From SaaS to Systems of Work: The Vertical AI Opportunity with Nick Tippmann

Speaker A: And you have both OpenAI and anthropic doing these massive joint ventures and acquisitions, pushing these armies of forward deployed engineers into corporate America. It tells you that intelligence alone isn't going to be enough. Vertical software and B2B SaaS in general was generally being priced around, call it, uh, a 300, $350 billion market. Whereas Vertical AI now gets to compete and go against a 4 trillion plus services and labor budget. That's really the TAM expansion that we're excited about. As the cost and time of uh, building software collapses, the price of standing out continues to increase. And so it's never been harder to garner attention or stand out in a very crowded market.

Speaker B: Welcome to Investing in Startups. I'm Joe Mager. Our guest this week is Nick Tippmann, founder and managing partner of Tiptop dc. Nick is an early stage investor with an intense focus on vertical AI. We talk shop all the time in real life and so it was fun to get him on the show. We talked about the opportunity around vertical AI, the size of the AI, uh, market defensibility against the frontier AI labs, and lots more. Please enjoy. We talk all the time in real life and here we are doing the show. Good times.

Speaker A: Good times indeed. Excited to be here, Joe.

Speaker B: Well, I love our chats, I love your domain expertise and we have a lot of fun, really deep, nerdy discussions. So it's fun to finally do it here on the show. Your sweet spot is vertical AI and I thought a good place to start would just be for defining for listeners vertical AI and how that's different from what people might think of when it comes to listening. Just say B2B SaaS which I'll, uh, let you, I'll let you run with.

Speaker A: Yeah, for sure. It's a great question and it is a big distinction here at Tip Top. I like to think of it as an evolution in SaaS or vertical SaaS. And so I like to think about vertical SaaS or vertical software. 1.0, uh, was really all about the workflows. Think of companies like Viva Autodesk where at the end of the day they were selling the software and the benefit was efficiency. Then you had the next wave of uh, vertical software somewhere around the mid-2010s or so where you started to add fintech to the workflows and you got big winners like Toast, Shopify, Service Titan, and so now you are selling the software and the fintech rails. The way that I look at vertical AI and think about vertical software, what I call 3.0 is you still have the workflows, you still can have the fintech, but. But now you have the agents as well. And no longer is it just about the efficiency. It's about actually doing the work and selling the outcome. And so the way that I like to think about the distinction again from vertical software to vertical AI is really the unit of value has shifted from time saved to work delivered. Uh, you could think about really all B2B SaaS sold time savings or efficiency. At the end of the day, vertical AI looks to sell the actual outcome. And that changes a lot of things. It often will change the buyer, change where the budget's coming from, and probably most importantly, changing the pricing model as well.

Speaker B: Okay, that's helpful. And one thing I wonder about with vertical AI is TAMs and how to think about those market sizes. Because you could say it's a progression, some kind of evolution. But how do you think of TAM for a vertical versus the traditional vertical SaaS? Incumbents appreciating they're doing different things. It's not like one is necessarily replacing the prior version like Cloud might have replaced on Prem or desktop. But how do you think about those TAMs for a vertical on the AI piece versus the traditional incumbency that's there?

Speaker A: Yeah, uh, I will, I'll talk about the tams here in a second, but you slipped something in there that I want to come back to that I think vertical AI replacing SaaS or vertical SaaS is likely at the wrong framing. And I feel like the best vertical AI is vertical software with an agentic outcome layer and harnessed around that workflow. Basically all the top vertical AI companies think like, even up Harvey, gcai, Abridge, et cetera are really all software underneath them as well. And that builds into the evolution of vertical software to vertical AI. Also builds in the evolution of, um, Software TAMs to AI TAMs. And so going back to the 1.1.2, 3.0 framing in one point, uh, zero with workflows and software only, uh, you could generally expect to Capture maybe about 1.1percent of a market's GMV. And so if there's, uh, a billion dollars flowing through a market, you capture a billion dollars as the tam. When folks started layering fintech onto their products, we saw a TAM expansion of anywhere from 2 to 5x. And so now you see folks like Toast and Shopify and Service Titan able to capture anywhere from 2 to 5% of the spend going through a particular industry segment vertical. You can kind of chop it up and down, but wherever you're focused, you can expect to get about 2 to 5%. Now, with agentic AI or vertical AI, you are massively expanding the TAM. Now all of a sudden when you're doing the work, that opens up the TAM to the entire service budget plus all internal and external labor spend. And that's really the bet with vertical AI or applied AI companies more broadly or in general. Just a, uh, couple of stats to validate that that we've seen flying around. McKinsey's saying that labor spend is anywhere from 5 to 10x the IT budget in every single market. Yet Sequoia's now famous services or the new software piece where they mention there's about $6 in services spend for every dollar in software spend. Using the example of a company paying maybe 10k a year for QuickBooks, but 120k to their accountant who actually closes the books. There's some other markers out there talking about outsourced services sized at anywhere from 3 to 4 billion or trillion. Excuse me, wrong, uh, letters.

