Han Yuan: 0:03

Welcome to Career Journeys Revealed. In this episode, Hitesh and I sit down with Marty Abbott, whose career spans from serving as an Army officer and paratrooper to leading technology organizations through some of their most challenging growth phases. After his time as DTO at eBay, during a period of explosive scaling, Marty co-founded AKF nearly two decades ago. And since then, he's worked with everyone from high-growth startups to the White House. Yes, AKF was brought in when healthcare.gov crashed during the Affordable Care Act launch. In this conversation, we explore the principles that have guided Marty's approach to leadership, the patterns he's observed across hundreds of organizations, and how he sees the fundamentals of building great teams evolving in the age of AI. So Marty.

Marty Abbott: 1:36

Those things didn't really exist, especially for small businesses. And I started developing an inventory management system for his little furniture company while I was in high school. So that that got me interested. But I, you know, when I went to college, I I didn't have a traditional career. I I spent first five years after college as an army officer and another eight years, or pardon me, seven years after that in the reserve. So I didn't really get to use any of my uh any of my computer science degree for for quite a while after graduating.

Han Yuan: 2:13

Okay, so um your your interest in CS began very early. Um it kind of took a bit of a pause. How how did you eventually get back into get back into it?

Marty Abbott: 2:27

When I left active duty, the active duty army, I I really didn't have a clue as to what I would do. And I took the first job I was offered, which was as a manufacturing manager. So now, you know, I'm five years out of college. I spent another year and a half as a manufacturing manager at at Motorola. And then there was an opportunity to work in what was then called computer-aided manufacturing, which were you know robotics on on the assembly line, et cetera, much of which I had been familiar with as a manufacturing manager. And and when that job came open, I applied for it. It was part software, part electrical, um, you know, both developing pick and place robots and then then programming them to work for guys.

Hitesh Chudasama: 3:17

Going from the military role to the civilian, you know, a technology company, was that a big transition for you?

Marty Abbott: 3:24

In many ways, yeah. I mean, after, you know, a lot of time in well, a fair amount of time in the active army and and and living the army life, being a civilian is just really different. One, you you don't fear for your life every day. That's nice, right? But two, um, you the tools you use to motivate people are are slightly different. Leadership's still leadership, but on a day-to-day basis, you know, you have some tools you don't have uh when you're in the army versus a civilian, and vice versa is true as well.

Hitesh Chudasama: 3:57

I know that within the army is a lot of top-down. Uh, did you feel like you have to go ahead and change that paradigm as you, you know, shift it to working in more within Motorola and other companies going forward?

Marty Abbott: 4:10

Missions are top-down, but the approach to mission is almost always bottoms up. Right. So what a a group or a team or a company or a squad or a platoon will do is is handed down by superior leadership or those above you, but but the path to accomplish that mission is developed by the individual. That's not, I think that's a fairly common misconception, army versus civilian, um, because the true the same is true when it works well in the civilian world, right? You'll be given a goal or an objective or an outcome and need to develop the approach to get that outcome. In the uh in the army, it might be taken hill, hold hill 314 against all enemies, right? In the civilian world, it's make $20 million in revenue off of this new objective associated with the project. The team, when things work well, the team still needs to figure out the path to accomplish that.

Han Yuan: 5:04

So at this time, you're you're kind of figuring life out, um, you're in civilian life. Um, how did you eventually move all the way to like internet scale and eBay? Because that that seems like kind of a stretch. And the timescales that we're talking about are actually pretty short. I mean, because you know, back in 1986, um to you know, say late 90s, that's that's not a lot of time. That's it's kind of a lot of change very quickly.

