that's useful for people to understand when, and answering that question again, we've gone down a large rabbit hole. But I think it's important for people to realize that your experience as well as I mean, I was looking at your your your LinkedIn, educational background, and it turns out we're at U of W at the same time, actually. Oh, cool. Um, but you're, you're not an engineer, right? Oh, I'm a geographer, mathematician or computer science professional, and I think that none of us are, I mean, I'm a poli sci major, right? So it's like, you know, we don't come at this as techies. We come at this as, how can I use this tool?

Dan Kershaw 50:12
You didn't know Tamara Cook, did you

Marco Campana 50:16
Tamara cook? That sounds familiar, but that's poli

Dan Kershaw 50:19
sci at that time,

Marco Campana 50:20
probably newer than my wife. It's a small world,

Dan Kershaw 50:26
but, yeah, it's it's so easy because you just are interacting with it in the same way we all have. We chat, we chat with people. That's effectively what we're doing. We're chatting with an AI agent that has many superpowers, if you pay for chat GPT, it can see, it can talk. I did. There's a little experiment we're trying a furniture bank where I built a GPT, aka, think of a Excel macro. Again, I just explained in English what I wanted it to do. There was no code involved. And I said, when you start ask what language the client we're serving is speaking, and when now going forward, when you hear Urdu translate it to English, and when you hear English translated to urgent. So when if you came to furniture bank, the volunteers will guide families through the showroom to build their home based on the donated goods. Historically, we'd have a large portion where we couldn't speak the language, given the number of languages that are out there. So we're now experimenting, and that same little chat GPT tool is able to speak Urdu or French or Persian or pick a language. Is it perfect? Probably not, but it's a whole lot better than no translation services, and that was free.

Marco Campana 52:02
Yeah. And being created the Right Sector when it comes to that, because we know that machine translation is not perfect, and in some situations, like during covid, we learned it could be quite dangerous, in fact, with technical English. But in the case of when you're there with somebody and you're face to face, this is an Augment, as you said, an augmentation tool, right? That whether using Google Translate or chat gpts, machine translation, you're, you are you can check in. You can establish it's not a it's not a document translation that someone will be using on their own, isolated but you can actually read each other's vibes. Yeah,

Dan Kershaw 52:36
and it's important mind, the motto at furnish bank is progress over perfection. We often in our sector get paralyzed to Nope, it's not perfect, not going to do it. Nothing happens. I would much prefer to imperfectly translate all languages when you come to furniture bank and being mindful that it's not perfect, but it's better than because we're never going to have you know how many funders are like, give you money for all of these things? We're seeing a generosity crisis. We're seeing funding, government funding, going down. So we're looking at AI as creating more capabilities that we've never be able to afford with our own operating so and again, you try it, and what if it's not quite right? Go back into the GPT and give it more detail in English. Again, no programming required. You can all make gpts to repeat processes.

Marco Campana 53:32
And going back to your earlier answer to the question about losing jobs, this is, this is augmenting someone's ability to provide service and not turn somebody away or not, but it requires a human in the loop, and that's what's also very important. It's not like maybe one day, but right now, it's not that someone goes to pick up a tablet that furniture bank has provided and said, enter your language and then interact with their AI chat bot. That may well happen one day, but someone is there intermediate. So,

Dan Kershaw 54:03
and it's contextual in that we are looking at experimenting with on the furniture donation side, there's a large portion of people are like, hey, they'll do it for free. No, we don't pick up furniture for free because we don't sell the furniture, give it away. So we're looking at having an AI phone agent who will talk to Jessica one, Jessica Two, Jessica three and four and five and interview. What do you want? What are you looking for? And for the Jessica who understand this is a paid service to come and professionally remove your furniture and home goods that we can pass them straight on to a human agent to have a donation call. And for those who clearly are like charities are meant to do things for free, we can send them to the education side of things and not waste our very valuable humans. That's That in itself, is 50% of. Our call volume is people thinking we're we're charities will do it for free, and we're obligated to.

Marco Campana 55:07
So along the lines of augmenting someone, posted something in slido, which I want to read to you. So I have read this quote, AI won't replace people, but people who use AI will replace people who don't.