Speaker B: When you start getting this many zeros, it's easy to lose track.

Speaker A: Yeah, exactly. So I guess to wrap it up like vertical software and B2B SaaS in general was generally being priced around, call it a 300, $350 billion market, whereas vertical AI now gets to compete and go against a, uh, 4 trillion plus services and labor budget. That's really the TAM expansion that we're excited about and that's the opportunity that we're seeing with these vertical AI and applied AI companies.

Speaker B: I was going to ask you about incumbency and why more of the vertical SaaS players haven't built more AI tools into what they're doing. But then I started to think about some of the companies in your portfolio where there wasn't an incumbent doing what they're doing. There were plenty of, like, if you take legal tech gcai, there are, there's almost an endless number of legal tech companies, software players doing all kinds of different things. And yet the actual function of what GCI or Harvey or Lagora do was greenfield relative to what existed. So maybe the actual question I'm leading to is when you're looking at vertical AI companies, are you specifically looking for more of a greenfield situation, or are you comfy with and happy to go? With situations where there is an incumbent, it's a little more clear what the incumbent might want to do, but there's someone coming in with a different approach to it that you think can just out hustle, outplay out, distribute the incumbent.

Speaker A: Yeah, I'm right there with you. I'm actually probably a little surprised that we don't see more of the incumbents system of records or vertical operating systems doing more with AI. And I think really at the end of the day, having worked at Opera, being an operator at heart and being in companies like change is really hard and they definitely probably have some architectural and complexity from a technical perspective that hampers them and holds them back from really executing the way that they want. But at the end of the day, like changing people is hard and changing a behemoth of an incumbent that's been around for 20, 30, 40 years and has a cash cow on their hand, it's even harder to get them to move. And so I think there certainly is the opportunity for the incumbents to go up into the agentic or the action layer. Time will tell on if they are able to actually pull that off. One way that I do like to think about it and look at it when we're evaluating opportunities and looking at different markets, I like to think about how nimble is the incumbent in the market. Are they a 30, 40 year old sleeping giant that hasn't innovated anytime recently and have used their pricing power, their system of record lock in their control points to kind of rest on their laurels? Or, or are they still very nimble, fast moving, sharp executive team? Uh, I think of companies like Toast or Service Titan or Shopify, a couple of the what I call Vertical software 2.0 Companies that are certainly not asleep at the wheel by any means and yet you see them coming out with agentic products and moving quickly as well. And so there's certainly a lot of attack vectors, but also ultimately at uh, tip top. Our thesis and belief is that the ultimate end state of what these best vertical AI companies look like is a system of work. And that generally is a combo of what maybe traditionally would be considered a system of record and now has been often called a system of action or a system of intelligence that sits on top of the sars. Really love and are excited about companies that are taking on the whole approach and going after building these system of works in niche verticals and industries.

Speaker B: You'd mentioned you're an operator at heart. You and I have had a lot of conversations about marketing. As you were a former CMO at Greenlight Guru, I'd love to hear about how you view the go to market motion and model with vertical AI versus vertical SaaS. And obviously given the lenses you have, you'd have a better line of sight for what makes for a good model at early stage, but would love to hear more about that.

