How Our HR Mistakes Cost Us 30% on Our Exit

Matthew Schmidt: To do effective integrations and automations and those sorts of things. You don't have to have an AI behind it. Right. Um, AI is really good at sifting through lots of information and understanding things that look like what it's been trained on. Right. Uh, and now to be, you know, generating something. Like next. Right? It understands the, the next thing it should be saying. So, you know, I think as you look at streamlining your business, you know, AI is really good at under, at looking across all the different sources of data within your organization. I. And helping you understand the, you know, potential relationships between that data or the places where it's, you know, it looks like it may potentially go off course, right? From an automation and integration perspective, the, the AI piece is sort of secondary, right? But it may point you to where you can use automation and integration to be able to, uh. To have an impact on the business.

Sean Weisbrot: Welcome back to another episode of The Wheel of World Podcast. I'm here today with Matthew Schmidt, the founder and CEO of People Logic, which makes organizational health simple enough to be actionable. They used more insights, fewer surveys, and one score to track, to measure and optimize your team's health. In this episode, we'll be talking about HR and knowledge graphs and things about automation and integration and ai. Uh, there are a few things I'd like to talk about before we get started. First is I've been checking into the statistics and the analytics that YouTube provides. And unfortunately, 99.1% of the people that watch these videos are not subscribed. So if you're not subscribed right now, I would love for you to subscribe. It really helps the algorithm and gets me closer to that 1000 subscriber number. We're currently at 2 57. We've been growing a lot faster because people have been subscribing finally, which is amazing. I'd love to get to 90%, so. Thank you very much for that and for following with us. The next thing is, uh, this episode I was supported by Chachi Bt. I did something really interesting. I went to Chachi Bt and I said, Hey, create some questions that I could ask Matthew Schmidt, and it gave me some questions. I didn't think they were very good, so I changed the prompt and I said, Hey, those questions suck. Why don't you give me questions that are unique? It spit out some really interesting questions. I may not ask all of them. I'm gonna ask some of them. So I would love for you to tell me in the comments which ones you think were mine and which ones you think were chatt, and maybe I'll prompt Matthew to tell me what he thinks about that as well. And, uh, I have been recently creating other content outside of the podcast that'll still be on YouTube. So look for that. I'll be talking about chat, GPT and AI, and my investing philosophy and some of the me investments I've been making and why I made those investments and things like that. So get ready for a lot of really interesting content. Now, let's get onto the interview. Thank you for taking the time to talk with me, Matthew. I appreciate it. Why don't you tell the one a little bit more about People Logic, and we'll go from there.

Matthew Schmidt: People Logic was created with the goal to help companies be able to grow faster with less risk, uh, and really came about from being able to build a growth company, my last company, which we exited. Um, but seeing that, you know, we made a lot of missteps along the way, but had all this data in the tools we were using to be able to, uh, be able to make better decisions. Right, and to be able to make better decisions about our people, uh, so that we could improve their engagement and their retention and their employee experience. Uh, and so, you know, we, we have been working on that mission to really be able to help HR leaders be able to get a better pulse on their organization without the need for surveys. Um, and to be able to have it be right in their flow of work without, uh, having to continually bother the, all the people on their teams to, to try to get feedback and, and insights. What were some of those missteps you made? Like most companies, we, we burned out a lot of folks, um, predominantly because we weren't paying attention to where there were process inefficiencies, right. Um, or. Where we weren't really understanding the, the true social network behind the organization. Uh, and so, you know, what that comes to, where that comes to bear is really where, you know, you have your lead engineer that's actually spending all this time helping your customer success team or your customer success teams interfacing with your sales team to improve sales and not doing the work that they should be doing, or, you know, doing work. In a manual way that's, you know, not creating as much efficiency, uh, as they could be. And so, you know, they, that led to higher burnout, higher attrition, um, you know, just lower satisfaction, more, uh, frustration. Uh, and you know, with people logic, we're able to see where those bottlenecks are and where people are, you know, really, you know, spread too thin across the organization.

Sean Weisbrot: You learned from the experience of. The mistakes you made from that last business and that led you directly to making this new company.

