Transcription
[Music] Hello everyone. Great, it works.
Okay, so I want to tell you my story about AI and why I got into it and why I got so excited about it and why I'm on stage talking to all of you here today.
I spent my entire career helping big brands understand the new consumer. When I first started my career, the new consumer was the millennial generation. They were the first generation to grow up in the internet in the household. I'm a Gen Xer. I grew up before the internet was in the household. Which means when I tell my kids that when I grew up there was no internet, they think I'm an absolute dinosaur.
But the millennials were the first generation to grow up with the internet in the household. Then the new consumer changed. He became Gen Z, known as the iPhone and social media generation, otherwise known as those who have the iPhone as an appendage to their body. Bring the P. How many of you have a Gen Z kid at home? Right, you get it, right? And uh that changed the world in incredible ways, right? The power of the internet in everyone's pocket, the power of at the time the iPhone came out really had the power of only what NASA had. And now look where it's taken the world.
And now there's a new generation coming in, generation alpha, the topic of my recent book, Generation AI. Gen Alpha is going to be known as the AI generation. Currently age 0 to 15. Gen Alpha will never know a world without AI. We'll never know a world where they can talk to technology the same way that they talk to people. Think about that. This is such an incredible movement that we're seeing. Yet, for some people, it's the only movement that they're ever going to know.
So, back to me and my journey. I currently run a venture funded software company called Suzie. We have about 300 employees. And when AI, namely ChachiBt, came out, my first inclination was give AI to my engineers and have them figure out how it can help our software product. And I waited and I waited and I waited and nothing happened. And it dawned on me that AI is unlike any other technology we've seen before. It's a new realm and how humans interact with computing.
So I made the decision to take AI as the CEO of the company back under my own reigns. and made the decision that before I was going to build AI for my business, I was going to build AI for something personal to solve a personal problem I had that I knew I'd be passionate about and then take those learnings and apply it to my business.
So, I'm 50 years old. I have young children. And the goal I want to solve personally is I want to stay alive as long as possible. So, I made the commitment that before I was going to use AI for my business, I was going to use AI to solve the biggest challenge or opportunity I see, which is staying alive as long as possible. And everything I've read about AI and everything you're going to hear about AI today is it's all about data. In the case of most of you, it's financial data. In the case of me and staying alive, it's a data on my body.
So, I went through 25 years of personal health information, MRIs, X-rays, blood test information, notes from doctors, anything I can get my hands on. And I pulled it all together and I organized it in a structured way. Blood test, PDF, X-ray, PDF. I've lived a relatively healthy life, but there's things that have come up over the years and markers in my blood tests and all sorts of stuff. And with that information, I built my own personal health bot and I trained it. I told him to act like a leading doctor from John's Hopkins University with the sole purpose of keeping Matt Britain alive.
The first question I asked it was, "If I'm going to die in 5 years, what's the most likely cause?" And it answered, and it freaked me out. And unlike a doctor that might care about bedside manner or hurting your feelings, AI does not care about any of that. And obviously I was like, "Whoa, what do you mean? What?" And you think about that compared to WebMD where you go to it and we all think we're going to die, right? Um, this provided incredible insight that was customuilt for me. It talked about things that were in a blood test six years ago. It gave me dossier I could give to an endocrinologist or an allergist before my appointments so they knew everything there was to know about me and how to actually then dive right into treating me versus spending all the time getting to know me. And it dawned on me that that's really the true power of AI.
Now, there's a lot of learnings here because some of you who work for larger companies are probably experiencing right now that you can't just download any AI tool that you want and upload all your customer data, right? There's privacy, it data security that really encumbers many of you from really diving into AI. So, you're hearing it from everyone. You need to understand AI. You need to futureproof yourself. But how if you can't even use the tools?