Marty Abbott: 5:36

Yeah, uh I think the answer to that is a bit unfortunate and and is likely true with most people, I think. And that said, I don't think I was ever completely prepared for any of the jobs that I had. That's true when I was a platoon leader in the army, um, you know, into being a captain in the army and then becoming a manufacturing manager, uh, you know, at one time a software engineer and electrical engineer, manager of the same skills, you know, software and electrical engineering at Motorola, and then on through my career. I don't, I think with every promotion, every promotion was a stripe. So I left the Army in '95, was at three companies, you know, between '95 and 2002. I joined eBay in 99. By 2001 or 2002, I was on executive staff at BBA. And there's just no way you can be prepared for that in the course of seven years of being a civilian, right? I I wasn't, I wasn't prepared. I think, frankly, if I evaluated today my performance back then in the fashion that we do within AKF, which you're familiar with, Vaughn, I wouldn't have done well. I don't think I would have given myself, I don't think I would give myself a great grade today, but you grow and you learn. Right. And as long as you apply yourself, and as long as it's, I think this is an important point, as long as it's not about you, and you're willing to drive yourself and you have an appropriate level of intelligence. I don't think you have to be super intelligent. I think you can have, and I think I've told you before, Han, I have a 98 IQ, right? I'm the I'm the dumbest guy in every room, guaranteed. The best day of my life was when I found out the average IQ in the US was 98 and not 100, because I became average. Like I was below average, right? But if you can do that, if you can drive, if you can make it not about you, about something else, about something bigger than you, you can grow into jobs. You just have to be willing to fail and have a handful of folks around you who are going to help you when you fail.

Hitesh Chudasama: 7:50

So, Marty, uh, going back to your eBay experience, I know when I joined eBay in 2003, there were a lot of discussion in regards to what had happened in the past, especially in '99. I think this is before you joined. There was a bunch of different outages. And the stories I've heard in regards to the infrastructure that was set up and the rigor that was put into place. Maybe if you could just maybe uh talk to us in regards to like when you came in, how was that environment? And like, what are some of the things that you have to do once you, you know, within the first six, eight months that you arrived?

Marty Abbott: 8:22

Yeah, I think um so the largest outage that I can recall at eBay was a handful of months before I joined. I think it was in the June time frame of 99, sometime in the summer, maybe early July. Um, and and I joined in October. I think there a lot has changed since then, and and then there just weren't a lot of teams, and and I'm including myself in that, that understood how to operate things at the scale that eBay was in 99. Now it's much larger now, but but by 2000, it was the largest commerce system in the world, right? E-commerce system, anyways. And and by 2005, I think in terms of gross merchandise sold, it was um approaching the size of Walmart. Not in terms of sales, right? It what eBay collected, but but how much volume was going through the eBay platform and total total sales. We none of us knew how to handle that. We we had to collectively invent approaches, but all of it centered in discipline, right? Ensuring that that we did the same things every day, that we never allowed something that broke once to ever break again, which requ required a lot of discipline. So you all might remember, you know, the morning operations meetings, et cetera, that were established. Those, we started those shortly after I arrived, um, all for the purposes of ensuring that no one ever took their eye off of past failures, given the best indication of what's likely to fail in the future. And then then we used all of that, right? The learnings from past failures to inform future architectural decisions. And I think that life cycle of data, information, and knowledge, right, and the virtuous cycle of fixing permanently the things that are broken and learning from them and applying that knowledge across the platform, not just on the thing that broke, but where else might that happen? That helps keep you from having problems in the future. And it's it's not like eBay immediately went without outages, right? There were still outages up through the day that I left in 2005. And there's still outages today. Just you, I think over time the team's gotten better. Lower frequency, lower duration, and that's the goal. We're never going to create something that's completely perfect. No one knows.

Han Yuan: 11:11

With that being said, it was it was a lot harder to scale back in those days. Like you were dealing with like physical machines, data centers, so on and so forth. And a lot of those things have been abstracted away. I'm I'm curious, since you know your history with technology is so deep, do you think the the scaling challenges today are easier or harder than they were in the past?

Marty Abbott: 11:37

I think it's it's easier and more cost-effective to scale with solutions out of the box today than it was when I was at eBay for sure. Um, but I think that's created a different problem, and that's that people are now comfortable, engineers are now comfortable not completely understanding what's going on under the covers and assuming that some other component that they've purchased will solve their problems. Um and while we believe as a firm that you should you should be careful about where you spend engineering time, that doesn't eliminate the need for the team to completely understand how something works. The lack of that knowledge, when you buy something and you don't understand what's going on in the covers, invites opportunities for failures that you're not going to understand how to fix because you don't know what caused it. And that increases outage duration. So I think engineering-wise, absolutely things have gotten easier. Architecture-wise, too many architects assume a box on their architecture is going to solve a problem and don't understand enough under the covers of what's going on.

Han Yuan: 13:04

The current environment must drive you a little nuts then, because like you you you have admittedly non-technical people like you know, vibe coding, um, generating software, deploying it. Um, there's there's there's a lot of hype around some of these solutions, like lovable and whatnot, where people, you know, claim that they can build like hundred million dollar businesses from just like a prompt. Um underlying all of this is that there's very there's very little understanding of what's being generated. Like, do you do you feel like there's going to be a comeuppance at some point?