Matthew Schmidt: Yeah, exactly. Exactly. So, you know, I would say that the missteps that we made around our people, whether that was around their, you know, their engagement or because we lost great people that we shouldn't have lost. I, I would say those missteps probably increased our time to exit by at least two years, but possibly more. Right. And so, you know, as I was transitioning out there, I, I felt that that was a problem worth solving. That if I could help companies that were growing quickly, make fewer mistakes, or to be able to get those early warning signs before they stepped on those landmines, uh, that that would be a valuable business to build.

Sean Weisbrot: I completely agree that it's valuable to help other businesses to make better decisions, and that's one of the missions of my company we live to build. Uh, so I'm glad that we're aligned on that. You said that it took a probably two years longer to exit. Do you think there was a negative impact on the end valuation? And if so, even if you don't put a specific number on it, do you have an idea of maybe the differential? We had

Matthew Schmidt: multiple product lines, you know, two very distinct things that we sold and, um. You know, if we had solved some of these problems, we, we could have built up the second invested more in the second product line to actually, uh, drive its revenue higher, which, you know, probably could have given us, you know. Probably not 50% more on our exit, but I would imagine 30%.

Sean Weisbrot: And typically when you're exiting, that's a huge amount because assuming you're one of two founders, that that could be an extra few million dollars potentially in your pocket.

Matthew Schmidt: Exactly. Right. And for our team, uh, who were also all owners, right? So there was, you know, that was a, a challenge. Worth, worth biting off.

Sean Weisbrot: I have to make a disclaimer here now. Uh. A few days ago, I completed an investment in a brand new company that will be focusing on automation and integration and business intelligence services. So they're not a competitor to your company, but their philosophy is similar, where the goal is to streamline operations in an intelligent and efficient way. With that in mind. One of the types of businesses that I'm curious about working with and and promoting the service to for the agency are companies that are thinking about selling themselves. The reason being is if someone's gonna buy your business, they would like to inherit something that's hopefully not a mess. And generally, if you've been running a company for 5, 10, 15, 20 years. It's probably a mess inside because you probably got comfortable with the way things are working and therefore you're probably not gonna get the best value for your money when you're trying to exit because there's a lot of inefficiencies. And one of the reasons why people will buy your business is because they can spot those inefficiencies and wanna improve upon them. Think about it now like this. If you. Work on streamlining your processes, automating your workflows, and integrating your different departments. When you go to sell, there's less than efficiencies, which means they can potentially have a much faster, uh, growth period where they didn't have to waste six months to a year figuring out all the problems and fixing them. They know that everything is kind of working and then they can scale with it. I think HR is a huge, huge point that a lot of company owners really ignore when it comes to, you know, the operations and the management of, uh, how the company works. So, you know, going back to your product, I think people logic is extremely valuable. In that regard. And again, another disclaimer, this is not paid content. They're not paying me to say this from the bottom of my heart, with psychology in my mind and having experience as an operator and an HR manager, HR is severely lacking across most companies. They generally just don't know how to take care of their people.

Matthew Schmidt: Yeah, no, that's a, that's a great point. John and I, I think like what we see routinely is. HR leaders that are just so completely overwhelmed that they cannot even begin to think about how to, you know, even open their eyes to the fact that the entire industry's changing. Right? 99% of them can't even get to that, you know, to where they have a moment to think about how they might take that evolutionary leap, and so. We see the HR teams being responsible for everything from, you know, internal comms, uh, in the business to running the all hands to, you know, setting up the offsite to benefits, to, you know, laying people off to, you know, and then down in there is okay. Yeah. We also have to like, make sure we take care of our people. Right. And they certainly say they care about retaining people. But it's a, you know, it's a never ending challenge for, for those teams to be able to, you know, just continually keep up.

Sean Weisbrot: How do you see AI revolutionizing the recruitment and talent acquisition process within large organizations, and what specific challenges can it address?

