Episode 823 | Hot Take Tuesday: Is A.I. Killing B2B SaaS?, ChatGPT Ads, OpenClaw

Rob Walling:
This podcast is brought to you by Mercury, the banking solution I use across all of my businesses. I manage half a dozen Mercury accounts from my personal single member LLC to MicroConf, our seven figure global events and education platform to TinySeed, our venture fund and accelerator. Mercury handles every one of them. Traditional banking forces you to duct tape tools together and work around slow processes. Mercury doesn’t. The dashboard shows me exactly where each business stands at a glance. The interface is simple enough for daily banking and paying invoices, but can also handle the multi-step approval processes we need when wiring large sums of money to the dozens of companies we invest in each year. There’s a reason more than 300,000 entrepreneurs have made the switch. Anytime founders ask me where to set up their accounts, I send them to mercury.com. It’s free to get started with no in- person visits and no minimum balance. 

Visit mercury.com to apply online in minutes. Mercury is a FinTech company, not an FDIC insured bank. Banking services provided through Choice Financial Group and Column NA Members FDIC. You’re listening to Startups of the Rest of Us. I’m your host, Rob Walling. In this hot take Tuesday episode, my two co-hosts and I dig into topics like is AI killing B2B SaaS, ads hitting ChatGPT, the explosion of OpenClaw and more. I did not intentionally make it an all AI episode, but these are the topics of the day. Before we dive into the episode, I want to let you know that there are two never published episodes of startups with the rest of us. And the only way to get them is to sign up for our email list, head to startupsforthrestofus.com, enter your email to subscribe, and you’ll get an episode titled Eight Things You MustKow When Launching your Startup and 10 Things You Should Know As You Grow and Scale Your SaaS. 

These are audio episodes and they are accompanied by a PDF guide for each that summarizes the two episodes. And honestly, whether you’re interested in these episodes or not, we also send out a weekly recap of each episode of Startups for the rest of us. And you’ll be joining several thousand folks on that list who get the show notes and all the links of things we talk about in the episodes. Startupsfortherestofus.com and subscribe to get all of that. And with that, let’s dive into Hot Take Tuesday. 

It’s Hot Take Tuesday time again, and I’m welcoming my usual band of Misfits. Kicking us off first is … Tracy Osborne. So Tracy Osborne, what do you do at MicroConf TinySeed?  

Tracy Osborn:
I feel like these sessions are actually a really great way for people to see what a dork you actually are. It doesn’t come out. You rest the podcast. Yeah,  

Rob Walling:
Totally.  

Tracy Osborn:
Hi, I’m Tracy. I am the head of product of TinySeed and of MicroConf.  

Rob Walling:
So what do you mean head of product? What do you do?  

Tracy Osborn:
I’ll say the thing I said earlier to you. I am the Riker to your Picard. So I run all the day-to-day. You tell me what to do and I make it happen. Let’s put it that way. I  

Rob Walling:
Love it.  

Tracy Osborn:
Yeah.  

Rob Walling:
Oh, that’s  

Tracy Osborn:
Great. That’s what I  

Rob Walling:
Do. And my other panelist is? It sounds like you’re being announced at a Giants game.  

Einar Vollset:
Yes. Finally. It’s less than one Scaramucci until spring training starts. Did you know that?  

Rob Walling:
I didn’t, but I do now. And aroll set, what do you do?  

Einar Vollset:
I am your co-founder at TinySeed and also the co-founder and, well, founder and managing partner at Discretion Capital.  

Rob Walling:
Very nice. And Discretion Capital is a boutique M&A firm that advises SaaS founders from what, two million to 20, 25 million ARR, and you advise them on the sell side if they’re looking to sell. Speaking of that, you and your team have put together a guide, the definitive guide to M&A for B2B SaaS. It is at discretioncapital.com/guide. You want to give us just a brief summary of what’s included?  

Einar Vollset:
Yeah. It was actually one of my colleagues over at this question was like, “You should write something because we’ve been doing this since for nearly a decade now, and most people don’t really understand what it goes into.” It goes into M&A for B2B SaaS in this sort of ARR range. So really that’s the goal. The goal really is like … And look, obviously it’s marketing. I want the companies to come to us and we’ll help them make a bunch of money, but not everybody will do that. And so even if you’re going to DIY it, it’s worth understanding what does the market actually look like because it’s actually one of this sort of size range is very difficult. Once you get larger than that, if you’re into like 50 or 100 million dollars, then very quickly it becomes kind of known what the market is and what things cost and who the buyers are. 