Speaker A: Yeah, for sure. So there's maybe one of my more contrarian takes that I've had over the last year or two is that more has been staying the same from a go to market perspective than changing. Maybe I'm coming around a little bit on that. But a lot of the principles in the foundation stay the same. And so let's just run through a uh, couple of points that I don't think are changing the importance of building trust at scale. Scale, things like content, community, events, real connections, domain expertise, authenticity, empathy with your buyer, none of that is changing and I would actually probably say maybe it's even more important today than it has been in the SaaS era. I also don't think having domain Experts as your AES or uh, your CSMs or now as we like to call them and rebranded them forward deployed engineers is really important. So like at Greenlight Guru we would hire medical device engineers that had 10 plus years of experience as our customer success managers that now often would be called forward deployed engineers as many of them were classically trained engineers that were coming into a services role, which is probably exciting for many of the folks back there. But that context and judgment that those domain experts have build trust. And uh, I think of that really back to the first point of building trust at scale. Some of the other fundamentals that aren't changing, like having a tight icp, value based, selling quick time to first value and then back to that building a brand. And again I think maybe building the brand is as important or more important than ever in the age of AI. And last thing that I'll say is really the same, but again becoming more important is distribution as a differentiator. I'd say it's never been easier to build software as we all know. But as the cost and time of building software collapses, the price of standing out continues to increase. And so it's never been harder to garner attention or stand out in a very crowded market. So those are some of the things that I think have stayed the same. What I think the, one of the biggest changes that we've seen is really pricing is being completely rewritten right now and it's really all over the board but and uh, on the fly. And on the fly a lot of testing, which is actually I say a uh, best practice in iteration. The best companies were often always testing and reevaluating their pricing models. But yeah, it's definitely changing. And so at the highest level Think of outcomes over seats. So outcomes per action, per agent, per workflow, they're all on the table. Think of some examples of vertical AI companies like even up per demand package, customer support like people like Sierra Fin per result ticket, you got clay per enriched record or as you mentioned GCAI a uh, couple of times now being priced against outside council spend. But actually one of the things that gives a little bit of comfort for us folks that have been operating for a while. One of the things I don't think has changed is the principles that goes into best practice pricing which are generally a model that has a good, better best format, has some sort of value component or volume component tied to it and then some sort of value metric as well. But the levers inside of that framework are what's greatly shifting. You can kind of think of it like hiring an AI replaces this software. It's a lot there maybe one or one or two others on what's changed and stayed the same. But I uh, go back to this. A lot is staying the same but maybe the importance of things are changing. And I think the importance of data and rev ops and go to market engineering has never been Greater in the SaaS era. Uh, I like to think of them as a differentiator. Having your data pipeline set up properly, having the right foundations in place could be a differentiator for you. And was one of the ways that we were able to really punch above our weight class at greenlight guru at a bootstrapped company going up against many incumbents that have been around for a long time was that we were able to target better segment, better sell better, better conversion rates, et cetera because we were always looking at the data. Now today I believe the importance of data and go to market engineering and rev ops uh, is really table stakes if you want to keep up in the age of AI. Data is fuel to the LLMs. And it needs to be categorized, saved, uh, cataloged all properly to be able to be retrieved and edited in the right way. And so if your data model isn't in place then no AI on top of it is coming to save you. And so it really is core and foundational to get your data right. And then maybe the last thing that I'll say is change has really been the expectations for, for executives and ICs with everything moving so quick. If you are not AI native and learning AI using AI in your day to day workflow or day to day work, now's the time. It's been long enough since the ChatGPT moment that if you're looking at your team and they haven't adopted, uh, it's probably no longer the time to upskill but probably more of the time to upgrade talent, you definitely want to look at automating your business end to end. There's probably a lot of org chart changes coming. I think today an AI native go to market Engineer Rev ops professional could probably do the work of a team of five marketing ops professionals in 2018.

Speaker B: An existential question I think all software investors have when looking at whether it's vertical AI companies really anything with AI is how an AI software company positions themselves to offer a better solution than what a more general horizontal model can come in and offer. And I'm curious maybe at the general level how you think about that, which of course is very meta question but it's something we all deal with and kind of the high level framework how you think of it. But maybe then also when you're looking at individual companies, like a little more tactically how you approach thinking about defensibility from the big LLM providers.