Matthew Schmidt: Yeah, so I mean, talent acquisition's an interesting one, and certainly there's been a ton of money even before the, the most recent advances in generative ai. Um, you know, I think we're, right now we're in a really interesting place because it's, you know, we have a ton of tools for talent acquisition folks to be able to use on one side. And then on the other side, we have these amazing new tools that can create a persona of someone in, uh, in a matter of minutes, right? You want a new headshot, you upload some. Uh, a couple of selfies. You, uh, you need to create a person. You want to put a blog and let's create some content making you look like a, a thought leader or, you know, you want a, a fancy new resume. Like, okay. All those things can be done in like 10 minutes or less if you know what you're doing, right? So that has to be blended in with all these new AI tools that are looking at. The, you know, all these resumes that are coming in Right. And trying to, uh, help to, you know, have the talent acquisition folks be able to manage an ever increasing number of applications. Right. Uh, you know, I think for most organizations, the biggest challenge, particularly larger organizations, the biggest challenge will just be making sure that the AI that they're leveraging. Really aligns with their, uh, their various initiatives, right? Because it's not simply filling the position, it's filling the position with, uh, you know, meeting certain diversity metrics, for example, right? And making sure that the, you know, you're not excluding great people, uh, because there still is a, you know, a human side to, to the hiring piece. Um, and I think, you know, yes, it's gonna allow people to do more with less. But at the same time, you know, I think, you know, going all in and letting AI decide who you're hiring is probably not gonna be the right move either.

Sean Weisbrot: I completely agree. In my last company where I had about 16 employees, we never used AI for any part of the process. We, we created a funnel by hand using LinkedIn and Calendly, well, LinkedIn and Clickup, and then. Put into, put it, put Calendly in email and all that. Um, I talked about the hiring process we used before. I'll mention it real fast and, and I'm curious to see what you think of it actually, since you have a lot of experience in hr. Basically what we did was we would generally post an ad that we wrote by hand. 'cause chat to PD didn't exist yet. We, we, we created the job description. We created all of the, you know, the, the understanding of where that person would fit into the company. Um, and then we would create an ad on. LinkedIn and it was just a job post. We didn't actually pay for the ad because why give them money? And then we would set the region that we were looking for. So if we were looking for a developer, maybe we were looking in Southeast Asia. If we were looking for a marketing person, maybe we were looking in Europe or whatever, and the, you know, the, the full job description would be there and at the very bottom it would go. Do not click on apply through the LinkedIn button, go to this link. Fill out the form there and it was a, a clickup based form. I can't tell you how many damn people clicked on the damn apply button. I can't tell you how many people added me or mark my COO as a friend on LinkedIn. I can't tell you how many people sent us emails directly through our company emails 'cause they've scraped the data off the internet. Guess what? They all were disqualified immediately because our goal was we wanna hire people that follow instructions. Sure, we want them to be creative, but we want them to follow instructions first. 'cause if you can't follow instructions, then your creativity doesn't matter, right? Because there's a hierarchy. At some point, you're gonna have to listen to what someone else says. Fair. Then if they, if they filled out the form on Clickup and they were a fit, we would send them an email. Now, let's say we were hiring for a marketing manager. I can't tell you how many people we got who were cooks or who were developers or who were lawyers or who were college students. So many people were not even qualified for the job. They were automatically disqualified. Of the people who filled out the form and who were qualified, we emailed them and we simply said, thank you for your application. We're happy to do a call with you. Pick a time on this calendar. Here's the link. Guess what? So many people said, I'm available at 3:00 PM on Wednesday. Does that work for you? Disqualified? Why they didn't listen? So if they went and they picked a time on the Calendly, guess what? We'd have the call with them and we would ask them a series of eight questions to see, you know, do they actually fit for us? Like, you know, how, how, what's your current situation? Why are you leaving your job if you have one? Uh, when would you be able to start? What is your asking salary about? Just a kind of a basic thing. If they were a fit, we would then. Send them another email introducing them to the next person. And that person would do a hiring test with them, a, a hard skills test. And if they passed the skills test, they would then go and have a final call with, you know, the, the COO and, and myself, whatever, uh, we might have 400 applicants and only like three or four were qualified and actually paid attention and followed instructions by the end of it. It was annoying, but we got really good people because. They were the only ones that survived. Do you think that's how a lot of companies do it, or do you think that's maybe not a smart way to do it anymore? Are there tools that we can use to streamline this process without dehumanizing the process? Because I, I, I'm, uh, I'm afraid of using AI in a hiring process personally, a