Dan Kershaw 55:16
Do you agree? Or what do you think I generally agree, and I think it applies to organizations as well. There's a conversation that I hear often. I'm 11 years young into this sector. I come from a for profit background. The problem is worse in the US, but the number of it's a commentary. Imagine Canada. And have talked about is that there are a lot of organizations out there, many of which are trying to do the same thing. So there's just, there's a lot of organizations chasing after the same resources. And I believe it'll be the those organizations that lean into AI and demonstrate, you know, in a impact per dollar, etc, they're going to be the winners in the long run. If you sit on the sidelines and wait, it'll be too late. As I said, today's the worst AI you'll ever use, and it changes weekly. Marco, I don't know how much you try and keep up with it, but like, I was there when the internet started it, it changed, but like, literally every week, it is moving at speeds that is hard to stay on top of. That's why the podcast is really important.

Marco Campana 56:34
That's a great segue to another question in in the in on slido, which is, is AI tech evolving so fast that it can be hard to keep up.

Dan Kershaw 56:41
Yes, 100% so you're going to find areas that are of interest to you personally, that you you know, you can add it in on social media, follow them on LinkedIn. I'm I'm a former marketer, and there's another podcast I didn't put on there, but it's every week. It's a one and a half hour deep dive on, like, really, you know, aggressive use cases of AI in marketing, communication, social media and all of those things. So there are given your interest, there are going to be specialists that will provide the detail I haven't found for our sector as a whole, a similar set of conversations, because not a lot of organizations put their hand up to say, I self identify as being AI augmented. They've done surveys in the UK. I haven't seen here in Canada, but I assume those in Britain are similar. Is that they actually want us to use tools and be professional and be, you know, as productive and efficient as a for profit. And they don't, you know, I've not had a single donor like, How dare you use this new tool. I've had a number going, Wow, that's amazing. How can we help? How can we support or I'm picking you because you're being innovative. And I guess that's why I like being here. Is like you all have the right to be innovative, and this is the first tool that's free, and it's just start dabbling. Marco's self admitted. I'm no nerd, I'm now, and you just started. So you all can be where I am and where Marco is, but you have to start. And it the effort to start is so low, there's no excuse than other than being afraid. And you don't need to be afraid about this. You need to be informed about it, which is easy to do.

Marco Campana 58:51
You bring up a really good point about funders that came up in our last session, actually, which is, should we be asking permission? And obviously the answer then and now is no. But where are funders when it comes to this. Have you been so they're encouraging you. And in our case, IRCC is the biggest funder, and they have, in fact, funded a few projects with AI that are AI related, and most of them, as you know, initial, innovative projects to learn from. Have you been asked by funders help us understand how we can support the funding of AI, is that something they're looking to learn more about?

Dan Kershaw 59:25
I'd love to say it's often right now, if you go to grant connect and type in the word AI, you'll get zero search results. So there are no funders today in Canada that explicitly say yes, we would happily encourage you. I am working with counseling foundation and some foundations out of Quebec who are interested in supporting and that was me teasing them, which is, you funders love restricted funds. Why don't you force a restricted fund of $2,000 or something that you have to use? Ai? I in certain ways, read the book, play with Po, try perplexity, like, really low, low risk stuff. Because once you start, you can't, like, you're like, oh, it's moving from and math to calculator to Internet. Like, you can't go back, because it's just so such a productivity boom. In my case, I just embed it. You know, we're allowed our admin. I also put a impact measurement tax, and I put a AI augmentation technology tax, plus the program. So I don't ask for permission. I just say this is the way it is in order for my organization to remain relevant, productive, Best of Class marketer, these are the things that we are doing. So be bold. You have a right to be bold. You're changing people's lives. You should be bolder than business. I

Marco Campana 1:01:01
like that. Don't ask for permission. Just tell them what you need and build it into your budgets.

Dan Kershaw 1:01:07
I've yet to have anyone say no, I'm not saying I'm N equals one. I'm not I'm not statistically significant. But I've never had a funder push back on asking so put it in there, because you don't need a lot of money to to bring it to your wider team. You don't need a lot of money to experiment outside of yourself. Excellent. There's

Marco Campana 1:01:31
a couple questions I'm going to combine a little bit, because they I think they're related. So one, you've you've talked about in your presentation, which is the potential for bias and discrimination and misinformation and and related to that. So as a AI generated answers pull from its their chosen sources. So how can we help AI help to provide current information and links to Main and up to date sources? Yes,

