Michael opens by describing the unusual breadth of his role, protecting an agency that handles business taxes, the state lottery, and three distinct law enforcement divisions, all of which face constant daily attacks. He's direct that the nature of those attacks has changed fundamentally: attackers no longer need to be skilled, since AI agents can now be set up once and left running indefinitely to probe for openings, discovering the kind of zero day exploits that used to require deep technical expertise. When asked whether fighting AI with AI is simply the obvious answer, Michael pushes back with more nuance than that framing usually gets, warning that an arms race mentality can leave everyone burned, and that the real fix is architectural. He describes a least access philosophy that extends zero trust internally, treating even an organization's own network as untrusted, requiring every user and every AI agent employees build to be segmented down to only the specific applications and data they actually need, monitored the same way a service account would be.
The conversation turns to the practical challenge of getting organizations to actually fund this kind of security, and Michael is clear that the winning argument isn't fear, it's framing security spending as a business enabler rather than a cost center, forcing leadership to weigh what could be lost in time, money, and trust against the investment required to prevent it. He's candid that responsibility for security can no longer sit only with IT, and that everyone in an organization has to own some piece of it. On the question of whether AI will eventually make user security awareness training obsolete entirely, Michael offers a genuinely distinct perspective from the more optimistic take Matthew proposes: even if workplace tools someday catch everything automatically, employees still won't have those same protections at home, and Michael continues training specifically to close that gap, since an employee who gets hit personally often ends up missing work to recover from it.
The back half of the episode covers two threads that stick with the listener. First, Michael explains that many of the AI hallucinations organizations blame on the model are actually a symptom of poor internal data, meaning the fix is rewriting and cleaning up an organization's own source material before feeding it into any system, not distrusting the AI itself. Second, he shares a story from his own small community of roughly 4,000 people: after a local fire captain passed away, someone used the publicly posted funeral details to determine exactly when his house would be empty and broke in, only to be caught waiting by police, the intruder turning out to be a deputy sheriff. Michael uses the story to make a simple, memorable point about privacy: any piece of information posted publicly will eventually be used by someone, for something, and understanding that risk is now as much a part of the job as any technical control.
Resources mentioned in this episode
This episode is brought to you by CyberLynx.
CyberLynx is a Bethesda managed IT and cybersecurity company. Local techs you know, not a call center. Month-to-month. 24/7 intrusion detection.
We help growing companies with managed IT, help desk, backup and recovery, and a fractional CIO.
Talk to us at https://cyberlynx.com/contact, info@cyberlynx.com, or 301-798-9170.
Shared Governance in the Age of AI: Keeping the Institution Safe with Bill Guerrero - Ep 240
Playing With Fire: Securing AI Without Shutting Off the Stove with Nakeea Neischer - Ep 239
Build Versus Buy: The Risk of Vibe Coding Your Security Stack with Andrew Dutton - Ep 238
Michael Foster
CISO
Wisconsin Department of Revenue
Matthew Connor: Matthew Connor here, host of the Cyber Business Podcast. Today we're joined by Michael Foster, CISO of the Wisconsin Department of Revenue. Michael, welcome to the show.
Michael Foster: Thank you.
Matthew Connor: Well, thanks for coming on. Before we get too far in, a quick word from our sponsors.
Sponsor Break — CyberLynx: Hackers are using AI to conduct machine-speed attacks. Is your security keeping up? CyberLynx sells industry-leading AI-powered cybersecurity solutions that detect threats in real time, so you know about an attack before the damage is done, not after. Learn more at cyberlynx.com. That's cyberlynx.com.
Matthew Connor: And now, back to our show. Michael, for those who aren't familiar, can you tell us about the Wisconsin Department of Revenue and your role there as CISO?