Matthew Schmidt: startup. Probably shouldn't use AI in the hiring process, right? The, the choice and the impact of the individuals at that early stage is too important, right? You don't want to hand off any of that or overlook a, you know, a hidden gem that you know might be a great fit ultimately. But you know, maybe is, you know, a career changer, right? Or is, you know, changing industries or whatever, but they have good skills, right? Uh, and they can follow the process. Um, you know, so I think, you know, that process probably, you know, it's, I, I think it, yeah, absolutely it works, right? Um, I think, you know, probably. It reduced the volume to a place where you were able to, to make it manageable. Right. Um, because I do know, as soon as you have those, you know, you have the apply button and you run it through LinkedIn, you're gonna wind up with, you know, you could wind up with 400 or 500 applicants and then somebody's gotta go through those. Right. And then you have to decide who you want to to follow through. So, you know, when you're dealing with that kind of volume, does it make sense to, you know. At least have some tool that, you know, breaks it down into different buckets. Like I could see that, like where it's a classifying sort of thing, right? So we, you know, this person has all the skills that we. Uh, that we wanna see and this person, uh, you know, follow the instructions and, and so on. Right? I'm not sure of the current state of the art in the, you know, in that world right now. We tend to play a little bit later in the HR stack than the talent acquisition space. But, um, you know, I think it's all about a question of volume and how many of you're trying to get through and how many on your team, um, are responsible for the hiring process.

Sean Weisbrot: What's an example of. How you've helped a company to kind of transform their HR functions and, and improve the way that they manage that.

Matthew Schmidt: We have a couple of, uh, you know, really key case studies and the, the big thing that we tend to help them do is allow the HR teams to, uh, push down some of the responsibility, some of the management of employee engagement down to the individual managers of teams. To allow them to understand where they have their risk and bring that into their one-on-ones and their team meetings. And whether that's around operational inefficiencies or whether it's around the, the more human side of, you know, people being overloaded or having, uh, you know, those, you know, becoming that influencer across the organization. Um, you know, that allows the HR team to then become a bit more strategic. Uh, instead of being constantly putting out fires. And so, you know, with that in in mind, we, we really stay very focused on, for the most part that, you know, the middle of the employee experience from the after they've been hired and, and helping, you know, we helped to some degree to, you know, help them measure onboarding, uh, all the way through then to, you know, the, you know. Just beyond where they're, you know, before they're getting ready to, to perhaps lead the organization. Ultimately our goal is to make sure that you're keeping great people. Um, but you know, if you've got people that aren't great fits, you need to know that too.

Sean Weisbrot: So you're saying that your goal is to push the information down to the lowest level manager who has boots on the ground or is managing boots on the ground. But doesn't that start from the high level and if so. How do you create a high level strategy with a company in order to enable that? Because it really starts from top down. I think

Matthew Schmidt: the initial place that we start within in an organization is that highest level, uh, person. Uh, whether that's, you know, in HR typically, right? So we, we start by working with the HR teams or the people teams and we. We work with them to create sets of enablement content, right? So whether that's, you know, here's how you message this to the organization, here's how you, uh, you know, work to train the individual managers. Here's our understanding of your goals for the year. Um, you know, here's the, you know, your current engagement levels from your last surveys. Um, and here's your, you know, your current, uh, attrition levels, right? And from there then, you know, we can work to kind of guide them through towards a, you know, a, a path and a strategy that allows them to, um, to achieve the goals that they are responsible for, right? And then we work with them occasionally on. Producing reports that they can use for executive team meetings and all hands and those sorts of things around, uh, some of the findings that, that we see that are useful for, you know, showcasing the fact that HR is now data-driven, uh, and proactive rather than simply being reactive and and focused on compliance.

Sean Weisbrot: How can AI be employed along this process? So that what gets implemented for these people teams and these HR management teams to be able to streamline what they do, automate what they do, integrate what they do, so that. As they scale, it doesn't break, and they can make sure that new teams or new people that come on get onboarded correctly and get trained correctly, et cetera,

Matthew Schmidt: to do effective integrations and automations and those sorts of things. You don't have to have an AI behind it. Right. Um, AI is really good at sifting through lots of information and understanding things that look like what it's been trained on. Right. Uh, and now to be, you know, generating something like next, right? It understands the, the next thing it should be saying. So, you know, I think as you look at streamlining your business, you know, AI is really good at under, at looking across all the different sources of data within your organization and helping you understand the, you know, potential relationships between that data or. You know, it looks like it may potentially go off course, right? From an automation and integration perspective, the, the AI piece is sort of secondary, right? But it may point you to where you can use automation and integration to be able to, uh, to have an impact on the business.

Sean Weisbrot: Would it make sense to use this enablement content to train the AI on how it can be effective? In helping the team or, or is it possible to maybe create an AI chat bot that would disseminate this information or make this information easily available at, at the ask of a question from someone who's being onboarded or even someone who's actually quite good already at the company.

