Okay, Dave. So you and I are gonna talk about AI today, and I just want to, you know, throw this out there. I think, you know, what you and I were talking about earlier is we're kind of beyond this initial wave. I think there's a lot of mixed feelings right now about AI. I think initially, it was really exciting. Now there's some people out there wondering, was this a fad? They're comparing it to things like blockchain.
I would even say that, historically, there's been a lot of fear, but right now, I would say that maybe there's more fatigue, if anything. So you have different parties in different camps. Right? You have some people who are beyond excited about this, continue to see, you know, so much, you know, progress, potential, etcetera, and you have others who scared, tired, whatever the case may be, maybe even starting to be a little bit bit of anti AI. And I think so today, if, you know, if we kinda break down our goal is is let's kinda share our thoughts about it, but, you know, what does the second wave look like? Does that make sense?
Yeah. Definitely. I I think it's a good time to talk about it. We've had GenAI in the marketplace for a while now, and I agree. I remember when ChatGPT first came out, people said this is every bit as impactful as when they created the Internet.
And other people said, but what about blockchain? Right? So it's always been very divisive. And, one thing that I've noticed more recently is that there is a, you know, a fatigue.
I think that a lot of people went and bought solutions that were AI driven, and they didn't really get the outcomes that they were hoping for. And that's leading to some skepticism around this entire concept. You have other people who are so convinced that it works that they're fearful that they might lose their jobs. And I think that, you know, our job as leaders in a technology company, as always, is to provide clarity on what changes in technology mean for our ecosystems and what they don't mean also.
So I think now that, you know, if you follow the fundraising, we're still very much in a hype cycle. But if you follow the actual customer experience, I think you could say that first wave has crashed, and now is a good time to look at AI a little more analytically and what it actually means for you and for your business. So, excited to dig into it.
Yeah. Two things came to mind came to mind to me. First, you know, the the comparison to blockchain, I've certainly seen that. One key difference for me is when I heard about blockchain, it almost felt theoretical.
It almost felt like being back in college where there was this idea, and it was like, oh, okay. That makes sense. But I didn't see people with apps for blockchain on their phone. It wasn't it wasn't like you and I signed up for a blockchain account and we're incorporating that.
Does that make sense? So like that feels like a little bit of discount where the difference where right now it, you know, feels like people are using AI all the time, and that's one key differentiator.
I think about whether a technology is truly disruptive. One of the in order to actually disrupt, it needs to be adopted. If people are not adopting the technology, then it's not disruptive. If it if it's disruptive to a handful of super smart grad students, that's not the same thing as being disruptive to an entire country of people.
And I'm with you. I think when I look at what ChattCPT has done and therefore what all generative AI has done, I usually use my mom as the litmus test. I say, like, is my mom able to interact with this? And I've seen her interact with ChatGPT.
My mom is actually in the camp that this is the beginning of the end of human civilization. But regardless, she was able to submit a chat to ChatGPT and get an answer. So it's like something that has the ability to influence consumer behavior, I feel like is a must for the ability to disrupt business behavior. And I think especially because the way businesses work nowadays, what we want is a consumer grade tool experience.
I know spend management definitely operates that way. So you need something that is consumer grade that you do see everyone using, that you do see showing up on Instagram, to actually have that disruptive impact. And I think that we are seeing that with generative AI in a way that we haven't seen with any other technology for quite some time.
Yeah. The other thing that came to mind for me, I would completely agree with you. If there's people like your mom who are all of a sudden using, you know, AI, that's that's pretty cool and certainly litmus test for this idea of adoption. The other thing, though, that came to mind for me is people are sometimes underwhelmed with the response or what they get out of AI.
But what I think is missing is they're comparing AI to what it is today, and they're ignoring how fast it is changing and how fast it's growing and how fast it's it's increasing its ability to add true value. Right? So even even six months or a year ago, you would get on to, you know, one of the tools and you would put something in there, and it it was cool, but there were a lot of mistakes. And now those mistakes are decreasing, and the value is exponentially increasing.
And so, yeah, is is it solving all of the use cases today? No. But is it quickly and rapidly accelerating? The answer is yes.
And and and the pace the change of pace is incredible.