Michael Foster: So the Wisconsin Department of Revenue is a lot like the IRS, but just for the state of Wisconsin. We manage a bunch of different things besides just taxes, the business taxes, we manage the lottery, on top of that we have three police forces that manage everything from soup to nuts, whether people are trying to break into the lottery, or whether folks are trying to take advantage of some of the different alcohol and tobacco laws, things like that. So my job encompasses all of that, anything security-related, to include some small portions of the physical security, falls under my purview. So managing the information we have inside, to keep it inside, and then managing the allowable access from outside entities, and everything in between.
Matthew Connor: Yeah, well, it kind of reminds me of a combination, on the commercial side, between Chase Bank and MGM, where you've got all this money and people wanting access to it. So when you talk about the lottery, obviously that attracts a lot of attention, that's going to attract a lot of attempts. Do people try as frequently as you might imagine, or is it surprisingly less than you might imagine, to get into the lottery?
Michael Foster: We see attacks every single day on pretty much everything we have out there, so we only make available what's truly required to be available. With the lottery, a lot of that stuff is handled by our vendors, including a couple different gaming systems that help keep people out. So we see the attacks, we see them every single day, and I'd say they're probably more frequent than people probably realize. Now, well, you mentioned it, AI is out there, those attacks are constant, no longer do you need somebody sitting behind a computer, like they used to, wardialing numbers to try to get into your system, they have AI agents now that they can just set and forget, like the old Ron Popeil commercials, set it and forget it, and then come back and get your chicken later. So we're seeing that every single day, where it doesn't take a skilled hacker to get into these systems anymore, they can just buy a tool, or use some of the free tools out there.
Matthew Connor: Yeah, and I think that's the real challenge, because even with that, now these threat actors, thanks to AI, there are now so many zero-day exploits, things we thought we'd patched, we didn't know were an issue, and now they're being exploited at machine speed. So that becomes the real challenge, how, in this modern age, and I have my opinions, I'm curious where you land on this, where the attackers are using AI to find zero-day exploits and attack at unbelievable speeds, how do you combat that with traditional tools, or are they just not enough anymore? Do we need to be fighting fire with fire?
Michael Foster: Yeah, well, fighting fire with fire, I think it can lead to more trouble, is everyone burned?
Matthew Connor: Yeah, you're exactly right.
Michael Foster: Some of the tools we have are sufficient, but I think what's really troubling is we need to get to that world of least access for every single individual user. Gone are the days of a regular VPN where you just open it up and let people in, just like they'd be sitting at work, you can't do that anymore. Your internal network is no longer trusted, don't even trust that, and then hold your users, your employees, to that same level of, I guess, non-trust. So even when they're in the office, your network shouldn't have full access to everything, you should be able to segment it down to applications for each individual user. Where we see the problems pop up is if you don't do that, and you let your employees start messing around with different AI things, and they start building these AI agents, if you don't have visibility into what they're doing, what they're allowed to do, and the ability to actually monitor and shut them down if something happens, just like you would a service account, you're doing yourself an injustice. So we really have to get to that world where everything is least access, you're only going to get access to what you're allowed to, to do your job, whether it's internal or external. And that same thing goes for when you have external customers coming in, like our business taxpayers, well, lock it down to what they need to do, they need to pay their taxes, don't let them get in there and mess around with anything else in their account, unless they really need that access, then give them that access for a limited time.
Matthew Connor: I think that all presupposes that our infrastructure, as it stands, is secure and sufficient, right? And what's interesting is you said you can't trust your network, and I don't think you can, I don't think we can trust our systems, I don't think we can trust any security. And which is to your point, so you've got to limit access. However, if you can't trust your network, and you can't trust your systems, then you also can't trust your segmentation, right? So even at that, if you're up against an advanced threat that's using AI, and they can basically just drill right through the wall and get in and have access, then isn't the only real solution, the only way we can kind of trust our networks, if we have, let's say, machine learning that's monitoring the network, and when it sees something weird, says that's not normal, because I can't do it, humans are real bad at that, there's so much data flying, we drown in it, I can't do it. But machines are great at it, they see the pattern, they see something weird, that's where they're great. And I think isn't that where we have to go with our security, to where we can say, you know what, this is the thing I can kind of trust, I can kind of trust that we've got eyes on our network and our activity, and it will find the weird anomaly, and it's like, whoa, how'd that happen, they got in this way, that's so weird, okay, cool, didn't know that was a problem, they literally walked through the wall, how they did it doesn't matter, they walked through the wall and now they're in the vault, now we have an issue. I know that's just some weird analogies there, but I think that's really where we are today, and the future's going there, because the bad guys are doing that today. And I think as AI continues to advance, will we find ourselves in a world where we really need to be leveraging the machines to fight the machines and protect us?