But between that sort of two and 20 million ARR range, it’s an extremely opaque market. And so what that translates into is like, you can very easily end up in a situation where you as a founder get bad advice, so you end up selling your business for like a third of what it’s worth, which pisses me off. So that’s what’s really the impetus for writing this guide. So we’re putting this together. It’s mostly written. We’re actually launching it publicly and we should have some hopefully physical copies for people at MicroConf in Portland in April, but we’re putting out one chapter a week basically. We just launched the chapter two here going into, on Monday, going into who actually buys these companies, are they? Because people don’t even know that. People don’t even know what is the typical makeup of buyers in the universe.  

Rob Walling:
Yeah. That 70% of B2B SaaS acquisitions in this range are private equity.  

Einar Vollset:
Yeah, private equity or at least private equity owned. So a lot of people, they don’t know what that means. And a lot of people have a view of, I like to joke, like a private equity is like a Bruce Springsteen movie or a song where it’s like they’re always bad and always pay poorly, but that really is a misunderstanding of really the biggest buyer class for B2B SaaS in this size range.  

Rob Walling:
That’s discretioncapital.com/guide. And by the time this episode goes live, I believe you have at least five, maybe six chapters of that will be live because this is a few weeks out. Speaking of that, I actually wanted to ask you a question upfront and it’s in relation to this very podcast, episode 819 where I was asked about QSBS and someone was saying, should we become a C Corp? And I was like, obviously I can’t give you advice, but here’s what I would do. I would do it because I figure I’m going to sell and then I’m going to save on the taxes. And now even after, isn’t it like three years you get 50%, four years you get 75, and five years you get the full thing. And I explained that all on the episode, but the thing that I wasn’t sure about is QSBS means you have to sell the stock. 

So you can’t do an asset sale. And I know there’s a mix of asset and stock sales. So my question that I couldn’t answer on there, and I specifically called out and said, “I should ask Einar in a future episode. Here we are at that future episode.” Around what ARR range or … No, it’s probably more like at what purchase price is a better way to ask this. Do you see SaaS acquisitions changing from asset to stock purchase or vice versa?  

Einar Vollset:
I mean, you can have an asset purchase pretty much any size. It really depends. Probably 90% of the transactions we do, if not more, are stock transactions. Really, I think around about five million is roughly where it starts. We don’t do a lot or any deals that low, but like certainly once you get past 10, most of them are stock. And actually cover this a little bit in the guide, but it’s sort of a give and take. Usually the buyers want to do an asset purchase for tax reasons because that means they can reappreciate the asset and it’s simpler. So this is the industry, right? So what is an asset purchase versus a stock purchase? Asset purchase just means you’re just buying specific assets. So like you’re buying say the contracts, the website, the thing, the IP, but you’re not buying the rest of the company or any obligations or whatever that they have or have not made over the years. 

And so the reason why you typically, typically don’t do a stock purchase is because it becomes a lot more expensive because it becomes just like the legal cost of papering a stock purchase becomes kind of not worth it if your deal isn’t very large. So like my rule of thumb is like around about a million of ARR typically, it starts to become much more common to do stock purchases. But I mean, it does happen, particularly if you have large companies that are LLCs come along and they’re just like, “Yeah, we’re going to structure an asset purchase.” And all things being the same, you’re going to get slightly more money in an asset purchase than you would do in a stock if you’re negotiating correctly. But again, like usually, certainly if you’re a C corp, but even so that there’s usually some tax issues around non-stock purchases that you should be aware of.  

Rob Walling:
Cool. So give or take rule of thumb, you think about a million ARR and maybe five million exit price.  

Einar Vollset:
Somewhere in there. Yeah.  

Rob Walling:
That sounds about right. And I had hinted at that in the episode, but I just wasn’t certain. I’m sure I’d heard you say it at some point. I was kind of just quoting you on it. So with that, let’s dive into our first topic of the day and I’m just going to give a warning to everyone. This is all AI all the time. I went out to look for any topic that would impact us in B2B SaaS and I sifted through all the things and it’s just AI, AI, AI. So that is the episode today. If you don’t like AI, I’ll be back again next Tuesday morning. But first topic is an article that we will link up in the show notes called AI is Killing B2B SaaS. And I think my question, and I’ll kick it to Tracy first is, is AI killing B2B SaaS or will it just be subscription software by another name? 