So, my advice to all of you is think about a problem you have personally. Whether it's staying alive, whether it's managing your personal finances or trust in the states, whether it's something I did for my mom who's 75 and lives alone in Florida who's calling me three times a week about her Netflix password, right? And now I created a bot for her where she can find all of her password information and everything she needs. Or you can create a family chatbot that allows you to easily share schedules and information about your kids soccer practice. go through the process of figuring out one problem you want to solve, aggregating the data you need to solve the problem and building something.
Now, you're all probably like, "Well, how do I do that?" It's incredibly easy, so much easier than so many things you guys probably solve. Certainly more less complicated than submitting corporate taxes every year, right? And I will at the end have a QR code where I'll send everybody step-by-step instructions. I think the biggest learning I had the whole process is you just need to be focused. So many people are overwhelmed by AI because it's such a powerful tool and they get enamored with all these tools. Figure out one problem you want to solve. Get the data for that problem. Have an understanding of what a solution looks like. And you can go to chat GPT and say, "This is what I want to solve. Give me step-by-step instructions. Do not give me step two until step one is done." And by step 57, you're going to find that you've amassed a skill set that will separate you from the rest moving forward.
Now, not everybody in this room is going to do this, but those of you who will are going to come back and thank me because the experience that you'll gain by building something will really futureroof you. And that's the passion I have right now is I want to share what I've learned about AI with as many people as I can. So, they on the right size side of this incredible change we're about to see because this is bigger than the internet itself, which the millennials experienced. It's bigger than the iPhone and social media Gen Z experienced. It is really going to change the way that humans and technology interact. And I hope today what you take away from this presentation is one a good grounding of what AI is, two what are the applications of AI today that are meaningful to business. And three, what does it mean to everybody in this room? So we're going to dive in and I'm going to share the video and this entire presentation with everybody after the fact. So you're welcome to take pictures or slides, but you certainly don't have to.
So why is AI so powerful? I think one of the biggest misconceptions about AI is it's for the tech whiz kids. The internet was like that. Not everyone could build a website or code. But the reality is to use AI, you just need to have curiosity and a certain level of perseverance and you can be great at AI. This notion of prompt engineering is going to go down as sort of like a BS term of the mid2020s because the reality is AI is getting good enough now where it knows what you mean. You don't need to necessarily engineer prompts. You just have to talk to it just like you would talk to anybody else.
But AI is also improving at an unprecedented rate. Today, Apple's having a big event where they're announcing, I think, the iPhone 17 or 18. And every development cycle the iPhone has is about a year apart. I would argue the leaps that AI has had in the last 6 months is greater than the leaps the iPhone has had since the very first iPhone came out. Think about that for a second. The rate of change is mind-blowing. And because of that, it really does have unlimited potential. The the length of task and the power and potency of AI is doubling every seven months. So many times I hear from people, I tried that over the summer, it didn't work. Or maybe you're saying, I tried to have AI analyze a balance sheet in historical financials, it didn't work. It doesn't matter what it did in the past. It doesn't even matter what it can do today. Because at this rate of change, you have to assume that anything AI does okay now is going to do incredibly well a year from now. And you have to work to where the puck is going and not have these self-limiting beliefs on the past because this rate of change is something that we're not accustomed to. So it makes sense and stands the reason why you would say I don't this hasn't worked for me. Just you need to trust the future of this progression because the data is really undeniable.
So as I mentioned I have small children. And I think a lot about what AI means for the future of education, right? Because we've all grown up in what's called the knowledge economy. And whether it's knowledge of doing somebody's taxes as an accountant or knowledge of writing a contract as a lawyer or knowledge of reading an X-ray as a radiologist or knowledge of coding as a software engineer, your ability to learn and regurgitate that knowledge to people who are willing to pay for it was a tried and trueue path to success during the knowledge economy. Right now, I believe we're coming to the end of the knowledge economy. And the notion of education being about memorization and regurgitation, study as hard as you can, spit out what you learn the next day, and do good on a test, I think it's far less important moving forward. Instead, it's about creativity and problem solving and critical thinking and these skills that we're not really taught at young ages. We are taught to spit things back out. But in the world where we can spit out any information instantly in any context we want, what it means to be successful and powerful and productive and effective are not what it used to be.