Matthew Schmidt: So I think that's where you're gonna see a lot of the impact of AI in hr, right, is changing, freeing up the HR business partner in particular. From spending their days answering questions. Right? So when you're feeding the AI chatbot, the sum of all your company knowledge, whether that's around your benefits or around your projects that are in flight, or what the current OKRs are or what have you, right? Um, you know, the, they can ask that at the drop of a hack from Slack or Teams or from the web. And those are, um. Things that HR business partners and other people will no longer have to answer. And I think as we move forward and the bots continue to get better, that's going to free people up to do things that are more strategic, that are more focused on driving value for the business and increasing the the value of the company.

Sean Weisbrot: I asked you that question knowing something really interesting, so I was already aware of the answer. I wanted to see what you thought. So I'll follow that up with, I recently read about a study that was done at a company with a few thousand employees, and this company provides kind of like customer support. Like it's, it's like outsource customer support basically. And what they did was. They took some people and trained them on this kind of a bot where you can constantly ask questions. So let's say for example, you're brand new, you get a question from a customer, you don't really know the answer, you can just ask the bot and it tells you the answer like this. So they found that people who had been around in the company for one month were performing a as well as, if not better, than people who weren't trained that way, who had been at the company for six months. That's not to say that the people at six months should be fired, but to say that this kind of a tool can drastically speed up the time for training people and making them proficient at providing at least customer service support, I. But that doesn't mean it can't also be put into it. Couldn't be trained on your technical documentation for your development team to ask questions of the backend. That doesn't mean it couldn't be put into another thing about product, so that your product team could ask questions of the speed, the feature, specifications, documentation. There's no reason why these words. On a webpage or on a document, can't come to life and support the development of the knowledge of your teams so that they can help each other, or so that they can ask questions even between departments so that you can prevent them wasting time trying to reach out to other people. That's not to say that the communication between humans isn't vital for the development of, and the thriving of a company, but. When people are so burned out by meetings, these bots could save tremendous amounts of time wasted on emails and wasted on meetings.

Matthew Schmidt: Those sorts of things, I think make a a lot of sense. Right. A question answering is, is probably the simplest place for those to, to get injected into an organization. And it might be that it's, uh. It might be that it's about a feature spec, or it might be that it, you know, saves engineers time because they're not being broken out of their flow to answer a question. Or it might be that they can speed up writing a new piece of code by having intelligent blueprints. Right. So, you know, I think that there's, you know, as far as efficiencies, we, you know, looking at it in terms of how can we use. The question answering capabilities and the generation of, uh, text, uh, of all crimes to be able to, to speed things up. Right? I, we already use it in the sense of being able to, uh, you know, create the initial presentations, right? Using some of these new, uh, presentation generation tools that are, that are coming out, right. Um. And so, you know, when I don't have to spend time, you know, staring at a white set of slides, um, you know, it starts to, starts to really add up.

Sean Weisbrot: Do you think HR professionals should be worried that AI will replace them, or do you think they should be excited that AI will allow them to do less of the menial tasks that they are currently doing? In order to focus on providing a higher level function for their company,

Matthew Schmidt: human nature is to be fearful that they're gonna take your jobs, right? Uh, that's the, whether it's AI or it's some other person, that's the, the human nature part of it. Um, the people who are growth, growth-minded and are gonna be really, uh, exciting about the opportunity that it provides. For you to do the less menial work and focus on driving value for the, for the organization, right? Uh, simple case in point, right? The, you know, I know folks that, you know, take all their notes and notion and then notion summarizes them and pulls out the items, right? Um, that saves them a huge amount of time and has them be more prepared for meetings and, uh, for having conversations. And makes them more effective at their job. And so, and causes them less stress because they're, um, you know, they're able to stay up to speed better with a lot of things being thrown at them. So, you know, in my mind, those that embrace the, uh, the efficiencies that those types of tools can bring and. The acknowledging that it's gonna make them better at their job and take away the stuff that nobody actually likes doing. Right. Um, nobody really wants to answer yet. Another question about benefits, right? Uh, or the 401k, um, then they should be super excited that they get to do something that, that adds more value to the org.

Sean Weisbrot: Do you think AI has a place in performance evaluation, which could lead to. Letting someone go or giving them a raise or giving them a promotion? And if so, what does it look like now and what should it be? How should humans work with ai? Or should AI have complete controller?