Marty Abbott: 13:41

Well, I think especially when you uh look at this most recent event, right? The advent of of AI and ML, Chat GPT, Gen AI, I I think that solves a lot of problems. And again, on that balance sheet we were talking about, the new problems it creates, creates those, because I was talking about buying a black box. Now you've got another black box creating black boxes for you. And if no one's understanding how it's doing its work, it's very difficult to go run at a Chat GPT or Gen AI system and yell at it to go fix something it broke. Right? Today, as we said, real engineers have to get and figure out what the hell's going on. And if they didn't write it, is it easier or harder to understand it? Now, I'm not saying they should all, you know, all the code should be written by people. Not at all. That's stupid. But I am saying it is creating a new problem where issues, defects, failures, outages last longer than they need to because fewer and fewer people understand what's going on.

Han Yuan: 14:57

Do you have predictions on how this will net out? Like what because that's that's probably ultimately the real scaling challenge, right? It's not it's not building the site, generating the site by talking to an AI, but it's it's actually evolving the product over a period of time without burning your customers.

Marty Abbott: 15:17

I I don't I don't have a prediction because it's so much can happen. And and the real question to me is whether or not folks start to better understand how to employ these things and stop thinking about them as an additional engineer and start thinking about them as things we need to properly control in our environments. And you know, it we have to have code generation vis-a-vis AI. We have to have it. We haven't produced enough engineers over time in the country. We produce roughly the same number of engineers, plus or minus, as we as we did in 1945, right? 50,000, 60,000 degreed engineers a year in the U.S., not of all of whom are U.S. citizens, even while the need has increased significantly. When you take a look at the Great Recession that happened a handful of years ago, the only group of people who were at or near full economic employment during that recession were engineers, right? At or right around 2% or so. Everyone else unemployment went significantly higher than that, right? 10, 15, 20% by by some disciplines. So we we needed, well, and as a result of that, we we created all these boot camps, right? Let's get let's get people, technicians out who can take the place of engineers and write code. They may not understand how to architect things, they may not understand failure rates, et cetera, but they can write code. That's gen one. Gen two now is gen AI, right? Why put people through a boot camp when software can do it even better than they can in much less time? But we're still left with this too few people understand what's going on. If we don't start managing AI in the same way we we truly manage employees, we're gonna have a problem. We're gonna have too few engineers, too many problems, and not enough people to fix it.

Hitesh Chudasama: 17:20

Yeah. No, I agree, Amarty. Uh, so in regards to this particular situation, as things evolve, what are some of the skill sets from your point of view that you see that needs to be gained? Because I think one of the things that you talked about was discipline, especially making sure the fact that as you're as you're developing, having the discipline to go ahead and make sure you have different kinds of validation and checks. So, what are some of the things that you feel like going forward, especially for the next generation of developers or engineers, some of the things that they need to uh have as part of their arsenal to be successful going forward?

Marty Abbott: 17:54

I think we all have a bias towards learning what we enjoy. And software development has become, in one respect, so easy now with the tools available that that many people could do it, and especially once we we throw Gen AI into it. But I think a problem we've been dealing with for greater than you know 15 or 20 years is probably 30 years, is we don't produce great cross-functional engineers anymore. Everyone seems to either be or want to be a specialist. And that really hurts us, right? When I talk to software engineers who don't understand the concepts of DNS or networking, right? Many of those things, or or or don't really understand what a database is doing. Or on the flip side, when I talk to a systems administrator, a network engineer who has no clue about software development, that hurts all of us, right? It creates identities that puts us in conflict and means that we've got to get larger and larger groups of people cross-functionally to solve problems. If if we spent a bit of time trying to get more people to understand the entire domain of what's necessary to produce a product, I think we would produce better products. Like and that starts in education systems, right? What should we change with with the way we teach folks? And what we teach it.

Hitesh Chudasama: 19:26

Yeah, over the years, I've seen the fact that there has been a lot of specialized kind of jobs like front-end developers. And now we're moving more towards full stack and now going to more of the AI engineers. But you're right, the the fundamentals need to still be there because I think if you're developing something and not knowing how some of the fundamentals work, you're going to run into problems going forward as well.