Is the market overreacting? Because as we’re recording this, the stock market is brutalizing some public SaaS companies. What are your thoughts?  

Tracy Osborn:
The landscaper down the street is always going to need some sort of software to help them run their business. And so subscription software for businesses is going to stick around, but the way that that software is built and the ability for people to start building solutions for those B2B businesses or excuse me, for those kind of businesses is fundamentally changing. It is so much easier now for someone to be like, “Cool, I just want to build a software for five people on my street that are running these businesses.” So as someone who is supporting folks who are building software for B2B, B2B SaaS is not going to go away, but for those folks building that software, they need to be very aware that it’s going to be a way easier for people to build something to do what they’re doing and they have to figure out some way to make your software unique and defensible against these folks who are vibe coding.  

Rob Walling:
One thing I was really surprised by in this piece, which is actually pretty well written, there’s some things I quibble with, but one fact that he was displaying was that HubSpot and Klaviyo are down 30% and analysts are writing notes titled No Reasons to Own Software Stocks. And when I think of how complicated HubSpot and Klaviyo is, it’s like you’re not going to fucking vibe code those. It is probably more powerful than Drip and Drip is complicated as. Anyways, with that in mind, AinR Vollset, what are your thoughts on this topic?  

Einar Vollset:
I mean, I’ve been asked this question a fair bit. Obviously, fundraising for fund three by investors and things like that. I’m on record to say that, I mean, I think even public SaaS companies are oversold. I actually personally rotated into some more public SaaS in the last couple of weeks, which may come back to bite me, but fundamentally-  

Rob Walling:
When oversold, you mean people have been selling them so they’re going down and you think it’s  

Einar Vollset:
Overreaction? Basically, I think it’s a buying opportunity is what I think. And so I bought some. But fundamentally, my thing with it is like this, it’s like, I think the only reason when SaaS will be dead is if we’ve basically invented God. If you’ve invented an AI that is super intelligent to the point that like, okay, we’re ants to it and it can just do anything perfectly straight off the bat when you just ask you to do it, yeah, then I think SaaS is in trouble. But if that’s happening, the least of your problems is like, what’s going on with SaaS. All sorts of problems become a problem then. Instead, where I think we’re at is like, look, I think the models are incrementally getting better, but not as much as fast as they were, relatively speaking, is sort of my view. So it feels a little bit like we’re sort of at the, at least the starting on the top end of an S curve of capability. 

I think that’s extremely disruptive to a lot of incumbents. And so I think yes, there is some risk to some large incumbents because it’s sort of one of the situations where like, okay, there’s a lot of change and a lot of uncertainty. And so if they misplay it, I can see that they are disruptable in a way that they weren’t prior to this AI thing. But the flip side of that is like, okay, I think people lose track of the fact that what we’re dealing with here is software. I mean, people sort of anthropomorphize AI in such a way that they think it’s fundamentally different to like, for example, like I have people who will very happily, like they will not give their email to Gemini, the Google AI, even when they turn off like, “Don’t use my data for training.” And I’m like, “Why? 

Why? Why do you care?” And they’re like, “Oh, because it’s not secure.” And I’m like, “Google already has your data. What are you talking about? It’s in Gmail. You’re giving the data back to the same company. It doesn’t matter. It’s the same thing.” But I think people have sort of like completely lost that a lot of the time. And so they anthropomorphize and think about it versus like what it is is software. It’s a new software capability. And so applying that to public soft stock is like basically the bet. If you’re betting against SaaS or certainly established public SaaS, if you’re betting against that, what you’re basically arguing is the companies that made billions of dollars deploying software are going to do a bad job deploying software. That’s the bet. That is what you’re betting if you think that SaaS is dead completely, unless you believe this other thing, which is that we’ve built God or are building God, which there are people who do, there are people who are convinced that the world’s going to end within the next three or four years.  