Matthew Schmidt: That's a really difficult question. Right. And it's part of the reason why we step into, you know, leveraging the latest AI very carefully, right? Frankly, because we deal with a lot of, you know, is servicing whether someone's at risk for attrition and we wanna be. Pretty damn right, uh, that we're not surfacing the wrong information in performance management. You know, there, let's be honest, right? That process is entirely broken, right? The current way that people do performance reviews is hugely time consuming. Uh, not particularly effective, right? Filled with tons of bias. Um, and, you know. Hasn't seen a major overhaul in, in quite some time right now. There are some interesting companies that are out there doing really interesting things. Companies like confirm that are using ONA to uh, kind of turn the, the 360 process on its head and to save you a bunch of time. Uh, but by and large companies are, are pretty inefficient at how they do performance reviews, right? It takes days, uh, for a single manager to do that, and then it all has to be collated and. You know, there is, I think, a huge amount of risk in removing the human components from that process, uh, by leveraging ai. Um, you know, do I think that it could, you know, you could use AI to, you know, try and score the reviews on a couple of different quadrants? Um, yes, you could. Right. You could see, okay, how does this relate to, you know, alignment with our values? Right? Or how does this relate to, you know, completion of their OKRs or their, their goals, right. Um, but I think, you know, offloading that from the person who's most familiar with it is, um, you know, probably going to create more problems than its work.

Sean Weisbrot: I feel like there's a huge problem there when you try to train an AI on your corporate culture and give it the flexibility to go, yeah, I don't think this person is in line with our values. Well, how do you make that determination? Right? If they hit their A, like if you set an OKR, that's easy. Okay. Did you know their, their KPI did. They get five articles done a week. Okay, yes. I've looked at the data. Yes. This, yeah. Right. So if they hit their goals, fine, but determining alignment on on values is very, very nuanced that even humans struggle with.

Matthew Schmidt: Certainly today's AI is lacking intuition. Um, and so, you know, I think that's where, yeah, you, you'd have a real problem. Um, I think that would cause, um. Again, more, more problems than a worth. What would you even train that on? Right. It, it's, I think is the, the biggest, the biggest challenge there. So it really, again, you have to look at it through the lens of what are you optimizing for? Uh, and I think if you're trying to optimize for, uh, you know, giving the managers time back, uh, or, you know, making, you know, reducing the. Recency bias, uh, that's present in performance reviews. Um, you know, you can do those things, but you just have to know what, what the ultimate goal is that you're, that you're trying to do. If you're trying to simply remove managers from the, from the performance review process, like that's not gonna be super effective, but it might be effective in summarizing the performance reviews so that the next level manager is able to. Um, to identify the, you know, where they need to pay attention and drill a little deeper, right? That might be really useful versus, you know, them having to read all of the, the reviews. Um, you might find that the AI is able to suggest, uh, a broader set of peers to provide, uh, a 360. Than simply having the person suggest things. So it it, it all, again, it all depends on what you're trying to optimize for, um, versus, you know, just simply trying to remove the manager from the, from the process.

Sean Weisbrot: Should AI have the ability to promote diversity, equity, and inclusion? And if so, what's the best way to implement that? Don't

Matthew Schmidt: think we should be having AI making decisions for us without some oversight, right? Like right now that we're not to that place where like, I would trust ai just like without spot checking it, right? So, you know, should we let it be making decisions around the, our diversity metrics, right? That's entirely dependent on. Everybody's individual organization, right? You should, you know, ultimately that comes down to, okay, what is the, the shape of our organization that we're trying to achieve? But, you know, it should not be making decisions, particularly in the hiring process based on those types of things, right? Uh, it should not say, oh, this text feels like it was written by someone who identifies as a female. Right. Like if it starts to introduce those sorts of components, um, you know, we're just setting ourselves up for a massive, uh, a massive fail. So, um, yeah, I, I think, you know, you, we still have a ton of work to do on making sure that, uh, we're not introducing more bias into the data sets that we use to train the ais and, um. Making sure you know, that we are, you know, continuing to be focused on the, you know, a fair and equitable hiring process. Right. Um, and for the truth is that most companies who are in the HR tech space don't have the resources that someone like Open AI has to, to train their content. Right. And to train the ai. They, they just simply, they don't have a billion dollars from. From Microsoft. Right. Uh, and so, you know, I think we have to, we just have to be really very careful when we start to let it afa. You mentioned knowledge graphs. What is a knowledge graph? A knowledge graph is really about, uh, there's, if you sort of taking it to your low level, it's a, it's the connection of all of the different pieces of, uh, knowledge that exists within, you know. In our context, we'll take it as your organization. Right. And so, um, but it could be, you know, if you think about it, a, you know, organizational network analysis could be a technology graph, right? That's the graph of the people and how they interact, uh, within an organization. Um, for, you know, one of the neat things that People Logic does is it actually builds up a knowledge graph underneath the, uh. From all of the data that it's collecting across your tools of, you know, how are certain projects connected and so not just the, or, you know, certain deals and customers and those sorts of things. And so, you know, we build not just the graph of the people and how they're connected, but it's the people and what they're working on. Um, and the information within the organization so that then we can start to understand the, the impact to a business around sort.