Michael Foster: Yeah, you hit the nail right on the head with that, because you have to know your environment, and how do you know your environment, you monitor it, you monitor everything, all the way down to each individual appliance, and have that as a consolidated picture somewhere, to make some correlated sense out of what's going on in your environment. Where folks struggle with that is the cost of these things, trying to justify that cost, and people in my position have to show it as a business enabler instead of a cost, what could we lose in time, money, and effort if we don't do something like this right now? And there are lots of tools out there that can do this, some are much better than others, and it's just a matter of being able to sell that, as I said, not as a cost, but as a way of doing business, because no longer is IT the sole holder of security, everybody has to be responsible for security. And how do you do that, how do you get that ownership to everybody, to realize that IT is not going to be your savior, we have to have some tools, and we also have to have you helping us out, looking for these different pieces, but monitoring, yeah, you expect, you inspect.
Matthew Connor: Yep, and I think that's really well said, positioning these things as business enablers and not just some cost center, like, oh, we need more security, yeah, of course we need more security, because the bad guys are using more advanced weapons, we need more advanced defenses, and that's just the reality. And I'll be interested to see, there are a lot of people who I think fear that these tools are cost-prohibitive, and I don't think that's true anymore, I think early on that was the case, but as more and more tools come in, the prices are becoming more and more competitive. And I think for the general population, it's a matter of getting over that initial, that's going to be super expensive, that's kind of like science-fiction stuff, I know that's going to be expensive. So I think part of it is finding out that, one, it's really not, and two, seeing tools like Darktrace, super cool stuff, using machine learning to monitor the network, your endpoints, your email, I think it gives us a glimpse into the future, where you were talking about bringing in the users to be part of security. I think if you look at an organization as it currently stands, if you were to roll out machine learning for your email, your endpoints, your network, and everything is monitored, I think you get a glimpse of the future where, unlike the traditional email tools, where so much good stuff gets filtered out and so much bad stuff gets through, that it's basically worthless, the old filtering, well if I'm still having to rummage through the garbage to find stuff, what's the point? But with machine learning on filtering your email, it gets so good, and the email security training is so personalized and so good, that I think we see a future, and this is a hot take, people are going to disagree with me all day on this, but I think we get to a point in the not-so-distant future, because things are advancing very fast these days, where Jane in accounting gets to focus purely on accounting, and we never talk about user training or awareness training for users ever again. I think we can get to that point where security becomes a thing where we, as security professionals, finally have the tools to fully do the job between us and the tools, to where the user can just do their job. And I may be an AI optimist, I may be Pollyanna and way too optimistic, but I'm curious, feel free to disagree and tell me I'm crazy, but I think that could be a future we see in the not-too-distant future. What do you think?
Michael Foster: Well, it's possible. The only counter I'd have to that is, they'll have the tools at work, but they won't have those same tools at home. And the way I approach it, as far as security awareness, is I want to make sure they're just as secure at home as they are at work. Even if we have tools that can take care of it all for you, it's still beneficial for us to look at that and help them understand they can protect themselves at home by doing certain things. I've found that if employees don't take that to heart, and they get hit at home, they end up taking time off to try to straighten out their life, whereas if they'd taken a few things into account beforehand, they might be able to protect themselves from it. So I can still see there's a disconnect between what's out there for business tools that can help with that, and what's out there for the home user, so I like to bridge that whole gap with our security awareness training. So even if it does fully go away, I may still provide some of that stuff in our environment.