A hundred percent. And like I said earlier, if you follow the funding, this is where growth and investment is happening, and that's usually where real innovation is going to happen. Right? If you play out where that money is going, what r and d it's it's funding and developing.
And I think that that's actually made me think of this really good, HBR article that just came out, that I was actually reading this morning. Let me see if I can find the title. We could send it out with the notes. But it's something to the effect of, like, AI is just a prediction machine.
And if you think about the history of computers, that's all computers have ever been, right? It talks about like the Mario Brothers video game. You know, what that game is doing is you're telling the AI to react and you're telling, sorry, the technology to react in certain ways when certain conditions are met. And that's all computers have ever done.
The only difference is the the complexity of the data that's driving that decisioning and driving that prediction is at a level that it's never been before. And it's gonna continue to improve, but it isn't truly providing human level reasoning. And that, I think, is where people may have the wrong expectation and also why I think the concern about jobs is a little bit misguided. So like the AI, for instance, today, can tell you that it thinks fraudulent activity is having on your bank account.
Right? And I've had that happen to me fifty times or so. None of those was fraudulent. What it's doing is it's trained to look for patterns and recognize anomalies in those patterns and say, Hey, here's an anomaly that we recognize.
But the human still has to be the one to say, Well, I know what I buy and what I don't buy. Therefore, I can, you know, add a level of sophistication to this interaction if the AI can't. I think that's really, really important because that's where I think people have an unmet expectation. You can't ask the AI, you know, who the love of your life is going to be because only you can reason that when the situation arises.
So, when I think about spend management, right, I think that there is a very interesting transformation happening that mimics that exactly because the data is so complex.
The spend data lives in many different systems, maybe doesn't live in any system, and maybe is able to get pulled into a spreadsheet and to generate a report. And I think that the efficacy and the and the quality of a finance or procurement team's output was in some ways guided by their ability to generate the report. Right, to synthesize all the information and say, you know, here's a view into the visibility that everyone can now see this information that's otherwise very hard to get. And that's frightening in a world for AI because the AI can do that in milliseconds.
Right? Like, you can prompt just about any report that you want. And if the data is there, you can get it. However, the AI, the ability for AI to reason exactly what needs to happen next and the ability for AI to go make that thing happen, those are the areas that AI is not capable of delivering, right?
So I think when I think about strategic finance, you know, getting a good understanding of where you have opportunities to optimize cost and maybe even what those opportunities might be is still the same thing as getting that fraud report. You still need to go in there and assess as a human whether those are real opportunities for you or not. So that's the role of the human in an AI led world of spend management is to still provide that layer of reason and execution. And I think it's actually very much a good and necessary thing to not be the one that has to manually make the report every time, even if that is how we used to show value in our organizations, right?
So that to me is like a good example of how AI is changing things, but it's not solving every problem. Humans are still the best judges when it comes to how to solve the problem, but they need tools that make it very easy for them to get to the point where they can make that decision. And that's, I think, what AI is doing.
Yeah. I I would agree with you. I think back to, you know, I'm I'm I'm trying to be active with, you know, other leaders in the community, both finance, procurement, etcetera. And I think about, like, what are their pain points today?
And procurement for a long time has really struggled because, you know, they're they're reactive. They're brought in late, to to, you know, contracts or renewals or or purchases, etcetera. You know, the the reputation isn't the best. They're often a misunderstood function.
Oftentimes, they lack appropriate resources.
And then finally, you know, the you know, because they're often covering such a broad breadth of spend, you know, they're not specialized. So I would say that those are the biggest problems procurement is up against right now. And I honestly don't see AI replacing the humans and doing something better. I see AI supercharging those, you know, humans, the the professionals.
That way it solves those problems. Right? So the technology makes it so they're brought in the earlier. So, you know, the the reputation improves because they are, you know, driving more value.
You know, they're able to sort through the data quicker. They, you know, they can, you know, in terms of, like, lacking appropriate resources, they can do a lot more with a small team or a, you know, a team of one in a lot of cases. And then it and it you know, the lack of specialization is a big one because it can provide you with all these data and insights to where you don't need all the experience for every single category. You need that procurement experience.