Dan Kershaw 1:01:57
good question. Great question. So it's two parts a as a sector. I'm talking to Bruce at Imagine Canada, we as a sector have to aggregate, I'm going to call it our biases, and make sure they get to the models that are being trained. But that's a super nerdy Marco, and I will talk about it offline activity. But for your immediate use cases, the AIs are always being trained, always, every day, they get bigger, they get smarter and better, and they're always looking for new material to train. You'll see in chat GPT, if you're using it, if there's a button, you know, can we use what you're doing for training, most people turn it off. I'm like, Heck yes, you're meant to know what a furniture bank is. You're meant to know what furniture poverty is. You're meant to so the bias is there. It's not a bad thing. It's it's that the information that is out there has everything that's out there. And to combat that, when I'm using it is I'm uploading documents that are internal or I'm pointing them to websites. So you combat the bias by applying your own bias. And because that's the funny thing, AI is super excited to meet, make you happy, and that's where it'll hallucinate, if you're not clear. But if you say, please refer to these sources and what you know, it'll refer to your website or your long form, 50 page explanation of what is your organization that you can't give to a funder. It'll use that as a foundation for its answer, so you can avoid bias. But it like, don't treat biases as a swear word. It's just, it's like search results. You go to Google, there's an algorithm, and it'll give you a result of the top for a reason, there's bias. So it's, that's how I use it day to day, is I'm always attaching links or documents to help ground it, to make sure it's staying in the zone that I want it to.

Marco Campana 1:04:13
And increasingly, chatgpt didn't do this initially, but they are now providing footnotes to source links of where they got their material, a tool like perplexity has always been doing that, for example, yeah, so, so, yeah, you can you, it does. And you should check links, because it, it provides misinformation from some of those facility times it'll i Yesterday, I was using perplexity, which I find is a really good tool, because multiple ai, llms, chatgpt and Llama and Claude, I think so you can choose and see the different nuances in the responses. Perplexity usually is pretty good, but I clicked on a link, did a quick search for the data wasn't there at all in any form that they had quoted and very I mean, AI is very confident, right? Yeah, please. 19.3% of these people said this. And it's like that. There's nobody said nothing about nothing. And you have to check that, right? You have you. You know this is the someone described this to me as the AI human sandwich. You start with the human What do I want to put in? What do I want to generate? What facts you know that those what do I want to get out of it? Let ai do its work. And then you go, step back in and say, Now, how am I going to use this? But how am I going to evaluate it,

Dan Kershaw 1:05:24
right? Yeah, like, if you haven't been to perplexity, it's you, it's, I don't use Google search as much anymore. I use perplexity. So go to it Marco, I'm sure you'll put it in there. But perplexity.ai, if I'm meeting somebody new, put their name in, and it'll give you this beautiful summary with links like a stranger, I'm an introvert by nature. I like I I'm not gregarious in the least, except when we're talking about things I find interesting. Perplexity can give me a breakdown. Talk about this foundation, talk about this location, talk about anything, and it'll aggregate it from all sources, not just what Google thinks is the most important answer. The other tool that I've I use, I don't know if you've tried it, Marco is called Oh, P, o, e.com, and Marco was talking about llms. I use the there are all of these different engines, and I use the metaphor of walking down the supermarket in the junk food aisle, there's all the pop. All these llms are just pops, different flavors, different carbonation. They're just different. Oh, on one subscription lets you access all of them. And the most interesting thing is you can start with Google, ask the ask it, stuff, upload documents, whatever. And right after it say, Would you like to try that with sign, which is a different model. And then you can follow on. And what you'll very quickly see is some models are really good for strategy. So you're trying to come up with strategic plan for the next three years, or at this moment in time, Google, Gemini is amazing at bringing together. You can upload all the documents you have in your organization, feed it into Gemini and say, Please come up with a hype, three hypothetical, three year strat plans, and it'll do it. It'll do it better than if you give it to another so po.com that's, you know, one subscription. But you can try everything, and you're going to find some like Claude. Sonnet is great for writing beautiful English, far more the most human English that you're finding. You can find other models, like Facebook's it's linguists lousy. So they're all going to have personalities that you're going to gravitate to. In the same way, we have our favorite kitchen tool, we ever favorite. This is how when you use AIS, it's not a single tool. It is many tools that you're going to come across and use at different times.