Speaker A: Yeah, absolutely. It's the uh, multi trillion dollar question on everyone's mind these days. Right. And I will say like, yes, it is a fact that the labs are moving up the stack. Ah. And moving more into horizontal apps. And I think investors are right in many ways to be worried. And I think what you talked about there of uh, having a framework to try to understand what gets eaten by the foundational lab companies and what's safe or what's not in the kill zone. Look, I think the labs are coming for huge swaths of the application layer, but the application layer isn't one big homogeneous prize if you will. It's hundreds and if not thousands of potential segments, markets, industries, etc. I think one of the. And so I think it's really important to understand where the models could be heading and where they couldn't be heading and what will be defensible over time and won't be. If you're in these large horizontal categories like coding or writing or image generation and you're essentially a thin wrapper on top of the models and the models getting better is scary for you. Uh, as I think Sam Altman said a couple of years ago now, like if the model is getting better or scary for you, then you're probably in the wrong business. But if the model is getting better for you, actually enhances your product and makes things better, then you're probably sitting in a good position. Now, a lot has evolved since he made that Comment a couple of years ago. Now we have the history of them actually moving into the application layer, but I think one of the biggest, most recent tells that intelligence isn't going to solve everything is actually the model companies moving into the services layer. You've seen both OpenAI and Anthropic both complete either massive multibillion dollar acquisitions or joint ventures to essentially create armies of uh, forward deployed engineers to go out into corporate America and make AI work. And I think that in some ways that's an admission, um, that intelligence alone isn't going to be enough and you're going to need the context, you're going to need the harnesses and it is not as easy as just turning on chatgpt or Anthropic or spinning up quad and you're going to rework your business.

Speaker B: It's never been easier for founders to whip together a demo or product quickly, easily, and something that if they're showing an early stage investor, you or me, over zoom, it looks like there's a working product. As it turns out, some of these things are a little rinky dink. But where I'm going with this is how do you separate the real product at pre seed and seed versus something that's whipped together? And uh, maybe a question that's actually above that is to the extent that you actually care, like do you care that the product at that stage is more than something that's lightly built to show.

Speaker A: Yeah, it's a good question. And the early product is certainly important. It may not need to be the product that scales. And so a lot of it, I guess when I think about that very specifically on the product is probably an area I don't place as much importance maybe to your comment there. If you really care, if they have an experienced CTO engineering leader, someone that you know is going to be able to harden the system over time, then I think it's a positive that shows you how much they're able to do with how little and move fast. But then that goes back into the defensibility question, which is probably the area that I spend more time on. It's like what access to proprietary data do they have? Or is there going to be a proprietary data exhaust? Is there a reinforcement learning loop that will compound and actually make the product better over time so that when the next Vibe coded app starts, uh, it's not a, ah, one to one, but you do have some compounding value over time and so those are both really important. I also think about the workflow complexity today and how that grows over time. And so I'm often very keen to be hearing for a wedge pitch and a phase two, phase three. Now there's been some talk and a lot written recently that a lot of investors are saying kind of screw the three horizons framework and get rid of the wedge. And because the price of software has collapsed to uh, almost zero, like ambition is what's lacking. And you want to hear the pitch that it's not a phase one, phase two, phase three, but we're going after everything day one. I can see the merit to that today. Where I may be grounded in a little bit of truth or actuality in some sense is that in the old days you may, or in the cloud era, old days, however you want to look at it, you may have waited to 10, 20, 30 million in ARR with your first product or uh, your wedge product before you started going on or acting on App two Act three, waiting till you're seeing diminishing returns with your current product line. Whereas today I think as soon as you have something that has product market fit is up and running a workflow that people are using, paying for like now's the time, then what's the next workflow? What's the next workflow? And so this idea that rippling, I think popularized with the compounding startup is probably even more relevant today. And you can think about a compounding startup and compounding workflows that these companies are able to build. So those are some of the things I'm thinking of. I also ask questions around accountability, governance, security, enterprise grade features, are there network effects either in day one or over time? Like oftentimes a lot of vertical software companies end up becoming B2B2C marketplace businesses over the long term. Seeing if that is part of their Act 2, Act 3 or their bigger vision, however they're framing it these days. And then the last one, that is maybe harder due diligence in the early days and a little off track of your product comment. But do they care about brand? How are they thinking about brand? Brand was a huge differentiator for us at Greenlight. You've seen it work for companies like Salesforce. You think CRM, you think Salesforce, you think hr, you think workday. And so I think there's a lot to building a brand and focusing on it early. And again that's an asset that's going to be able to compound with you over time

Speaker B: and showing some intentionality versus someone who just vibe coded something over the weekend and sent you a uh, pitch deck which they Also vibe coded.