You need to you know, how to, like, the process works. But you can really manage a lot of different categories because you're armed with this data and insights from, you know, AI tools. And so that's where I see it, and it's not solving all of the problems right now, but it's starting to have an impact on that. And I think that's incredibly exciting for procurement people like me that, you know, have operate in this world where there there historically have been a lot of challenging, and it's a very difficult field.
And this is a way to to make the impact better, you know, to to make it so that way the reception of procurement's better. I I see a a beautiful future between AI and procurement.
Yeah. I I couldn't agree more. Obviously, being the CEO of a company that is building AI for procurement. But I think that what's really interesting about procurement and spend management is that those problems that you described, those are not traditional software problems. Right?
Traditional software problems are, you know, ease of use, connectivity, integration.
And procurement's problems, I would argue, are too complex to be solved just by cloud software. Right? So if you think about how a sales team works, the sales team has a tremendous amount of specialized data, just like you said. Right?
They're hyper specialized in selling their product. And the objections that are gonna come up when they try to sell that product and the personas that they're gonna sell to, they get to get tremendous pattern recognition because they're running exactly the same motion over and over again on the same exact product every single day. And, you know, what Salesforce and then the surrounding ecosystem did was put cloud software around that process to manage it and to further hone that blade and that ability. But the underlying data, you could very easily, as one company, and only as that one company, get all the data under the sun on how to sell your product.
But as you well know, practitioners and procurement have no such luxury. Right? You're only buying that one product maybe once a year, if that. And there's no way to build that pattern recognition.
And what that leads to is, you know, a bad software only experience. And when I think about, the early days of AI, ChatGPT used to hallucinate. The reason that ChatGPT and still does, but used to hallucinate more. Right? And the reason that it's it would hallucinate is it had incomplete dataset.
And the incomplete dataset is the situation that every practitioner procurement finds themselves in, where I now need to go, you know, source this thing that I've never heard of and the data that I need doesn't seem to exist. So that's an area where I think AI and procurement really do kind of come together beautifully because that practitioner will always be the best at what they do, which is running the strategic function in the company, but they have they're running it with a very incomplete dataset. And cloud software can't solve for that incomplete dataset. It can't predict better outcomes because it's only gonna get one data point on Salesforce every two years.
Right? So that's why we built, you know, Tropic on a community of data where we said, you know, what if you did have an easy to use tool, but it was sitting on a community of data and insights so that you don't have that limitation of data problem? And you can then use the AI on top of that to create the outcomes that you're looking for. So, lack of resources, lack of data, some of the other things that you said, these are not software problems.
This is a new breed of problem that I think are really AI problems. Right?
And only AI can actually solve them. So, you know, if you're sitting in a spend management position, I think this is a great time to be in your role because AI is giving you the tools to be strategic and to make an impact in your company in a way that you couldn't before. So, you know, very much the opposite of putting roles at risk.
Yeah. One example that comes to mind for me, you know, I've I've been doing procurement within the tech world for quite quite a few years now, And it feels like sourcing was always viewed as something very strategic, but it was often underutilized. Okay? And if I think about the reasons why procurement professionals weren't doing this idea of sourcing, what that means is, like, going out and truly exploring other options, trying to develop not only optionality at the point of purchase, but aligning yourself with the best supplier out there.
K. So so why wasn't it being done? Number one, frankly, it takes a lot of time. K?
I I think you have to get buy in from the business. There needs to be, like, you know, some compelling reason, and the data is sometimes missing.
I think number three, there there weren't a lot of good tools that were not aggregated except by humans. And so a lot of times, you you think about how quickly the software world is and has and continues to evolve. Well, a lot of times from a sourcing perspective, you're only looking at the known data points, the widely known data points, and you're ignoring maybe up and coming companies and and things of that nature.
And and so, you know, AI is incredible because I can go into a tool, and I can just say, hey. Here are here's here's my decision criteria. Here's the problem I'm trying to solve.
What companies should I consider looking at? And if you have a good data set, right, if you have a data set like what Tropic has, where you're looking across, you know, hundreds of different companies, you're looking at data points, like, in terms of sentiment and pricing and and things of that nature, all of a sudden you take what has historically been a pretty heavy lift that they've tried to solve through very arduous processes like the historical RFP process, which takes a ton of time and resources. And you can now use AI to come up with potentially one or two alternatives and narrow it down very, very quickly.