And this is going to hit hard for a lot of people. I think we're going to hit a big wave of job losses coming up. We've already start to see it. They say the future's already here. It's just not evenly distributed. But where the future is distributed is with big tech. And what we've seen with Meta and Amazon and these big companies is they're laying off five figure groups of people seemingly quarterly. And I think it's only a matter of time before that hits mainstream corporate America and the FedEx's and the Walmarts and the Home Depot start to see the same efficiencies that big tech has seen.
So what does that mean for all of us? It means we need to be on the right side of it, right? Because whether we think AI is right or wrong, it's here and we're not going backwards. And every technological innovation has had naysayers and fear-mongerers, but we always progress as a species. And this is a progression again like none other that we've seen.
So what is AI? I'm going to very quickly go through the AI value chain. And we're going to start with infrastructure. And really the only thing you need to know about AI infrastructure for today is you should have invested in Nvidia. But uh AI creates GPU chips originally made to power video games. Right place, right time. Just so happens their chips power AI. The infrastructure is also incredibly power hungry, which is why you see companies like Amazon and Microsoft investing in alternative energy sources like geothermal and nuclear to power the growing demands that we're seeing for AI chip uh sets which just companies have an insatiable right now demand for AI compute and we're seeing it basically play out um every quarterly earnings call.
The second part of the AI value chain is something that we probably all know fairly well, which is large language models. And right now, it kind of reminds me of a moment we saw a long time ago. You can turn it up a little bit. Okay. So, if you heard the sound, what he's saying is what is internet? What does at mean? What does dot mean? This is K Cork and Greg Gumble in 1994 kind of really trying to figure out what the internet actually means. And of course it seems crazy now. I'm just going to turn up the sound for my next video. But um it was just incredible how quickly we have evolved and how now the internet is sort of like a utility and running water here in society. The same thing sort of happened the same way AOL brought the internet to the masses. Chat GBT has done the same thing with artificial intelligence. If you look at the amount of time it's taken for chat GBT to get to 1 million users, it kind of illustrates the point that this is a technical evolution unlike anyone that we've ever seen before.
And what chat GBT is is a large language model. A large language model is actually fairly simple. It has an input, which is a prompt, which is basically anything you want to accomplish. There's technology that processes that prompt and looks through scores and trillions of different data points and aggregates it together and outputs or generate something which is why it's called generative AI. It's a pretty incredible technology not that dissimilar to Google where you would search something and Google would search the internet and give you a list of blue links. Of course, there's lots of differences here because it's multimodal and you can have conversations with it, but that's essentially what an LLM is. And now there's tons of LMS and they keep leaprogging each other. They're in a race to get to what we all know is AGI, which is essentially a place where machines get a little bit smarter than humans. No big deal, right? But it seems to be coming very fast. Um, in that note, in just one year, the smartest large language model improved it IQ score from 96 to 136 in one year. So again, if you used it a year ago, you were using a tool that had an IQ score that was 40% less than what you would use today and it's going to literally be off the charts before we know it. And there's all sorts of companies building their own large language models. Meta um has built a model where called Llama where they're open sourcing it and they have a completely different approach. And we're seeing a lot of companies toy with different types of large language models. So there's something called a reasoning model where instead of it just giving you information back right away, it actually thinks through the process a little bit different. Instead of reciting what it sees on the internet, it actually goes through a process of critical thinking to push out the answer which is using used for far more complex scientific type prompts. And of course, chat GBT itself recently announced or open AI, the parent company behind it, ChatGpt 5. And the interesting thing about chachi bach 5 which just came out a couple weeks ago is it decides for you what is the best model for whatever task that you have. So these models keep growing but again I wouldn't get too caught up on the models because ultimately it's about the problem that you need to solve. We all problems we need to solve personally and in business and if you stay focused there the tools you can find become slightly less important over time.