Marco Campana 1:08:12
And that's what the importance of those podcasts, those books, the fundraiser, fundraising AI. And for folks who are thinking, well, fundraising, AI, I'm not a fundraiser. It's here in part now, Dan's here in part to show you we can learn from outside of our sector and sector adjacent. And if you're wondering, part of the schedule, for example, for fundraising day, ai, ai, skills and roadmap leading your nonprofit to success. That's not fundraising related. Let's see here. Ai, from the top down, a fundraiser handbook to strategic integration. I guarantee you 80% of that is a handbook to strategic integration, not just fundraisers. Here's a really good one, which is run in part by Mina Das, who, you know Dan, I'm sure she's great in BC, and does a lot of work around data, equity and AI Equity and Human Centered AI. You can listen to her on my podcast where I interviewed her, but she navigating AI and data equity in nonprofits. Here's what we learned from a survey of more than 700 nonprofits in the US and Canada that's related to our sector, from data to impact leveraging predictive and generative AI for targeted philanthropy. Well, again, 80% of that is just, how do you leverage AI for prediction and generative AI for targeted? Well, you want to do targeted service delivery. So they're targeting people that they want to donate. You're targeting clients you want to serve. The techniques and the frameworks and the basics are all the same. So I think this is another thing that when you look at some of these things and you think, that's not really me, it is you. It is you 80 it was there, and I guarantee you 80% of the lessons will be useful for you. So it's something to think about when, when we talk about, oh, this sector, or that sector, or a word that doesn't sound like your role, like you're not, it's not settlement practitioners.ai Summit, although maybe we should have one, right? It's and this is kind of it. This is the first one that this four session. Reasons with PNS geo is the summit. But it's important for people to step out into these other spaces and realize that they're just as relevant for you as well.

Dan Kershaw 1:10:09
Thank you, Marco, that was a perfect use case of perplexity.

Marco Campana 1:10:13
Absolutely. Oh, and by the way, I oh you. You checked out for the link to perplexity. So for people who haven't gone in the chat, I asked perplexity, tell us about Dan, so I do have a question Dan, like, do yes, do you have, do you really have an eclectic career spanning over two decades? I love the language, right? It's like, it's very it's probably your own LinkedIn profile, right? I'm

Dan Kershaw 1:10:34
a geographer who went into Telecom, who went into Internet, gambling, payment processing. I was part of a collective that cleared 1/3 of the world's internet gambling, payment processing. I ended up at lavalized, doing mobile dating, then web hosting, e commerce, mobile publishing, and then a headhunter said, Have you ever thought of running a chair?

Marco Campana 1:10:55
I love it. So if you're wondering if your path can meander and get into AI, oh, my goodness, obviously, right. Anyway. So it shows us, that's a great example. And if you click on that link, folks, what you will see there is that perplexity sources. Right at the very top, it'll say, oh, here are the, I think, nine sources that are used. And you can actually click through, and then what it'll do as you're going through the answers, it provides these little footnotes to those links, which you can check for. And you should try

Dan Kershaw 1:11:27
it for your own organizations. Try it for yourself. Perplexity is a lovely job aggregating like, Who are you on the internet? If you get no search results and you really want to be known, then there's work to do, or you may see misinformation. They're like, we have to go correct that. So it's, for me, that's, it's a go to tool, professionally and personally, including, which dog breeds should you get for small children who are hyper allergenic and like beaches? It'll give you an answer with references,

Marco Campana 1:12:00
awesome, a practical question, yes. And this is because, you know, we're all most agencies are Microsoft 365, agencies these days. Copilot is the thing, right? So is copilot better than GPT? No difference. Or could you recommend other a

Dan Kershaw 1:12:15
Are there any Microsoft people on the call? No, seriously, no. It's horrible. I'm a power user of Microsoft. I have it and I have there's sometimes where I have, like, four different agents all trying to help me do exactly the same thing. I find it's very powerful, but it's slow and it's it can do things, but I've yet to find it do things as well or as quickly or to the level I want it I want the control. Copilot doesn't give me the control of the voice and the tone and the dictate. It's Microsoft's, you know, just kind of picking on you. Jessica's your photos there. It's like Jessica. I'm going to tell you how to write the email. It'll write a good email, but it certainly doesn't sound like Jessica, and it's expensive. There are no there are no nonprofit discounts. I have talked to them about it, and their message off the record is because of how energy intensive it is. They want to get it right before they make it cost effective for everybody to use. My sidebar is, personally, I would give Poe $20 a month. That's essentially even charging Canadian far faster than because it it like a calculator, can use it when I want to, whereas copilot, it's It's in everything, and you don't necessarily know where it is anyway. I'm sure it's the worst copilot I've ever seen. It'll get better every day, but right now, I would not be renewing my copilot experiment. I've been experimenting with lots of little tools trying to find the right balance.