And yeah, you'll still need to go through some demos and there's still that human involvement, but it takes a big portion of that lift and automates it very quickly, but in a way that's more, profound, more comprehensive that when it's ever possible. And so I think sourcing is going to be one of the ways that is impacted tremendously from a procurement perspective in the world of AI. But it's leveraging, like you said, all of that data, all of that those insights that make it all possible. So that way there is not the hallucination, but it's very comprehensive.
Yeah. And that's a strategy. A hundred percent. The implications for for sourcing, I think, are massive, especially as the space continues to grow and as companies continue to grow more data points. There's compounding value there in terms of access to information, right, which is one of the problems that you you raised. You called it specialization.
I think that's access to information. Right? Access to quality information that can help you do your job well and steer the company correctly. And I think that, the other thing it's interesting because, you know, I can speak from my own personal experience.
Like, when there was a set of things, I think about a marketing agency I worked at a really long time ago. There's a set of things that I did exceedingly manually to publish a blog post at this agency, you know, whatever it was fifteen years ago that only I knew how to do because I was generating all the content. And when I first learned that there was a better way to do that, my reaction was fear because my my sentiment was like, but this is how I show value. I know how to do this thing that nobody else knows how to do, which is like all the meta tagging and, you know, all the SEO stuff.
I was doing that manually in the back end. And then, I remember when I fully embraced this new platform that we moved to for content generation, I was able to write, I think, twelve different pieces of content in one day because I wasn't spending any of my time setting up the post in the black end in the back end and adding all the tags. So, like, maybe this is the wrong analogy, but I remember that feeling of, like, visceral fear was my first response when I learned that we were gonna be doing it more efficiently because I was like, this is how I add value though. Right?
And I wanted to protect that. And I think about how companies get the information that they need today to make decisions. So, like, if you think about risk. Right?
You know, if you have a very simple set of policies that, you know, we're not going to sign anything with an auto renew and, you know, the basic five point, you know, policy that you might have, for suppliers that you wanna bring on. The way that most companies are gathering those insights today to determine whether or not there's risk because they're sitting there and reading a document. Right? And, you know, when I talk to the people that are doing this to their companies, procurement operation specialists, sometimes finance people, sometimes legal, the sheer volume, like, of hours spent just reading very similar documents over and over and over and over again so you can check a box on whether or not something is on policy is, like, staggering.
But it's like, that was me. Right? And and I can understand that, you know, if a technology comes along that can read it for you and produce, you know, automatically what the risks in the document are, the initial response to that might be fear, because that was mine. But again, I kinda bring it back to the credit card analogy where it's like seeing that is not enough.
We know from, you know, you especially sitting in a strategic function in spend management that sometimes you need to break the rules. Right? You need to make good judgment calls on the decisions that your organization is making, and that human element of cognition is required to say, no. This is the right thing for us, and here are the reasons why.
Or no, this is the wrong thing for us and here are the reasons why. And I haven't seen any technology that can do that for the same, you know, data reasons we were mentioning before. But if, you know, if you told me today, knowing what I now know, you can automate every single one of those manual things that you used to do, I'd be like, sign me up for that. You know?
And that, I think, is what AI is actually doing. It's you know, it may be the most strategic, crazy, you know, innovative thing that we've ever seen. But what it's actually doing that seems to be adding the most value is reading documents for people, is extracting data for people and putting it into reports. Right?
That's stuff that we hate doing, but you have to do it to run an effective spend management practice. And, you know, I think we we started this discussion around, what's really going on under the hood with AI? How does AI actually create value? From what I've seen, it's those things that really are creating the value.
It's not the crazy pie in the sky, you know, ten year visionary stuff that you might hear about. It's like I had to do this really painful thing, and now I don't.
I think that's gonna resonate so well with procuring people because there are a lot of very tactical, boring, monotonous.
And and and, of course, you know, one of two things either happen. Either you don't do it and you kind of absorb the risk because, you know, somewhat, you know, you're just focusing on other things or you're doing it, but it's kind of a drain on your on on, you know, your time and your company and your your fulfillment and what you're getting out of it. So I'm I'm with you.
Hundred percent.
Yeah. Yeah. But to your point, you also brought up a good point that it it's only as good as the data it's built on. Right?