Matthew Connor: You raise a really good point, because maybe it's because of guys like you, and maybe it's because of the security awareness training people get at work, maybe that's why the working population falls prey to scammers much less than our senior citizens. The numbers for senior citizens, it's so heartbreaking, how many lose their life savings to online scams. And going back to my AI optimism, I look forward to a day where the processor is so good on your iPhone that Siri can be listening to your call, it all stays local, and when it's "Microsoft support" calling grandma and grandpa to help them with their computer, it says, grandma, we need to hang up, because this isn't Microsoft support. Well, thanks, Siri. And the same for that home user, on their personal computer, where it gets rolled out even in Gmail and Microsoft Outlook, when they're using more advanced AI filtering. I think it's possible we could get to that point, but I think you're right, there will be a period where at work we're going to be ahead of them, and at home there's going to be that lag, and it's painful for people when they fall victim to this stuff.
Michael Foster: Yeah, that's definitely true. And in my new-employee orientations and things like that, I ask people to bring up some stories, and a lot of folks will say, yeah, my folks have had this happen to them, and I'll have other folks in the room who are already part of our team, just waiting for their turn to talk, and they'll say, hey, I teach this to my folks at home, and we've prevented this and that. So it helps reinforce some of the things we try to teach there. I use those teaching sessions also to learn, to understand what people are seeing, but also where I may have missed something in our training curriculum, people who fall for different things, so I gear my training towards things we may have missed.
Matthew Connor: Yeah, I like that. And I think part of the challenge with the older population is that technology tends to be so confusing for them. And I think especially now, once again, I do hope AI will help them be more proficient with technology. Because just rewind ten, twenty years, you had to be so precise, you had to very particularly do things to make the computer work properly, and it wasn't very forgiving. But now with AI, you simply use natural language. So will we move, in the not-too-distant future, away from computer interfaces the way we're looking at them now, and toward a much more intuitive interface, where the senior citizen will be talking to their phone, hey, call my grand-nephew Bobby, and boom, Bobby's on the phone, instead of trying to find it, and instead of using it like a computer, using it more like a person? Do you think we're moving in that direction, where it becomes so much more intuitive and natural, and less type, click, moving a mouse? And does that help those senior citizens not fall victim, because there's nothing to open, the email comes in and AI says, oh, I filtered this out, this wasn't your great-grand-nephew Bobby, this was a scammer, we don't have to worry about it. Is that the kind of future we're looking at?
Michael Foster: Yes, definitely, we see it already, we see it coming already, with a lot of different tools that are out there already, even in your cars, talking to your car for directions, things like that, it's already out there in more plain language, you don't have to use specific keywords anymore. I think the only issues and risks we could run into are, where are your prompts being stored, and what models are they training, and what country has access to them, can they be hacked and taken advantage of to find out, oh, you know, Maude over here has a nephew, Billy, or something, that she talks to every week, and make some sort of a scam out of that type of stuff. It's all about the data, and right now, it's, we're moving that way, but I still see a lot of risk with the way AI is currently working. Everybody wants to use it, but they should want to build their own space for their own large language model, and only use their own data to train it, and if they don't, then they have no idea where the data they're using is going, and who else can have access to it. So putting those guardrails around their data is beneficial. But the other piece we're finding out with this data is that our own data is causing the hallucinations we hear about, it's not so much that the AI is hallucinating, it's because of the way we wrote the products we told it to use. So now organizations are spending time rewriting their data just to get it in and get it correct. So yeah, it's a lot of work with AI right now, a lot of work for these organizations. And honestly, I think if an organization is going to take some of this on, and look at their data and say, we want quick, easy answers for our call center based on our regulations, they need to do a lot of testing with the data behind it beforehand, and rewrite it first, before they put that stuff into production.