Rob Walling:
Yeah. And so here’s the thing, when I say SaaS isn’t dead and I roll my eyes, people say, “Well, of course you say that because everything you have is built on SaaS.” And you know what? That’s not actually true. I mean, I know right now it’s like SaaS and we’re B2B SaaS, but let’s say that agents became the new thing to subscribe to and that in fact, agents had built in databases and business logic and all this stuff. And maybe that happens eventually. Why wouldn’t we just start investing in them? Why wouldn’t I create startups for the rest of us content for them? Why wouldn’t my next book be called The Agentic Playbook by Rob Wallen? These are the same thing. It’s just software by a different name. SaaS used to be downloadable, right? It wasn’t SaaS then, but it was downloadable. Then they were called application service providers. 

ASPs was a name for SaaS before SaaS. And then it was like cloud software became a thing for a while and I think only like super enterprise like Fortune 500s call it that now. It’s just something by a different name. So let’s just … I get it’s a headline and I want to get clicks on social media or on YouTube or whatever. But if you are really stuck in SaaS … Now, maybe the reason that these like HubSpot or whatever people think they’re going down is that they can’t be agile enough. They can’t roll with this. They’re not going to implement AI in some Upstart is going to come in and uproot them. But to your point, if HubSpot just implements the AI in their model or in their app, they’re going to do okay. I think the hypothesis, there’s hypothesis that everyone’s going to vibe code their own app. 

And to Tracy’s earlier point, it’s like, no, we’re not. I used to write code for a living and I’m not going to vibe code a bunch of crap. I might do a little utility here, a little whatever, but the maintenance of that over time and the hosting, just the patches a lot, like I don’t want … And I get it, it could be lambda, blah, blah, blah, whatever. There’s no way that my super smart PhD wife, there’s no fucking way she’s going to vibe code a thing. She’s going to pay for software and most people will too. And we’re in these bubbles where we’re all techies and we can all roll our own thing in a weekend. That’s fine. Go do that. But that’s not … Tracy, you’ve seen how resistant that I’ve been to even writing, having any code that TinySeed maintains because it’s like, I don’t want to be a software shop. 

I know that the headache and maintenance of that. With that, I want to roll to our next topic, which is actually something Einar you brought up already. It’s just a question. There’s no article here, but I had a question for the two of you, and I’d like to start with Tracy again. My question is, are the AI models that you are using today noticeably better than they were six, maybe 12 months ago? We can kind of put a timeframe on it. And so Tracy, maybe talk about any of the models you’re using and what your experience has been like. The reason I’m bringing this up is, sure, it’s anecdotal. It’s an N of three people weighing in, but I do like to hear boots on the ground examples of different use cases because you’re using it for different things than I am.  

Tracy Osborn:
Yeah. I feel like I probably represent a lot of folks who are too busy. They see the hype happening, but they’re too busy to really dive into it. And so I’ve been using AI kind of in surface level haven’t gone deep. And so what I’ve been seeing from my work, I’ve been using ChatGPT and then I’ve started using Claude and I started using Claude because ChatGPT has been driving me up the wall twice now. I posted this on the TinySeat Slack, a whole screenshot of asking about a Zapier question and ChatGPT just lost its … And it just went for pages of not understanding and spiraling out. And it’s happened again with other Zapier questions. I don’t know what ChatGPT has with Zapier. So anyways, that’s my experience with ChatGPT of just seeing like, “Oh, hey, this thing is supposed to be evolving to be better.” And I’m running into these cases where I’m just like, “This is awful. 

I have to move off of it. ” And so I’ve started using Cloud. So that said, for my use cases, which is generally in and around communication, summarizing information, regurgitating information in a better way that I can use in my processes, I’m not really using it for programming yet. ChatGP is worse, Cloud seems to be better. I haven’t seen enough of a difference in the last six or 12 months to say to definitively like, “Oh, these things are better for me. ” They’re already, I would say, pretty much good enough other than ChatGPT just losing it on certain questions. What shouldn’t be happening? It’s very weird.  

Rob Walling:
How about you, AinR? What tools are you using and what are you using them for? And then the question of, are they better?  

Einar Vollset:
Yeah. I’m one of those revolting guys who has the Max subscription and everything. So I have Cloud Max, I have ChatGPT, Pro Max, whatever the most expensive ones. So I use both. The one I’m using the most now is CloudCode, specifically CloudCode, not just Claude. I like the style of whatever, the personality, let’s call it that, of Cloud better than ChatGPT at the moment. But I think the big unlock for me was like Claude code. It was the harnessing around it, the fact that it could go on and understand a whole quote unquote code base. None of the stuff I’m doing is a code base. It was always something else. But that harnessing and that structure around it is sort of, if anything for me, that change, which is not like an AI capability change, was a bigger step up and change than I’ve seen from any model change in the last, certainly last year, probably more. 