And you're right. We're we're seeing fewer hallucinations, but then again, I think that there still is some lack of trust. I think that's okay. But the better data we can put into it, the more accurate data it has underneath it, the better the better insights, you know, the the bit more we're gonna be able to rely on it.
Right? So, you know, I think that's that's really important that when you're using AI, you look at, hey. What data does it have access to?
And I think that's really, really important. So let's, let's migrate just a little bit. I don't wanna go completely pie in the sky, but I would love to kinda pick your brain on, you know, what you're most genuinely excited about, maybe some predictions. And let's not keep it, like, twenty twenty six and beyond. Let's maybe keep it for next year, the next twelve months. Like, where do you see the immediate future of AI, especially as as you talk to, you know, TalkWith's customers and and finance leaders and and spend management professionals?
Yeah. So there's you know, I would probably categorize this in two buckets. Like, one is things that I'll never do the old way again. Like, that's an area where I you know, once once you try something and have the experience of, you know, benefiting from that thing, just realizing you're never gonna do it the old way again.
Right? So, like, this is silly, but, like, building decks, I'm never gonna build a deck without AI ever again. Right? Like, I'm never gonna, you know, I'm never gonna do anything without logging in with face ID if I can ever again.
Right? It's like once you have that initial moment and realize how much easier something can be knowing that I'm never gonna go do that ever again. So for me, that's definitely you know, with my hands on experience and application, that's definitely being able to ask the platform things and get an immediate answer about what I should do because I don't like to think at all. So, you know, if I'm approving something and I can ask a question and get an answer and move to the next step, that's a big big win for me.
So I think, you know, as I think about next year, like, what are some things that we're gonna start to do with AI that are gonna be those things that activate moments of I'm never gonna do it the other way again?
I think that what we're going to see is a total transformation in how software platforms do reporting. Right? And, might seem silly, but I think there's a huge room to transform reports. Like, I never have exactly the report that I want in any of my tools, and we're using, like, the right tools to do that theoretically.
But I I always want something a little bit different after I see the initial report that's in the reporting dashboard of whatever that tool is. And we're still kind of living in this, like, legacy framework of you go to an application, that application has a one of the tabs in the menu is reports, and there might be some customization, but, like, here's the templates for the reports that we have. In an AI led world, you should be able to have any report that you want in milliseconds with any data point that you have. Right?
Like, I should be able to just go generate exactly the insight that I want, have the AI pull it for me, and then make decisions with it. And I think that every platform worth its salt is going to evolve that way. So I I hope I hope at least that we see the death of, like, the reports tab and the reporting template. And instead, what we see is, like, a prompt where I can prompt any report that I want, get a visual for it, and then make decisions.
So I think the I think I think you're right, and I think the user interface is gonna change a little bit too.
It's gonna be much more simplified. Right? Because you're not gonna need all these different tabs with, you know, hey. Click here.
Is it gonna be a I almost think like a matrix style, you know, like, line where you just start typing. Of course, it'll help you develop those prompts, but then it just takes you where you need. And it's it's custom too. Right?
It's like, hey. This is the question you asked. I I did it real time. It was so interesting.
I was I was talking, I was I was on a meeting, and we were, you know, working on our our approval process and and our thresholds and things of that nature. And I just went on to, you know, the AI tab, and I asked a very and I almost kinda did tongue in cheek because I was trying to be, like I I don't wanna be, like, super, you know, like, formal about how I ask it. Like, I'm I'm very much a, you know, very informal guy in a lattes, and and I used my tone in there. And and it just interpreted it so cleanly and so clearly and just popped it out.
And it just it kinda blew my mind, to be honest with you. But I I think that that's going to change the UI as well where these complicated platforms and solutions are gonna look very simplistic because everything that's happening is happening behind the scenes, and they're just gonna produce what you're looking for. You're not gonna have to dig anymore. And that's Exactly.
That's awesome.
And we don't need behind the scenes anymore. Right? And I think that, like, when I think about our product strategy, which I think does embody what I hope to see from AI, in the next few years, it's like we should never have to learn how to use platforms ever again. Like, I shouldn't have to adopt a a new way of working for a platform.
And, like, I was talking to somebody about the word workflows the other day because people at our company hate the word workflows. It's like every tool's got a workflow. We don't like workflows. And I said, well, look.