Many are asking what all this means for the future of Google who through the internet built possibly the best business model we've seen in history with their search product. Uh how many of you have used chat GPT in the place of Google for an everyday search in the last couple months? Wow. Right. So a year ago it would have been 5% of the people. It's probably 30% of the people. It's only going to continue over time. And many companies are trying to figure out what this means for the new search engine optimization. So I speak all around the world and what I'm increasingly finding is the companies that hire me to speak are actually finding me through chatbt versus Google.
Perhaps the most important part of the AI value chain is data. Data is what differentiates one company's use of AI from another. And what start to happen is like what we saw with the music industry where a lot of companies are leveraging data that they don't own to power their large language model. And the publishers are fighting back. So Reddit sued Anthropic, which is the company behind the large language model claude for allegedly using their data to power their model. And we're seeing something happening in Hollywood right now where a text to video tool. Midjourney is getting sued by Disney and Universal saying that their data is, you know, illegally fueling the the Midjourney model. So we're seeing how important data is.
And I spoke earlier about how I use data for my health bot. Well, you can use data in all different sorts of ways. Uh, a friend of mine's in real estate. He has agents that are all over Brooklyn and New York City who are often asked, "What's the closest school? What's the crime rate here? Is there new construction that's going to happen? It's going to keep me up all night." So, I built for him essentially an AI agent which taps in something called NYC open data, which publishes on a daily and weekly basis information like crime by city block and air quality and traffic by city block and new building permits. That information has already always been there, but who's going to go through a spreadsheet with millions of different lines? Nobody. But now with the power of AI and tapping into this, his agents could have access to all that data immediately.
The same thing exists for what we all do. And I think the same analogy in people in finan in the financial function organizations is not that much different than my healthbot. You're trying to keep your company alive. That's the opportunity you want to solve for. Instead of looking at X-ray and MRI information, you're looking at balance sheets and profit and loss statements and free cash flow statements. And with that, you want to understand the core insights. What levers does a company need to pull in the math equation of the business? You want to understand the business and you want to make sure the executives and the employees can understand it. That is how the accounting function and the finance function has become more strategic because the thing about AI that I've realized over time is it's far less now about how to solve the problem and it's far more now knowing about the problem that needs to be solved. If you're a graphic designer, you don't have necessarily need to understand the knobs and dials of Photoshop. You can go to an incredible AI text to video tool and just tell it what you want to do, but somebody still has to understand what they want to make and what a point this incredible technology do. So when we talk about the future of any knowledgebased function, it really is about leveling up your thinking. It's really about putting the tools and systems in place where you can easily access the data and then your job becomes far more about what are you going to do about it. And that's how the world is changing for every function because I speak to lawyers and accountants and software engineers and and doctors and all sorts of knowledge based workers and they all have the same concerns and to me not to oversimplify it but ultimately it is that simple. Data powers decisions that drive growth or keep people alive. You need to understand how to put these systems in place and understand what to do about it.
So the part of the AI value chain we all know is the application layer. And the thing about AI is the application layer isn't just text. It's video, it's image, it's voice. It could be data visualizations for the financial information that you spit out. This is a text to uh image model called midjourney. And look at this progression that we've seen with midjourney. Version one was about 19 months ago. Back then the the guy in V1 had seven fingers on one hand and six on the other. And a lot of people who use a then probably thought I can't make images with AI. And look at the fidelity of version 5.1. So tell me what's version seven going to look like? This progression exists in every part of AI including data analysis and me maybe many of you used AI when we were still in version one but we're a long way away.
This is a text uh to this is actually really interesting. This is a text to image creation tool that chat GBT had. And that guy on the left right that looks nothing like me. Um, that's actually not a real person. That person doesn't exist. He's an amalgamation of trillions of different data points on the internet and someone decided to make it. Think if you actually had to create that picture now, the photo shoot that you need to have. First of all, you'd have to find that guy. I don't see anyone in this audience that looks like him either. Um, and you'd actually have to make it. So, that that goes back to the problem that needs to be solved.