Han Yuan: 19:50

So, how do you think organization design will change as AI is integrated into the modern workplace?

Marty Abbott: 20:00

That's a difficult question. I think and and there are lots of factors involved in this, but I think that that organizations will be able to become more flat with AI. I think that they'll be able to speed up decision making processes, but in becoming flatter, um, it's also going to put more stress on managers and leaders that remain because they'll have a higher number of people with whom they're working, right? Once you cut out middle management, which I think generally is a good idea for speed of execution, it's a bad idea if you're running a nuclear power plant, right? You want rigid structures. You don't want fast change, arguably, right? You want things to work the way they are and you want a lot of overhead to ensure they work the way they are. You flip that around, you're high-risk businesses trying to create profit. You want flat. So to the extent, right, because that that impacts the speed with which you can make decisions. I have a friend who's a CEO. I'm not going to say who it is or or where he works, but he he said that every decision he makes in his company takes a day, 24 hours to make because of the remote nature of their company. Everyone's dispersed. So everything's done asynchronously, and it's very difficult to get people ad hoc into a room and just make the simplest of decisions as we used to when we all were at work all the time, right? You just go grab three people, grab, pull them in a room, decision was made, you move out and execute. Hard to do when you're remote. And I think the same type of thing will come up, or AI will help solve the same types of things in organizations where a manager or leader was just taking direction, applying a little something to it, and then distributing it. Right? The value he or she creates in many cases may not be worth his or her payroll. If automation can do that, we should absolutely do it. But I think that will bring to light a bigger issue with leadership, and that's that, you know, not all of our leaders are well trained. Not all of them are great. You know, and when you put that kind of stress on the people that remain, those who aren't great are going to fail more frequently.

Hitesh Chudasama: 22:28

So going back to what you're saying, Marty, so with AI being leveraged full-fledged by companies going forward, do you feel it'll be much more um effective and efficient because there will be less touch points? Because things will be much more automated and less people, and especially as you said, even middle management, there will be less of it. So is that the way that you're looking at it, where AI is going to be getting leverage from the top from from all different areas of the companies, but then they will have less touch points to be.

Marty Abbott: 22:59

Yeah, ideally, that's as long as it as long as we get as good or better an outcome, right, at lower cost, if AI accomplishes that and removing people and organizations, that great, that's that's awesome. That's what shareholders want. That's what everyone should want. Now, now, additionally, what should happen and has happened in the past with technologies advances like this is that that workforce, especially the highly skilled workforce, it's more of an issue if you're you're unskilled, you're a laborer, you're working on an assembly line, but educated folks in their domain should be able to be reallocated. It doesn't mean that'll happen right away, but new businesses should develop, right? The cost of entering a business and the risk of entering a business will decrease with AI, which means more businesses should start. That employment should go there. Right? Theoretically, it should increase GDP. It just, it's not going to do it immediately. And that's essentially what, at least if history is correct, every other technology advancement is done.

Hitesh Chudasama: 24:09

So as part of AKF, I know you work with various different companies across different industries. Like, are there certain industries that you've you've seen have been lagging in regards to that option? And maybe uh the other way around, which industries that you've seen have been forefront in regards to leveraging even some of the cutting edge technologies?

Marty Abbott: 24:27

Yeah, I think you know, the best research in that area is really the research that came out of Ames, Iowa, and I think the late 50s that resulted in what's today known as the technology adoption lifecycle. And, you know, it clearly indicates how new innovation gets processed through a market or through an industry and how new solutions get adopted. There are always laggards, um, and they're tend to be laggards by industry. Those that are most risk averse tend to be the slowest adopters of anything. And they're easy to point out, right? They're most often they're governments, they're banks, they're insurance companies, or the medical industry, right? Financial services, banks, insurance, medical, government, they're always slow because the risk of adoption is high. The risk of failure is high, and they don't tend to value the promise of something new. They they want they want repeatability, right? And they want to understand what they're using. They don't adopt new technologies. So, yeah, we we spent a great deal of time help helping those industries think through things that were commonplace 10 years ago. You'd be surprised the number of industries that are still debating whether or not multi-tenancy is okay for their business.

Hitesh Chudasama: 25:51

Marty, looking back at uh some of the work that you did, uh you were also at healthcare.gov. And I know within you know my personal experience as well, where being in healthcare, it it the companies it they take such a long time to make decisions, and there's a lot of things that are manual. Like if you had to go back to healthcare.gov, especially with the current technology suites that we have now, like what would you do different?