My answer is like, yeah, I think I don’t personally think that there has been such a big … Every time there’s a dot change, like 4.5 to 4.6 or whatever, there’s always like a big thing, oh my God, we turned the corner and like now we’re doing this thing and it’s the end times or the beginning of the beginning of whatever it is. And that I haven’t seen. My impression has been that if I’m perfectly honest, a lot of the time I don’t notice that I’ve been upgraded to version five or 5.1 or 5.2 or whatever, to be perfectly honest with you. I just don’t notice it. And so no, I don’t think that the models have changed. Where I might not pick up on it is like, I can see how potentially some of these updates will allow you to be more quote unquote agentic. 

So there’s some like long running things or particularly around coding where I can see like, okay, maybe they trained it in such a way that it can be independent for longer in parallel with a lot of other things that like if you are deep in big, big development projects, yeah, maybe you’re noticing that difference. But for the stuff that I do, like the most advanced stuff I did here was like, I got sick of Squarespace, so I had it port all of the discretion capital Squarespace site over to a static website generator and put it on GitHub. But that’s like, okay, then I got to spank it every 20 minutes because it screwed something up. It’s not like super long running agentic things. That might be the only difference, but personally, no, I haven’t noticed the difference.  

Rob Walling:
And for me, I had been using ChatGPT exclusively until Craig Hewitt busted my chops about it. He really doesn’t like ChatGPT and now I see why not. I started using Mannis a little bit, but then Claude became my thing. It feels like a lot of folks are moving to it. And I found Claude to be, it’s more expensive, but I don’t care. ChatGPT is 20 bucks and Claude is … I think I pay 200, I don’t actually need to. I’d probably be fine on the $100 plan, but I’ve been using it to repurpose some audio and video that I’ve recorded and trying to get it to write prose that sounds like me. So I’ve been training it to be like, “Here are my books, here are my blogs. I’ve written hundreds of thousands of words. This shouldn’t be that hard to match how I write to these topics because I’m also speaking them.” It’s me speaking, turn it into writing, right? 

And it doesn’t get a full finished job, but man, ChatGPT could not handle that. And Claude, I have trained it and I’ve had to give it iteration after iteration of, oh, it puts something out. I do a full edit of how I want it to be, then spit it back in and it gets better and better and better. So that’s one thing. I don’t know if Claude could have done that a year ago because I didn’t try it that extensively, but I do know that I’m impressed with it. Similar though, when I think back a year or two, it feels generally the same. It’s getting asymptotically better. It’s like a little better. I mean, Einar, you had a theory that resonated with me, which was the first generations of all these LLMs were trained on basically the entire internet or as much of it as they could get. 

And so then to get a lot better, you kind of need another entire internet’s worth of material, which doesn’t exist because it’s a whole internet. Am I quoting you correctly here?  

Einar Vollset:
Yeah. There’s a couple of different things, right? It’s like people are always saying, “Just train more data and more compute.” And then if you keep 30Xing the compute, you’re going to get whatever, a step up improvement, 50% improvement every time. And there’s two things there. One is like, you’re just going to run out of training data, which is actually why I’m slightly bullish, quite bullish actually on Tesla, despite their enormous thing, because one of the interesting things about Tesla is like, someone told me this, which stat’s probably not correct, but it’s something along the lines of this. It’s like every second Tesla, the network of Tesla cars gets a lifetime’s worth of a driving experience every second or every millisecond or something crazy.  

Rob Walling:
Interesting.  

Einar Vollset:
But that’s raw data, right? That’s like raw real training data, which like you’re not generating that kind of like textual data, but in and of itself. So there’s a constraint there. You could potentially like synthetic data, but then you have issues with like, “Well, are you really improving yourself improvement?” Yeah, dodgy. So there’s that. And then fundamentally also, it’s just the asymptote stuff just has to do with sort of the increasing cost of compute. I was arguing with someone on Twitter, which I know must shock you guys.  

Rob Walling:
What’s the first time for everything?  