Matthew Connor: Yeah, and another really interesting aspect of this whole AI guardrails topic is, even when you've set that up properly, and you've got your agreements with whomever, let's say you're using Anthropic, and great, they're keeping your data segmented, it's on your server, it's private, fine, let's assume all that works. Interestingly enough, though, what's challenging for organizations now, and we're starting to see things, I'm going to sound like a Darktrace fanboy, but they've recently come out with what they call Secure AI, which allows you to see what your end users are doing when they use AI, on any agent, on any platform. And so now you get insight and visibility, to where, oh, you used your personal ChatGPT, you dropped in corporate secrets, come on, we've been over this, this isn't the first time, and it's like, oh, I forgot, I meant to be in the corporate account, okay, but if you don't see it, you can't change it. And then being able to adjust that, to intervene, and say, whoa, hey Joey, do you really want to drop this into your personal ChatGPT? This looks like company, proprietary information, and this appears to be your personal ChatGPT account, should we maybe reconsider? Oh, thank you, yes, I didn't mean to do that. So it helps with that data loss and leakage, and I think stuff like that gets me excited about where we go in the future, because I think right now, first of all, I think it's a great time, right, I think we're in the most exciting time in the history of time to be alive, we've got self-driving cars, we've got video calls, we're living in the future. But I think when it comes to security and AI, we're in this really funny place, because just a couple years ago, AI was hallucinating all over the place, it was just kind of a glorified, it was a more advanced Google. But now, in September 2026, we're looking at an AI that's pretty darn smart, that in most regards will be smarter and better than the human at the keyboard. And now it's getting to this level where it's like, how are we going to properly manage this? We want the ability to leverage this technology properly, it's evolving so rapidly, how do we, and then we've got these advanced threats coming in, as we're living in a crazy time, how do we do all this, how do we keep it all straight, see it all, and handle it all? And I think that's the real challenge, that's our job now as security professionals, how do we grapple with these new challenges that have never existed before, and will be different tomorrow than they are today?
Michael Foster: Yeah, it's true. There's no real great answer to that right now, it's just looking at what we've seen so far, and then making adjustments as we go along. But monitoring, like we've mentioned quite a few times, monitoring and understanding what's going on in your environment is going to give you that picture, so you can make smart decisions going forward. Do you, as an organization, lock down all other AI avenues and only allow access to your own tool that you've created, so you can actually limit the exposure of all your data? That's easier said than done, because every day, one of your other tools that you bought, whether it's Adobe or whatever, just to throw one out there, they're putting some new function or feature in every single day, and trying to keep up with those is almost impossible anymore. I'll have folks who are the experts for these products call me up every day and say, hey, did you know this new feature exists, we think it's a risk, and that throws my whole day off, now I'm looking at something we've already vetted, because something just got newly put into it. So yeah, monitoring and understanding what these tools are doing is going to be the hardest part, but the most beneficial overall for us as security people, to understand what's going on, so the true decision-makers can get a good report from you, so they can make decisions based on true data. So yeah, it's a crapshoot out there, what you do as a person in our shoes, it really is.
Matthew Connor: And I can't help but wonder, based on what you're saying, we grew up in a time where privacy meant something different than it does today, and the people's right to privacy, and I think for us, in the digital age, it's like, wait a second, it still applies. And yet kids growing up, they're like, who cares, I'm not doing anything wrong, I don't care if somebody's got my data, if they see my stuff and look at my stuff, I don't care. And it's interesting, because I have kind of a visceral reaction to that, and I'm like, but it's the principle of the matter. However, if you take that same sort of privacy, or lack thereof, mentality of the younger generation, and marry that with what we've seen in all the breaches now, every security professional is preaching, it's not a matter of if we get breached, but when, and so we need to be prepared for how we recover quickly. So what we're saying then is, if we really think about it, we're saying privacy doesn't exist, that all this data can go anywhere, and so does it matter? Is that just the natural progression of things, like, well, yeah, the data's going to get out there, it's only a matter of time, so what do we care? Not that we should be flippant with it and just publish it, but where's the balance, obviously changed, from pure privacy being a fundamental right that should be protected at all costs, to now it's everywhere, and it's really not safe, your information's already on the dark web, you can't get it off of there. So how do we deal with that, or is it a matter of, in the future, data privacy is just not going to be a thing, and the rapid progression of things means your data is going to change so fast that by the time it gets out there it's already out of date, because AI is moving so fast, and we're moving so fast, that it doesn't matter? I'm kind of curious, where do you stand on that, am I missing something there, or are we headed to a future where there really is no privacy, between the cameras, the data access, the breaches, and so what we have to do as data professionals is leverage as many tools as we can to prevent it from happening, knowing that it likely is going to happen anyway?