Google has this incredible text to video model, right? And you might be like think like why is Matt talking about text to video to a bunch of people in the financial functions? Because if you want to drive decision-m based upon data, you need to actually communicate that in some way. And in the past, it was hard to do, right? So, for example, if you see that people are spending too much on TN, you could very creatively push that out and have data behind it in a way that you couldn't before. This platform, Google V3, essentially allows you to with any prompt have a 4K highquality video about anything you want to make and make it instantly. Now, right now, VO3 is only 8 seconds long, but a year from now, it's going to be 80 minutes long, and three years from now, we're all going to be making our own fulllength Hollywood style motion pictures from our couch, starring you as the action figure. And maybe you can look like that guy. Um, what's going on here? Um, oh, thanks. Thank you. I got it. Thank you. Oh, you're just changing that. Okay. I think we had a sound issue. So, I'm going to go back blah blah blah. You've already seen this. I've been through crazy tech issues, so this is nothing, guys. Don't worry. Back to 50. Nothing to be concerned. Oh, right. That's where the sound worked. Here we go. Okay. And we're back.
So, these text to video models are incredible, and I'm going to show you a couple that I made for this presentation today. I can't wait to hear Matt Britain talk about the future of AI and finance tomorrow. Yes. Or or my personal favorite, free cash flow, revenue recognition, depreciation. [Applause] I knew that one would work well. These were created this morning. Easy to do. None of those people exist. I did tell it to look like a casino that looked like the Arya and it's incredible how easy that is to make. So your ability to communicate charts, graphs, stories is unlike none other. And ultimately if your job is more strategic now to change people's decision, a big part is about how you communicate it.
Voice is incredibly powerful right now with AI. We've all been very frustrated with Siri and Alexa in the past. But I can tell you as somebody whose voice every single day, its ability to understand what you're talking about and transcribe it with incredible accuracy is something that's I think going to change the accounting industry when it comes to auditing and interviewing people. I get interviewed by my accounting firm every year, right? As part of our annual audit over time, that could actually be done by AI and be and they could interview more people, but somebody still needs to analyze the information to understand what it means. So voice is also incredibly powerful.
I'm fascinated with this notion of digital twins. One of these people is real and one isn't. And people are now able to essentially clone themselves and clone their voices. This looks complicated. It can be done in a matter of minutes. So, this is going to enter a whole new realm. And of course, we need to think about what are the implications of deep fakes and fraud on digital twins.
I mentioned earlier this whole notion of the knowledge economy and we're seeing it impact everything. Coding notwithstanding. There are tools out there like cursor and lovable that allow you to actually build complex iPhone applications with just a few prompts. And the impact of it we're seeing is on software development job postings dropping dramatically because in the past you needed the knowledge of writing code to create applications. Not so much anymore. And what these companies are doing is they're building 200 300 $400 million revenue companies with 50 employees. And that's another thing we're going to start to see over time is companies do a lot more with a lot less and become far more profitable moving forward.
So I decided before I'm going to talk about the implications of what you all do is to ultimately show you the real power of AI. And with that, as I couldn't sleep a little jet lag last night, I made a little music video using AI about me speaking at this conference right now to truly show you the power of this technology.
You get the lights from spreadsheets to speed.
The futures unlock.
AI in the mix.
Watch the industry shock.
Blackline soldiers.
We the first to move.
Finance stay sharp.
Future proof with the groove.
AI changing money.
Every legendary plan.
What took weeks to close now gets done in the span.
No fear of the box.
We adapt, we create,
turning, risking the power.
Automating the way.
Matt Britain on stage, vision clear, never late.
Dropping keys to the future.
Helping leaders navigate
from area lights to the world on the screen.
Accountants with swagger running AI clean
from spreadsheets to speed.
The future's unlock.