Marco Campana 1:14:08
Awesome, really good question from someone, how would you correct misinformation on perplexity? And I wasn't sure, so I asked perplexity, and I will share the response there are. That's a really important point. I wonder if you've had any experience with, with, with trying to correct or update information that you find

Dan Kershaw 1:14:24
I have not. So I'm curious on the link that you're going to give me, but yes, this

Marco Campana 1:14:30
is, this is specifically how perplexity says. So they have a flag icon somewhere below the answer, okay, let's see. Oh yeah, there's, where's

Dan Kershaw 1:14:39
the flag? URL, description, examples, oh, report, interesting. Okay, cool. Generally, it's, it's because you can go into the detail, and it's great for articles. Yeah, but

Marco Campana 1:15:01
it's great. It lied to me right away. Users can report inaccurate information using the flagged icon below the answer. There's no flag icon below the answer, man, there's a little three dots. And I can click on report, and then I can do it. So this is the beauty of live misinformation, right? I asked perplexity, and it sourced that with let me see, did it source it? Yes, users can report and it sourced its own support. How do I submit a bug? Okay, there's even an email address, so don't believe everything you read, but, but experiment to be curious, right? Like, I know three dots means more information. And instead of a flag, it had a downward thumb to report something, and you're thinking, I don't want to click on that. I want to tell them it was it was bad. And this is where the tech I find, I don't know about you, Dan, this is where I find the tech annoying, right? A who decided three dots was something I should click on for more information, yes, or three lines, right? Like the iconography on online is infuriating, and I looked it up once the cultural differences for those three dots. In some countries, like we call them the three lines is a hamburger. The three dots some people call like falafel or, you know, a skewer, for example. In different it's absurd, but we have to become comfortable with three dots, okay? And then I click on reports. I don't know what that means. Hopefully it'll open a new window. And in this case, it does, and I can say this is inaccurate, it's too short, it's out of date, it's too long. And I can and then space for me to say, how can the response be improved? Right? So things so you can all of that takes time, but that's part of our entering our own bias and our own information is But that takes time, and that's that's something that I think is really important to realize. Important to realize, if you want to do that and you're training it, not just on your own data, but as you're using the tool, you want to try to take some of that time from time to time. So I will after this, I'll go back in and I'll say in that, I'll report that they actually provided the wrong information according to the current version of their user interface may have been the last version, maybe three versions ago, but it's somehow referring to the wrong information. But all of that requires a commitment. So okay, this is interesting. It might not be as relevant for you, but can you give us some real life examples from your experience of AI, oh, you have actually other than media creation, more like client engagement. So the translation is an example translation. Yeah, right. So live with a person, yeah. And I just want to be clear, like we keep thinking about AI, is this thing that I use on my screen, but what you're providing is a live usage of a tool? Yeah? With somebody the phone, exactly. So let's talk about that. Yeah. So the