Speaker A: Fair enough. Yeah, that's a great point. Maybe to wrap that up real quick before the next question, just to double click, is that they do need to have a vision. I guess, like that's the number one. It's not just what did you build today, what's your tool, what's your product, but what is your actual vision for? In my case, often looking at vertical software of how does this revolutionize or change an industry from where they are today? And folks without that domain experience, without a unique earned insight, find it very challenging to articulate what their differentiation or what their big vision is.

Speaker B: Let's talk some more about GC AI, which you were an early investor in. We talked about it multiple times here. I know it's early in the grand scheme of things. The company raised a 60, uh, million round series B in November. So it's safe to say things are going really well. You're one of the company's earliest investors. I'd love to hear, and I know you and I have talked about this in real life, but I think it's a great story and great example for your thesis and obviously working well. But I'd love for you to tell folks about what you saw at the time when you invested.

Speaker A: Yeah, for sure. Um, so, yeah, very grateful and lucky to be one of the early investors in gcai. And if I think back to when I made that investment in, I believe April or May of 2024, there were a couple of things that really stood out to me. Number one, Chichilia Zaniti, the founder CEO, her deep, deep founder market fit that was going to build trust from day one in the market. And that trust, again, we've talked about it multiple times now, will continue to compound over time. Two, you could tell that her and the company were extremely focused on community and distribution from day negative. One, even from the first time I met Chachilia, before she'd even founded the company, which maybe I'm giving away some of the alpha, but that was part of the tell of what was this, make this a potentially exciting opportunity. Three, this was the early days of LLMs, and it had became clear to me that legal AI was one of the perfect industries for LLMs. You get language in, language out. Now that's where you talk about, okay, well, was there a risk of this being a rapper? Right. Well, that was really just her wedge pitch. And there was a, uh, really detailed and compelling phase two phase vision for the product as well, where they're able to continue to compound and build Additional workflows, compliance, regulatory checks, integration, enterprise grade features, all of that was is, has built defensibility over time and a broader higher level vision of starting as a copilot and moving to what we'd call an autopilot then or a uh, system of agents today. And then finally the last point that was really exciting was that the in house legal segment was growing eight times faster than outside counsel and that would allow them to have pricing power that's tied to outside counsel spend. And so can you talk to why

Speaker B: that is real quick?

Speaker A: Yeah. So, ah, if an outside counsel is charging you $1,500 an hour to do some work, to review a document, to try, check out your NDA to do some red lines or higher, higher level work. But the cost in general to do low level task based legal work is collapsing. And so when your internal GC understands how the cost structure of the external council is changing, then they're going to ask for better deals which over time is going to drive pricing power down. Or as you've seen in the legal market now going a uh, price for fee or a price for outcome. Uh, a lot of legal law firms are changing their pricing model. You've heard a ton about the death of the billable hour over the last couple of years. And so that's kind of the last distinction as well. That GCI does often get lumped in with Harvey and Lagora, but one of their key differentiators is that they are focused on the in house legal segment where both Harvey and Lagora started for outside counsel and law firms, which is actually a very different segment and ICP and value props and use cases than the in M house counsel.

Speaker B: Okay, all right, sorry I took you off course there, but just the fact that in house is growing eight times faster. Okay, well look, well done. I know it's still early in the grand scheme of things, but it's cool to see that working so well. I want to go back to the operator at heart piece and maybe that's true, but you seem to really love the investing side as well. You're an operator vc. What do you think there's been a lot of VC founder combat on Twitter X lately? Um, probably more than usual. What do you think most VCs misunderstand about being a founder? And what do you think most founders or operators misunderstand about being a vc?