Einar Vollset:
Yeah. But they were basically saying like, look, they had this graph, it was something along the lines of like, a 50% improvement 30 times in a row is still like an amazing, is like a fundamental change to everything we know about. And I’m like, yeah, but if you assume that say ChatGPT-3 costs $20 million to train in compute budget and you’re having for every step up, you have to 10X the budget, like by time like 23 or something, the compute budget is higher in US dollars than the number of atoms in the observable universe. So like, okay, what are we talking about here? Yeah, I know on paper you can make this work, but that’s just not how it’s going to be. There has to be, I think. And I think most people now, if you talk to like leading edge research and stuff, sort of agree with this. 

They’re like, look, there has to be another breakthrough to really start that, to get to the next kind of level, that capability where people start talking about like super intelligence or like something like that. It needs something else. You can’t just throw more synthetic data and compute on it, and I just don’t buy it personally.  

Rob Walling:
If you’re a founder with a strong vision, but no technical partner, you need more than a vibe coated MVP. You need a real foundation. That’s where Designly comes in. Their Solution Lab prototyping Sprint is a focused two week collaboration where you work together with a seasoned product owner, a developer, and a designer to define what to build, why it matters, and how it should look, feel, and function. You’ll turn your idea into a beautiful and clickable prototype, something you’ll be proud to show to early users or investors. Designly is the product team you can actually trust with a track record of helping founders move from being ghosted by dev shops to growing real businesses. And right now, Designly is giving my listeners $3,800 off their solution lab prototyping sprint. To get started today, check out designly.co/fortherest of us. That’s D-E-S-I-G-N-L-I.co/for the rest of us. Our next topic is that ChatGPT has started running ads, is going to start running ads. 

I actually don’t know. I’ve been, since I pay for it, I don’t see ads, but Anthropic ran a spicy Super Bowl ad pointing out … And it was a little disingenuous, but then Sam Altman got pretty angered by this and tweeted, “I wonder why Anthropic would go for something so clearly dishonest. Our most important principle for ads says we won’t do exactly what they’re showing. We would obviously never run ads in the way they depict them. We are not stupid and we know our users would reject that. ” And then it says, “Anthropic serves an expensive product to rich people. We are glad they do that, but we feel strongly that we need to bring AI to billions of people who can’t pay for subscriptions.” So there’s a lot here. Tracy, I’ll kick it to you first in terms of ads in ChatGPT. What do you think about all of it?  

Tracy Osborn:
I was thinking through this before we started talking, I was looking at your notes and it just reminds me that every now and then I pop on a Prime show and now it has the little pop-ups where it’s like, “Hey, this ad has…” Press the button on your remote and you can order the product via your Amazon Prime thing. And my first reaction seeing those was being like, “Oh yeah.” I was like, “I will never do this because I’m a brat and I hate those things.” And I was like, “Why would they do that? ” And then I was like, “Oh, but I’m not the average person. I’m sure that Amazon is only doing this because they see that it works.” And if I apply that over to ChatGPT, if you have the average user in there, and if I saw an ad in ChatGPT, I would just close a tab and move on. 

But I have to admit that I’m not the average person and I’m far more aware of, I think, advertising to the average person and more annoyed by it. So on one hand, I can see why ChatGPT is doing this. It’s not something for me, probably it’s for the person. The other part of it is like, “Really, this is your monetization model.” Going back to why I don’t like advertising, I’m like, “Really, is this the best you could do? Just serve up ads.” I don’t know. I don’t know if I have a really concise thing to say here other than my annoyance that the world is moving into this like, “Oh, got to have ads in there.” And then I think I’m not alone in being annoyed by this, which is why Claude brought that up as a way to needle their competition.  

Rob Walling:
Ainar, what do you think about this?  

Einar Vollset:
It reminds me of Google in the old days. This is how old I am. I think when Google co-founders was like, “Oh, never run ads. It’ll ruin it. ” And then they were like, “Oh, I will run it, but we’ll make it very clear that they’re ads and nobody’s confused.” And now I look at it and I’m like, “Okay, there’s no difference. They’re clearly serving equivalent ads.” I also am reminded of the old Google motto, don’t be evil, which I think that sort of went by the downside a while back too for a lot of stuff. But I mean, on the flip side, I get annoyed at ads, but I’m a little bit more understanding than Tracy. And I’m like, look, it’s the world’s biggest sort of economic model, the business model is advertising. And so yeah, they’re going to do it. Not only are they going to do it, but so is Claude, I guarantee it. 