Marty Abbott: 26:17

Well, our I what I I should point out that our involvement with healthcare.gov was was pro bono and a result of being called by you know the administration and them asking for help. So we didn't design it. We came in after the implosion and helped them with that's really important, Hitash, because they didn't but they did not cause that problem. They fixed it. That's what Marty's trying to say. Yeah, that that I was trying to say that a nice way, but uh and and I also don't think, I mean, frankly, just to be completely clear, I don't think we fixed it, right? We we had a number of suggestions for them. I think a majority of them were taken, but the folks who did the real work were, you know, it was eccentric. Um but it's I I think your problem was like what were the roots of the problem and and how should they be changed? Is that yeah?

Hitesh Chudasama: 27:22

So the roots of the problems, and especially within certain types of industries, and especially with the kind of solutions that we have now, as you look back on it, like how how would you have thought about that particular situation differently and some of the things that could be applied?

Marty Abbott: 27:38

Yeah, I think if you take a look at at healthcarecare.gov and its initial implementation, it would be it's an example of everything you could do wrong. Um, you know, assuming that the first point is is they assumed that they could build this thing based on requirements because legislation indicated everything that needed to happen. That was untrue. So they spent a great deal of time with waterfall processes, and then it ultimately a solution that frankly did not work relative to the needs of the market. So, point one, they should have taken a more agile approach. Point two, they developed three different solutions with absolutely no connective glue and no team responsible for the interaction. And they did so based off of, again, requirements. So when they joined them together, they had very little time testing them. The initial test passed and it failed under load, which leads us to point three. And that's it, there was absolutely no leadership. There were three different teams building three different systems and no single person whose back you could pat or throat you could choke. And, you know, just think about that. You want to farm out a product to three different companies, tell them to build exactly what you've done, and then you're going to integrate it with absolutely no oversight and expect it to work properly. You know, only an idiot wouldn't do that.

Hitesh Chudasama: 29:05

Yeah, there was not a holistic approach taken.

Marty Abbott: 29:08

Right. Yeah. So in and I think sadly that that is the case with a lot of government systems because the federal government doesn't go out and hire the best and brightest in many cases. Period. Right. It's it's not a slight against those those people. They're they're probably good at their jobs, but if you put a lot of them, not all of them, but many of them, into positions like we've had, they've flown.

Han Yuan: 29:40

You know, given your your incredible background, I think, both from like, you know, where you started, the systems that you've seen since. Um I'm curious, like, you know, going back to the present day, like what kind of advice would you give for um say the middle manager right now who Who is trying to figure out, like, hey, I'm too busy doing my day job managing one-on-ones and things like that, and I don't I don't necessarily have enough time to learn this Gen AI stuff. Like, what would you what would you tell that person? Because the conversations that I have in the in the world, there, there's a lot of like FUD out there where there's an entire group of people that are kind of like caught in the old world, but trying to figure out like how do how do I stay in the field?

Marty Abbott: 30:30

I don't think you have any of us have a choice, Han. If we decide not to keep ourselves current, we're making a decision that we become irrelevant. So if you ever, if if anyone's ever in a position where they believe they don't have time to keep current, they must understand that in making that decision, they will become irrelevant. Right? There's no other possible outcome of that. I don't stay current, I don't understand the tech. I want to be a technologist. Where's my job? It doesn't work. So they either have to find time at work or find time on their own. And it's not that different than talking to a high school kid who's thinking about going to a trade school or a college for nothing. Right? And the kid who says they're not going to do anything and they're just going to go out and get a job is saying exactly that. I can start making money now. Why would I wait four years? Like I want to make money now. It's the same discussion.

Han Yuan: 31:32

And what about the kid who, you know, comes out of college and ostensibly the education system failed them?

Marty Abbott: 31:40

Uh well, one assumes facts not in evidence. I don't know if the education has failed them in all cases or if if they failed themselves. Um, there's a number of them that get degrees. I don't I don't think you're asking about folks who graduate with degrees where there's a lot of history indicating they're not going to get a job, right? If you are, then my answer is get a different degree. Like if you're going to get a technology degree or an art history degree, don't blame somebody else that you can't find a job. So now once we remove those, we're talking about again the 50 to 70,000 people who are engineers or maybe the 150,000 in STEM broadly a year that we graduate compared to the millions of folks that we graduate. Those folks, it's just time, right? It's just a matter of finding the job. And the way to get there is to remove your own barriers. If you only want to work from home or you only want to work at work, you're going to have fewer jobs. If you're unwilling to move to another location for a job, you're making that decision. That's for you. Right? You can't hold that against the education system. And once we weed all of those out, then we would have a good base of people to discuss. And I don't know what that number is. I don't know if it's zero or or a hundred thousand. But but I do believe like all of our clients are hiring. Every single one of our clients has job openings. So it's hard for me to believe that the issue is the market not taking engineers.