Michael Foster: Yeah, leveraging tools, yes, understanding that it can and will probably happen to your environment is good to have in the back of your mind, knowing it may happen, and it may happen sooner than later. I think understanding what that data is, and how it can affect individual users, is going to be where we have to differentiate ourselves nowadays, so we can help them take action if the data was released. Gone are the days that organizations can come out and say, we had a breach, but we don't know what data was taken, we have no idea what was in that database, well, we'd better know what that information is, we'd better understand how it can be used by somebody, looking at not just breaches, but simple little things that happen on an everyday basis. People think it's no big deal to post, oh, the funeral for so-and-so is happening on this day at this time. And the reason I say that, a small story, in my local community, a small community, four thousand people, a fire captain was killed. They published his funeral, and there'd been a rash of robberies happening in the communities around the area, and they weren't sure exactly how or what these people were knowing. Well, this person was looking on Facebook to see when people would be out of their houses, so this person actually showed up at that fire captain's house during his funeral, and broke in and walked into the house, and it was a deputy sheriff who was doing it. Our police were waiting for her. So I look at it this way, any piece of information you put out there, somebody's going to try to use it some way or another, but understanding what you're putting out there, and what your risks are because you put it out there, is where you're going to differentiate yourself between being a victim or being able to prevent something. And I think that's the same way with businesses, you'd better understand what's sitting there that's a risk to the people who've entrusted you with protecting their information, their data.
Matthew Connor: Yeah, I think those are wise words, I mean, you really do, that's our job, at the end of the day. And yeah, I still can't get over that, it was the deputy sheriff, is that what you said?
Michael Foster: Deputy sheriff, yep.
Matthew Connor: That's like right out of a TV show, that's like one of those Hallmark specials.
Michael Foster: Yep, a hundred percent.
Matthew Connor: Geez, I mean, it just goes to show, you really can't trust anybody, and to your point, you have to be really careful when you're posting stuff online, it goes right back to the basics, you're posting stuff online, you have no idea who's going to use that information, and for what. And so that's really tough, balancing that, and so many people have published public profiles, they want, wow, yeah, that's really unfortunate. Well, you know what, I'm going to go with AI saves the day again in the future here, bear with me for a second, but you look at things like the Optimus robot from Tesla, if we get to that stage, and I think it's only a matter of time, I don't think it'll be that long before they're released. I think what that does for domestic violence, for home invasions, for elderly care, you name it, a lot of problems get solved if every household in the country had an Optimus robot. I mean, you're not breaking in and beating up RoboCop, right, you've got security right there, good luck with that one. You're not smacking around your partner, because you're not going to beat up RoboCop, that's going to stop real fast. I mean, there's got to be a downside, maybe, so what am I missing?
Michael Foster: I think that's perhaps the most wonderful thing imaginable, assuming it does a great job, and it's the perfect household assistant in that regard.
Matthew Connor: Right, is that not the ideal future?
Michael Foster: Sure, yeah, we've seen it on the Jetsons, with Rosie.
Matthew Connor: Yeah, she was great.