They might figure out something clever and do the thing, but there’s just not a universe unless Claude get bought by Apple or something that Claude aren’t going to have to do advertising if they’re going to be a public company and grow and if that’s their plan, there’s just no way, I don’t think.  

Rob Walling:
Yeah. Anytime you’re selling a free product, then the person using it is the product, right? That’s the old adage. And when you’re selling to basically consumer-ish folks, that’s been the monetization model for radio, for television, for all the mass media. So there’s too much money and value to be unlocked here to not do it, I think. In addition, I just want to call out that producer Ron and I were talking right before this call of about putting together a YouTube video on the opportunity here for startup founders because we have B2B SaaS companies that are like, well, Google SEO is getting a little scary and Google AdWords are too expensive and Facebook Meta ads are too expensive. You know when Google AdWords was really cheap right when they started, 2005 and 2006. In fact, Jason Cohen talks about, I think it was IT Watchdog, which is his precursor to WP Engine. 

He said he was getting five and 10 cent clicks at that point. And then I built Hittale, grew it from 1,500 bucks to 30 grand a month. It was not exclusively, but a lot of that growth was Facebook ads specifically. And the clicks were in the, I think I was getting 20 cent clicks at the time. It was somewhere in that range, like 15 to 25 cents because it was early. Now, it was the wild fricking West. There was almost no documentation. There were no real books or experts on it. There were no consultants who could do it. And I had to learn the interface myself and I did read like a Kindle book or two and try and err, trial and error to get it. So it’s Wild West, but if I were a startup founder myself, you know that I’d be looking, trying to get in. 

Now it’s not even available yet. But if there’s an early access list or something to get on, I’d be looking to do that. And I think it’d be fun, Tracy, even if we ran some just to test, if ads come out like to run ads for SaaS Institute or any other efforts that I think would warrant that. What do you think about the opportunity for SaaS founders? Do you think there’s going to be something coming up to replace what’s happening with SEO and have a new ad network that is cheaper, less expensive?  

Tracy Osborn:
I mean, yes, definitely. I guess if I look it through the lens of, I hate ads personally, but man, that is like jumping into this brand new way of work that people or the way of life that’s going on right now. Yes, absolutely. We need to. We at TinySeed need to take advantage of this. The folks that we work with at TinySeed to take advantage of this, it’s very clear that this is where everything is going. And the folks that are going to benefit the most from this are the folks who are going to jump on it early. And I know that so many things have come out around this, like crypto and all those things and people are like, “Oh yeah, jump on it early.” And they kind of, from my perspective, haven’t panned out as much as they have, but AI does feel like this is the true sticking.  

Rob Walling:
So a lot of usage.  

Tracy Osborn:
A  

Rob Walling:
Lot of eyeballs, a lot of attention to get in front of. And our final topic of the day is about OpenClaw, formerly known as Claude Bot. And when that came out, I was like, “Why did they name it that? They’re going to get sued.” And they get C indeed real quick. So is OpenClaw hype or is it substance? Is it here to stay or is it the next clubhouse? And we have two articles, one called Open Claus Changing My Life and another call Open Claw is what Apple Intelligence should have been in our Vollset. Are you a believer in OpenClaw and what it can do? Nodding vehemently, yes.  

Einar Vollset:
I love OpenClaw. I think it’s great. CloudBot, it’s amazing. I have it running for me. Looking at the Mac Mini, I bought for it. It’s great. Lots of issues with it, but it feels like … Actually, the title of this article is sort of how I feel about it. This is what Apple should have done. And I still think they might do because really OpenClaw isn’t all that helpful unless you give it access to all your shops, but if you give it access to all your shop, then that’s a giant security risk that you got to manage and all this stuff. It’s like it’s got to access this. Where is the nexus for where all that is securely enabled for me? It’s in my pocket and my phone. Apple is perfectly positioned to do that. And instead, what they have, which is always just complaining about on Twitter, is Apple Series is even suggesting, “Oh, do you want to add this event to your calendar?” And it’s a duplicate of an event that’s already in the calendar. 