Hitesh Chudasama: 33:20

The hiring that you're talking about, are they more recent college grades? I think just going back to what Hans questioned, and I agree with you, the fact that from uh from early, you know, engineers who are coming out of college, they do need to be open and they need to pivot based on where where the demand is. But I the the picture that was painted when they actually went in four years ago getting a particular degree in computer science or engineering, that this that would the environment then versus now is very different. And a lot of companies, at least based on the people I've talked to, and and just looking at some of the different kinds of openings, it's more for senior base and more seasoned.

Marty Abbott: 33:58

And that's a problem. That's a problem. Um, as I said, most of our companies, I I think all of them, but at the very least, most of them have positions. The other thing it's true, too many companies want senior engineers. And and that's a problem both for engineers and for the company. You have to you have to have principals, seniors. You've got to have brand new engine. They're your future. You don't want to hire your future when you need it, you want to hire it today. Right? You don't want to, you don't want to wait and then say, oh, we lost a guy, so I need another senior engineer. You want to have hired someone out of college that you've trained, that knows your way, that knows your culture, that thinks the way you do, that you've already interviewed for that senior job. If you don't do that, you're cutting yourself off at the knees.

Hitesh Chudasama: 34:56

Yeah. I mean, just taking the example from the military uh aspect where you have, you know, young cadets that are going in and then they're getting trained to get to the next level. And you need to be able to make sure that that process continues forward.

Marty Abbott: 35:12

Yeah. Absolutely. Yeah. So that I I think you're absolutely right, Natash. That that is an issue. There are companies, that was true even before AI, that would think I need 50 engineers and I can't afford one of them being new. You need 50 engineers, and you can't afford at least 15 of them not being new. That's reality. If you're going to grow.

Han Yuan: 35:41

We're almost at time, Marty. This has been such an incredible conversation.

Hitesh Chudasama: 35:48

Uh, this is good, Marty. Thanks, thanks for sharing your perspective on a lot of, especially as you know, not just AI, but like there's changes that are happening across various different industries. So the workspace is going to look really different going forward as well. So I think having your perspective on it was definitely helpful.

Marty Abbott: 36:04

Yeah, I think it all sadly it all comes down to us not having enough good leaders. Like, I I think that's the heart of Oliver. Like just like people who are about themselves who don't understand that leadership is not about them. We get that fixed. We get a lot of shit fixed. Yeah, I agree.

Han Yuan: 36:35

A lot of them are retiring. So that that's that's the that's the sad part, at least uh uh among my friends. Um yeah. Yeah. Well, this has been awesome, Marty.

Marty Abbott: 36:55

Hitesh, good to see you again. Han, always good to see you. If there's something I can do for you guys, just just let me know.

Han Yuan: 37:03

Hope our our listeners are gonna really enjoy this episode. I I think you um provide a fresh perspective, in a lot of ways, a calming perspective um around the the current environment. So I I think having that historical purview is um really, really interesting.

Hitesh Chudasama: 37:19

Hey Han, I have one more question from Marty. Yeah. So Marty, I know you have written multiple different books. Uh like let's just say if you were to write a next book, what would you go and talk about?

Marty Abbott: 37:31

Well, uh, I am writing a next book, but it's it's no, it's a piece of fiction. So, and who knows it'll if it'll ever be done. I think you my role in in what we do is largely over, fellows. I'm I'm tired of it. I got beat up, chewed up, and spit up, but when your book comes out, I can't wait to read it. Oh, it'll be a piece of shit like my career, probably, but no, I doubt it.

Hitesh Chudasama: 38:04

You had a successful you've had a successful one, and you're still continuing to keep on adding more, so which is great.

Marty Abbott: 38:10

Oh, it's nice. You guys are real sweethearts. Thank you. Make an old man happy. See you guys, thank you. Bye. All right, thank you.