Dan Kershaw 1:17:41
if you have chat GPT on your phone, all of you should do this volunteering. All of you get chat GPT, go to your fridge, take a photo of your fridge, what's inside, uploaded to chat GPT and tell what do I have in my fridge? It will do a pretty darn good answer. And then you say, what can I make with this in and pick a cuisine, and it will tell you, if you like drinking, you can take a photo of the spirits you have. What drinks can I make? I've watched somebody open AI who did a video, one which was who was visually impaired, help me secure a taxi. And the AI is smart enough to say taxi is coming. And he ordered a taxi, and a blind person in London, England got a taxi and got in it. So AI is, it is not one thing. It's not a screwdriver. It is. It has so many uses, and a lot of it for me are tools. I'm using scribe to help document processes and it will AI generate all of the documentations. Click on this, click on that. So things that I can mainly do AI is doing for me, I can do video avatars. I'm experimenting that with donors. You know, it is if 1000 people give us $10 Best Practices say I should send back a thank you video. But Dan, spending all day doing 1000 videos is a practically not possible. So I'm experimenting that I can get an actual virtual avatar of myself and say, Thank you. But again, it's experimenting. It's not. None of this is locked in stone. So if you find something that works, do more, if you find something that doesn't stop doing it the you know we're currently the question is, where else are we using it? Inside of the organization we're looking at. These are things we're looking at. So future tense, AI dispatching, so we have trucks all across the GTA, where, everywhere, collecting every. Thing, three sofas and this postal code, notionally, the data is there that AI can help sort it, that if we have multiple locations, because we do know from data that when a family has nothing, we know statistically, if it's family of six, what they need based on old data. So there will be a point in time in the future, not this year, that AI can say, make sure the following trucks go to the following places and offload these so that they match the incoming families for the next next two days. So that AI, there's predictive AI and there's generative AI. They both are predictive AI. It's been around for 25 years. Use Netflix, or half the tools we all use without even thinking about it. It's AI. They just don't talk about being AI. So it's really I could go into any organization and look at any function and point to ways, not that you should, but that you could. And that's where I always go back to the why. Why do you want to automate it? Why do you want to speed it up? Why do you want to change it? If you've got a clear why, and you rank it and like this, why? Absolutely, augmenting our phone system, absolutely is a great way to give donors a better experience and help our staff have better jobs, because it is not fun to be yelled at, like, like, really, it would be considered abuse every day because you're charity charging. So now I can filter that, and my staff can work with donors who understand that we provide a service, and those people who have a bias the charities we all do everything for free and operate on sunshine rainbows, we can point them to other ways that yes, free is the curve. Free is dragging it, taking it somewhere else. We're looking at it in accounting again. As you move down the spectrum, this is accounting. You've got to be really careful with it, but you certainly can apply AI to augment your accounting, but you'll go at the speed your organization wants. My reality, I'm the nerd in the building. I'm not surrounded by nerds. So it's a slow process. So everybody has the book, and not everybody's read it, and I tease them about it every so often I finish the book. And did you learn something? Yes, and we've got them into Jasper for marketing. And now social media is, again, it's not generate and send, it's we call them sfds, a word I won't say, first draft. And then we as humans go, pretty good. Make it better, and then we'll use it, but the amount of time that takes shrinks. So you literally, everywhere you walk in your organization, you can AI augment the function, but you don't need to, unless your organization says that is a profound improvement. On the surface, we're trying to provide that's

Marco Campana 1:23:19
helpful. That's helpful. That's really helpful. And I think that's important for people here. This is a long term process. You haven't resolved everything, you haven't brought everybody on board. You're still working on that, and that's useful for people to know two quick last questions, I know we're almost at the end of time. The first one I'll answer, which is, there are concerns using a as a search engine. Has that changed? No, but it's different now that there, there are AI tools that are revealing sources. You have a better sense of where the information in the sources are coming from. The last one is a big one, but going to ask you to answer it quickly, because it's not up to us to resolve but it is a concern. So is there a collective cost? Are there concerns for the environmental impact of use of AI? There is

Dan Kershaw 1:23:57
the there's a lot of factoids that are out there. Ai requires servers. Servers are expensive, and these are very expensive servers. I can go find a whole bunch of discussions and conversations that this will spark an energy renaissance and that it will force us to be smarter with how we generate sustainable energy, good thing. And then there's other conversations, just like politics, that will be polarizing. And part of it is it's here, the service exists, and whether So, I want to be aware and I want to be informed, and I do follow the different companies how they're actually powering it, and how they're disclosing and I'm sure at some point I'm not there yet, I will start choosing to use this because it is a more environmentally sustainable tool. But it's, it's nice, it's so we're two years, two years into this, and. Entire new industry that has spawned 10s of 1000s of companies across the world and changed everything on its head. So it is something to watch and be informed about. But it's, it's, I guess, be involved in the conversation by using AI and learning because it's already being used by everybody else around you.

Marco Campana 1:25:29
Awesome. I'm going to hand it back to Jessica, sorry for taking this right to the wire, but thank you. This has been fantastic. It

Jessica Kwik 1:25:36
had it was helpful, very helpful. And I think we're seeing that in the chat box, I think Dan, you helped step us through the process of learning like a whole new industry and a whole new, I guess, dynamic in the world that's shaping the future. And really, how do we step up to be informed, how to innovate in that space and be bold. So thank you so much for that. And I think you've given us permission to experiment and and try new things, and, you know, maybe even have more discussions amongst ourselves as we learn together. So

Dan Kershaw 1:26:14
yeah, there, if you all got into it uses, there is going to be next year, that sort of practitioners how to really work within the nature of what you do, there are going to be some very targeted learnings to share, like spells, like, once I know how to do something, I can give it to Marco, and Marco can do it. So it's I'll finish with a Harry potterism. We can all be muggles, but I choose to enter the Wizarding World of AI,

Jessica Kwik 1:26:44
wonderful. Thanks so much everyone for joining. I think it was a great discussion. Thank you Marco for moderating as well. And really happy that you could come. Dan, thanks so much again. My pleasure.