[Music]
Pretty cool, right? So, so I'm going to wrap this up. Thank you. In terms of what it means for all of you, what are the applications that you really need to understand so you can really drive impact in the finance industry and in all your roles? Because many of you are incredibly accomplished and spent your career learning skills that now are starting to see, oh, they can be automated or AI can do it. I do believe that you can use those skills to propel you if you understand how to combine those with the power of AI. And that's ultimately the challenge, but you're not going to be able to do it until you build something in AI. And you're not going to be able to truly build something in AI until you build something for yourselves. And that's been my experience is it's going to be incredibly hard for you to build something at your organization unless you're running a onewoman shop, right? But if you actually are working at a large organization, you're not going to be able to. And that's where building for yourself I think really matters.
So we've seen this adoption curve in AI where it first started with AI tools. That's call and response chat GBT give me a recipe for lasagna right then we start to see some companies I guess you know progress to AI 2.0 which is automation. Once I get a lead come in find out who it is send them a calendar invite to schedule something. And here in 2025 and on we are entering the world of agents. What is AI agents? AI agents instead of going from step one to two to three could go from step one to 11 back to six. Essentially, an agent is your co-pilot to use a variety of different tools to get any tasks done. But before you actually go down that path, I would again reinforce start with AI automation and that will help you understand actually what's possible. It we all really do need to be builders in this new world.
So to wrap up here, what are the trai what are the true applications I think most important applications? Well, first and foremost, it is about agents. 75% of leaders in finance expect AI agents to be routine in operations within 3 years. I'm not sure how many of those finance experts even know what an AI agent is. So I don't know how much you can actually put weight on it, but I want you to all understand what an AI agent is. And again, the only way you could truly understand what's possible is by building.
Two, FPNA teams struggle to predict beyond six months. I see it as an entrepreneur running a 300 person business where the world is changing fast. I struggle to have my finance team give me the right forecasts which allow me to make the right decisions. There's too much data coming from too many different places and there's version control. AI could really help understanding what the future looks like by modeling the past. And that's what it did for my health bot. It said people who did these things, they ate cheeseburgers each day, they probably died, right? So maybe you shouldn't do that, Matt. And that's basically what a what a forecasting tool could look at is it'll look at companies like the ones that you're overseeing and and what are the benchmarks and what are the decisions behind that benchmarks that will unlock growth because ultimately that's what it's all about, right? It's about unlocking growth and forecasting gives you the confidence you need to make decisions that unlock growth. So I think that's a huge area with AI and talking to data I think is probably the most important unlock. When I talk to my health data, when I talk to my personal finance data, when my mom talks to the buy for her to understand her Netflix password, she's getting information back and not having to dig it up. and your ability to take the financial data, the companies you work on and put them in a place where that not just the leadership but the entire organization could know that A and B equals F and F equals growth will help you level up to being more strategic. You can take and you can decide what data you share with who. But your ability to have this incredible information, the essentially the math equation behind your business accessible to your organization so they can talk to the data in ways that they understand and really understand what's happening with the business. I think is the ultimate unlock for people who are in the finance function. And that's why I think the best businesses are math equations. I was taught that very early on. But if there's a math equation of your business, for every thousand people we talk to, 50 people get a demo. for every 50 people to get a demo, three get a proposal, and one closes, how much does it cost to talk to a person and you basically work it back? Well, then you can pretty much know how much you spend on sales and marketing. And the more you can turn your business into a math equation, the more strategic you can be at driving change. And now AI really does tr truly allow you to unlock that power.
Um, I want to thank the Black Line team for having me today. They've been tremendous to work with. Um, this is my information. If you take a picture of this QR code, uh you can enter your email address and it will give you access to a ton of different tools that I talked about today. And I'm here as a resource and I really hope all of you do the work needed to future proof themselves. It's a scary world, but it's also a really powerful, exciting world. And I'm wishing all you guys the best here in Vegas. Thank you so much.