Michael Foster: Was great. Well, not only what you're talking about with that robot, but the things that could also help folks who have pets, or lots of pets, and can't take them on vacation, they can't do this or that. Now you can do stuff if you have that, and maybe there's a connection to that robot's eyes, so you can see what's going on at any given time, the pets are used to it, it'll let them out, bring them back in, whatever the case is. That's something we've struggled with in the past, when you have pets, I can't go anywhere because I've got nobody to watch them, and god forbid you use a boarding place, oh, we had a horror story with a boarding place, we took our dogs to a boarding place a few years ago, over a Christmas vacation, because we were going out of the country, and they both nearly died, one of ten other dogs died in the boarding facility, over a two-week period, and it was just ridiculous, and it was rated like the best boarding place in town, it was on the list, it wasn't like we found some hole-in-the-wall, cheap place, it was just pure neglect and a poor environment. And it was right during the COVID time, so you couldn't go into the place to see it, or see anybody, during that time, so they just took advantage of that.
Matthew Connor: And yet with a robot, oh, so great, your pets just stay there, have a great time, and they're all taken care of.
Michael Foster: Yeah, that's a great example.
Matthew Connor: Yeah, I would use that, I think that's going to be amazing, I'm really looking forward to that day, between, yeah, look at me, I'm just solving all the world's problems with AI, that's great.
Michael Foster: We're already doing a lot of it already, with our robot vacuums, our robot lawnmowers, let's get a robot car washer.
Matthew Connor: Oh, absolutely, how great would that be? And with self-driving cars now, I mean, did you see the statistics that came out last month, Tesla's at something like fourteen billion hours driven on full self-driving, something crazy, and they found the statistics are, it's eight times safer than the average US driver, eight times. And I believe it, because we were coming back from New York, probably a month or two ago, and we're in the left lane on full self-driving, and this car beside us swerves into our lane so fast, and had I been driving, if I could have reacted fast enough, and I don't think I could have, because of how fast it was, I would have slammed on the brakes, probably run into the median, the car behind us would have hit us, it would have been a big mess. And we're in the fast lane, so we're doing seventy, seventy-five miles an hour, that's a really dangerous situation, glad I'm in a Tesla, it's very, very safe. However, what happened was the car moved, like they were dancing together, it didn't overreact, it slowed down, it avoided, it couldn't have been a quarter of a second between the time it started moving and the other car started moving. I was like, I'm done, I'm done driving, I couldn't have done better, and I'm still a good driver, I think I'm good with video games, I've got good reflexes, nope, that was, at best, a Formula One driver might have been able to react that fast, maybe, while they're dialed in, but we're just cruising back from New York, I'm not a Formula One driver, there's no way, it was spectacular, it was the most incredible thing I'd ever seen, and I was like, yeah, I'm done, that's all I know. It's like, yeah, eight times safer than the average driver, for sure.
Michael Foster: There it is, so we're going to see it.
Matthew Connor: And were there still accidents?
Michael Foster: It's apparently eight times fewer accidents.
Matthew Connor: But we hold the machines to a higher standard, right, and so it's really interesting, that's a whole other topic. But I can see it, I think that future of people not driving, the household chores, everything being done by robots, I don't think we're that far away.
Michael Foster: Yeah, I wouldn't disagree with that, we're seeing it, and every little accident, or every little issue they find, it seems like they're working on it and fixing it as quick as they can. Some of the folks who have Teslas, they come up to me and say, boy, I like it, I just don't want to pay for it, so it's always that extra charge these companies are looking for, they want that subscription on top of you buying it, so they keep the money going. But yes, I see it just getting better and better, and we're going to see more and more of them out there, I was in Nashville and the Waymo cars were everywhere.
Matthew Connor: So yeah, it's, we're living in the future, it's pretty amazing. Michael, this has been an absolute blast, I can't thank you enough for coming on. But before we go, can you tell everybody where they can find out more about you and more about the Wisconsin Department of Revenue?
Michael Foster: Well, I do have my own spot on LinkedIn, they can just search for my name and find me out there, I post a lot of good different tidbits here and there, my thoughts for the day, things like that. The Department of Revenue has our own website out there, it's called revenue.wi.gov, if you want to see what the Department of Revenue has going, so it's the Department of Revenue and the Wisconsin State Lottery.
Matthew Connor: Fantastic, well, Michael, thanks again, and until next time.
Michael Foster: All right, thank you.