Whether you have a platform or not, you have a workflow. Right? You have a way that you do your job. Right?
So, like, my workflow is I I look at Slack first, and then I check my to do list, and then I pull up, you know, the deck or whatever it is that I'm working on. That's my workflow. And when you think about the workflows of almost every single person that I've met in finance or procurement, it's a brutal workflow. It's got all these different systems that I need to go into and pull things out of and synthesize.
And to do that, I have to learn how to work with all those systems. And it's like when I join a new team, I'm in a new role, one of the first things that people teach you is, you know, I went through this at Microsoft. Here's, like, the process for how you have to do all the things that you need to do for your job. You're gonna go to Power BI and you're gonna run this report on how you're forecasting your territory and take a snippet of it and then drop it into a PowerPoint deck and, you know, times a thousand.
We all have seen this kinda thing before. And it's all because we are you know, technology is forcing us to adapt to it. Right? It's like this is how you have to adapt to the technology that's here if you wanna get your job done.
AI will be able to do is it will be able to create technologies that adapt to you and how you work. Right? Like, what I love about ChatGPT is that it remembers me. Right?
When I'm going in to ask it for something, it it can connect dots between something I'm asking for today and something I'm asked it I asked it for a year ago. Right? And it knows the things that I care about and I focus on. I don't know why every technology platform that I use isn't like that too.
It's like if you know I'm gonna come in here and do exactly the same thing that I do every single day when I come to this platform, you should be making it easier for me to do that thing. You should be recognizing this is what that person does when they come to our platform. And then if you can nail that, you know, here's CEO Dave, and he wants to see, you know, upcoming spend for the next six months in budget versus actual, so I'm gonna give it to him as soon as he comes to the platform. If that's what Dave likes, then the next thing that AI can do is start to realize who else looks like Dave at other companies.
Right?
I noticed you have a CEO in your title. Here's what other CEOs usually wanna see when they come to Tropic. Right? Like, that's the experience that that we should be delivering because the technology is there.
ChatGPT can do it already. So why are we not doing that with business applications? So, like, I don't ever wanna learn a new tool ever again. I want all my tools to learn me.
Right? And that I think is a big difference with the AI can deliver on.
That's that's cool. What I tell people right now is, hey, at at a minimum, like, experiment. Just try it out. You know, even if you're skeptical, go try it out.
But here's maybe my bold prediction for twenty twenty five is you'll still need to do that to some degree, but it's less and less and less because AI will be so built into a lot of tools that it's naturally occurring, that it's it's it's there and and that, you know, it's already making recommendations. It's it's changing how you, work, and it's prompting it. So right now, you have to go to it and say, hey. Here's a document.
Can you review this for me? And I think people can and should do that. But I think in a very short period of time, you're gonna be reviewing a document that's gonna say, hey. Here's some risk that maybe you'd I I you know, that maybe you should be on the look for, and and, it's it's gonna be way more proactive.
That's my bold, kind of, you know you know, prediction. But I I agree with you that, you know, technology should certainly work with you. It's gonna help you do more with less, and, and and the future is is bright. That's not to say that, you know, it it it I think those emotions of of fear and and fatigue are real.
But, you know, I I almost feel like it's inevitable, honestly.
Of course. Yeah. It's I mean, you can't put the the genie back in the bottle at this point. And I think that to lay concerns as the CEO of a company that does this, like, the things that are, as I said before, in my opinion, by far the most impactful that AI can do are the things that you never wanna be doing yourself.
And, you know, I would never there's a million things that I would never do without the Internet as the tool that I do it with today. Right? And I've been starting to feel the same way about AI. And I think as the technology evolves, we're gonna really see that showing up inside businesses.
Awesome. Well, this is oh, you know, I I, I'm I'm always excited to talk about it. I admit, in the very beginning, I was a little bit skeptical. I I consider myself an AI convert at this point. I don't believe and told trust in it in entirely and implicitly, but I do think that, that trust is, you know, building and growing, and, it's it's pretty exciting.
Hundred percent. Good chatting. Always fun.
I hope that that's really you and not a, you know, an AI, a I'm actually on vacation right here now talking to, an AI, Awesome.
Good rapping with your shields.