Sean Weisbrot: So do you think an AI would be useful in sorting through that data to make higher level distinctions that could then act like a business intelligence system to help these companies make better decisions around how the organization operates? That's sort of

Matthew Schmidt: the foundation of the business, right? Um, so yeah, I, I mean, AI and, and truthfully whether it's full AI or, or more machine learning, um, those are things that businesses are good at, right? Or that those types of systems are, are good at, they're good at seeing whether, you know, this type of entity is related to this type of entity. Because, you know, these three people all still are all working on it together. Right. Or because we extracted it from this piece of content and we know that it's over here in this system. Right. So, you know, if you feed it enough of the, the data and you can understand, you know, how to structure it in a way to, uh, find the interesting signals in it, then yeah, absolutely. It, uh, it can surface the information you need to make better decisions.

Sean Weisbrot: What is your hope for the future in regards to AI in general and then about AI in relation to the running of companies? I am

Matthew Schmidt: generally, uh, an optimistic futurist, right? Uh, like Caleb raised on, on Star Trek. Um, um, you know, seeing that there is a better future, uh, available for us, right? So, you know. What I, when I look to the future, and I look to the reasons why we might, you know, invest in fusion power, right? Or, uh, you know, better desalination or those types of things, right? Um, or ai. I look for, you know, how can it make our lives as humans better, right? How can it make our time that we get to spend on this earth or any other? Better. Right. And make us do less of the things that, you know, really are better suited for machines. Right. And so, you know, earlier we talked about HR folks, you know, not having to do the medial tasks because AI can make their lives better. Right? That's a very specific but very simple, straightforward thing. Right. Um. You know, by having ai, you know, do some of the sort of more rote or, you know, it might be service oriented tasks out there, right? Yes. There will be jobs that are disrupted as part of that. Uh, it's on us as a society to figure out how we can help people reskill, uh, to be able to accommodate for that so that they be working on their, something that. Is more rewarding and fulfilling and more valuable.

Sean Weisbrot: So you think AI's role in our future as a species is to free us from the monotony of work so that we can live better lives. Would you agree with that?

Matthew Schmidt: I would agree with that in my mind, right? As a, an optimist, uh, looking at the future, right? Just like I believe that, you know. Using, you know, future power technologies can save us from a world of scarcity, right? Um, those things are, are important in the ways that we, that we look at how, uh, the future can evolve. Now, certainly they, you know, that requires us all to, you know, think in a similar way and to have the proper social safety nets and. Uh, yeah. My general feeling is that it can free us from doing the things that aren't rewarding.

Sean Weisbrot: Right. Some people want AI to make all of the decisions for humanity. Would you fear that as a potential future? Yeah.

Matthew Schmidt: I mean, that's pretty scary because you're asking about, I mean, look, what makes us human is free will and choice. Um, and I think to give the, to give that up to anyone, frankly. Uh, whether it's AI or not is taking away what makes us human. Um, you know, I think yes, if you're looking at it from the perspective of, okay, these are, you know, the people who are currently making decisions as representatives are filled with, uh, plenty of flaws. Yes, but our solution to that isn't to hand it all off to some AI that, uh, you know, we hope is gonna make better decisions because it's only gonna make decisions based on the data it's fed to it. And if you made that decision based on. You know, even the limited history of the US here, right? You would, uh, it's likely to make a decision that's not, not gonna be in the favor of us.