Speaker A: Yeah, there certainly has been a lot of chatter on Twitter and X recently. Yeah. And maybe we will get to what the heart of it is today. Or uh, at the end of the end of this question. But I'll start with what do VCs misunderstand about founders? I think they many lack an appreciation for just how fucking hard it is. And I would say don't internalize how much of success comes down to execution at the end of the day. And execution really comes down to people and people are attracted to culture and a big mission. And you can make culture a uh, differentiator if you become a talent magnet. And I don't know that VCs make that full connection between just how hard it is and how much of the actual success of a company, maybe up to 80% of it, just comes down to execution. Now you can't get the strategy wrong. Uh, it's important to have the right strategy, but it's not the place where you spend your time. And it's not often what makes or breaks the company. For founders having now sat on the other side of the table for a couple of years now, I don't think most founders really understand VC fund math and the outcomes that most VC funds need to get their returns and what they're looking to invest in. And that at the end of the day a VC's job is about putting the next dollar into the next best opportunity. And oftentimes that can mean a, uh, company can be a thesis fit in a solid market with good growth. But the growth bar has risen over the last couple of years. You see these AI companies and Bessemer put it out in their state of AI last year of like supernovas and shooting stars and these companies going 0 to 100 million in a year or two and 0 to 10 million in a year or two. Which I think has left a whole cohort of cloud era founders confused and upset about why the next round isn't coming. Because they were told that this is the milestone that they need to get to. They got to that milestone and they feel like the rug's been pulled out from under them. Um, and I think that's causing a lot of tension in the founder, uh, VC market these days.

Speaker B: What's a question you love to ask founders?

Speaker A: Oh, that's a good one. What's the why behind your, you starting your company? Uh, that's almost always one of the first questions that I end up asking the founder. Yeah, you get some interesting responses. It tells you a lot about their motivation, background. There isn't a right answer, but I am looking for the why. Maybe these days the founders that are a hammer searching for a problem, maybe not so hot These days again with the fact that the cost of software is collapsing to zero. And so if it's no longer your technical abilities or talent or ability to build great product, that's going to differentiate you. Going back to listening for an earned insight. Our unique learning is part of what I'm looking for.

Speaker B: What's something you've changed your mind about in the last couple of years?

Speaker A: I'll tell you that VC is a hard job as well. Both sides of the table have it hard and uh, I think we all, it would be good for the ecosystem to have empathy. Back to your point that anybody raising capital, particularly from people that are other people's money, there may be some egos and what goes into it, but yeah, it is a difficult job all around.

Speaker B: We'll end with question. I always ask folks, what's a conventional view in venture that you think is bs?

Speaker A: So a conventional view inventure that I think is BS is probably one that thread through this whole conversation that we've had here today that the foundational labs going up the stack into the application layer actually kills vertical AI. I think that's actually backwards. As we chatted about their own moves prove that there likely are going to be a lot of winners in many verticals. Um, and you have both OpenAI and anthropic doing these massive joint ventures and acquisitions, pushing these armies of forward deployed engineers into corporate America. It tells you that intelligence alone isn't going to be enough. And um, you don't pour billions into FD consulting shops if you believe the next model is going to take care of things. And so along that line, I think the better mental model for AI is more similar to the cloud era than the industrial revolution if you think about the model companies. And one of the points that I was going to chat about was like dude, dude, where's my moat? And uh, kind of the early innings of AI it was the model companies or 22, 23 it was the models and maybe 24, 25 it was all about the data. Now in 26 it's all about the harness. But all that goes back to the models really have no, no network effects. And you could very easily see it being a three, four way race where they're selling a commodity at marginal cost and then the value ends up moving up the stack. And so similar to how cloud gave us AWS, Azure, Google Cloud and then thousands of SaaS companies on top of it, I think we'll see that AI ends up looking pretty similar.

Speaker B: Cool. Well Nick, this is really fun. I always love our conversations and I'm glad we were able to finally get you on the show and bring it here. Thanks for coming on.

Speaker A: Thanks for having me, Joe. Glad we could do it and hopefully there are some good nuggets in there for your audience.

Speaker B: Thanks for listening to the show. We really appreciate the support and would love it if you'd subscribe to the podcast on itunes. Square Spotify, your preferred podcast platform One more thing Investing in startups does not have sponsors, but I want to give a shout out to one of our portfolio companies, Micro One. Micro One helps businesses ranging from startups to enterprises to source world class software engineers. Micro One uses their own AI tools to identify top 1% engineers and get them working for customers in as little as 48 hours. And if Micro One could help one of the largest AI companies in Silicon Valley hire 250 engineers and 30 days, they can probably help you too. If you'd like to learn more, visit MicroOne AI and thanks for listening. Joe Mega is the founder and Managing Partner of Seaplane Ventures. The content here is for informational purposes only and should not be construed as investment, legal or tax advice. The opinions expressed by guests are, uh, their own and do not reflect the views of Seaplane Ventures. Our host guests and clients may hold investments discussed in this podcast. Please invest responsibly.