This is how stupid Siri is. They can’t even look at the calendar, be like, “Oh, maybe I shouldn’t suggest an event that’s already in the calendar.” So yes, fundamentally, I think it’s … Do I think it’s as big a deal as people think? No. There’s an awful lot of noise at the moment and there’s an awful lot of people I started looking at it’s like, “What are you actually doing?” It’s like, “Oh, it’s looking at my email and my calendar and it’s arranging my to- do list.” And I’m like, “Great, but your Tooly-Do list is still kind of the same.” I don’t know how to say difference there. But I do think, this is what I was talking about too, with how CloudCode that harness around has been the biggest change. And I feel like that’s actually where Apple could come in and even without necessarily owning a fantastic foundation model could come in and just be the interface by which normal people interact with AI day in and day out. 

For sure, I think that’s true.  

Rob Walling:
Tracy makes, what’s your experience or understanding of OpenCloud?  

Tracy Osborn:
I’m scared of it because of the security issues and I am someone who … I’m just like, “Oh no.” Even small things that are not even AI, but I started using Superhuman and it was all like, “Oh, we’ll start managing your email for you. ” And I immediately missed a bunch of stuff because the system itself thought was like, “Oh, you don’t need these things.” And I was like, “Wait, I actually do need those things.” And so I imagine that being put into across my entire computer and the little AI agent just being like, “Don’t need that, don’t need that, don’t need that. ” And I’m a type A person who likes to have everything in the right place, for me to categorize myself and fix myself and update myself and all that. So that’s where I am right now. However, I can definitely see myself when a consumer solution comes out that allows me to tie in all my systems in a way where I know that the chances of something going wrong is vastly decreased because right now it looks like the chances are going wrong are nearly guaranteed because it’s early, but it’s going to get to that point. 

And when it gets to that point, then that’s the future. That’s what’s going to happen. And I know that that’s something I’ll switch over to. So I’m not there yet, but I’m looking at it and seeing like, okay, I can see where this is going and I’m excited, but I’m going to wait until I’m not scared of it.  

Einar Vollset:
But this is what I mean. This is why Apple is the perfect interface because if Apple came along and said, “Hey, here’s the thing, honestly, privacy is secured, blah, blah, blah, blah, blah.” You can just chat to a super capable AI that knows all the context for everything you’ve ever gotten, like who reminds you, “Hey, it’s birthday on next Tuesday or whatever.” That’s helpful.  

Tracy Osborn:
You’re talking to the person who has Siri turned off on her phone, so  

Einar Vollset:
It’s brain dead. It kept poisoning my calendar with duplicates.  

Rob Walling:
Yep. Yeah, it’s bananas. OpenClaw here to stay or the next BeReal. Remember BeReal, that social network that was here for like a week? I have no  

Tracy Osborn:
Memory of that. That’s so  

Rob Walling:
Funny. How about this one? Is OpenClaw the best idea since Greedo shooting first? This what? No, 

Just me. All right, Trace. I got chance. Have you ever seen Star Wars? That’s like an old Kevin Smith joke. All right. So those are … Oh, for me on Open Cloud, yeah, I’m in the same boat as you, Tracy. I’m not at all. I would buy the Mac Mini and I would install it. The technical issue doesn’t concern me. It’s the security of it at this point. And I don’t want to spend all that time and then have Apple or someone else do one. It’s like, “Oh, hey, it’s one click and it’ll do it. ” I’m kind of waiting for that. I don’t need it so badly that I’m so desperate that I’m going out buying Mac Minis. I’m  

Einar Vollset:
Pretty sure, I don’t know if you guys are football fans because I’m not. And so the Super Bowl was just happened this weekend. And so you know the big ad was ai.com. Oh, I know.  

Tracy Osborn:
Yeah. I heard something about this, but to do.  

Einar Vollset:
Yeah. So that’s pretty much, I’m sure that’s what they’re doing. It’s ai.com and it’s like lives in your iMessage kind of thing. And there’s a bunch of startups now that are sort of like, spin this up and let it host and do all the thing. And people are still, there’s still ways that you’re like, you can spin it up as like a cloud front thing, Majiger worker, Cloudflare worker or whatever. And that’s fine, but that doesn’t actually give you what you give why it’s the most capable, which is it has access to your stuff as if it’s sitting in front of a log desk. It’s what I’m saying, you’re almost never going to get 100% security because it’s not helpful unless it sees what I see. And if it doesn’t see what I see, then it’s just like, “Yeah, okay, can you do some web research for me? 

” I’m like, “Yeah, so can ChatGPT. Who cares? That’s not interesting.”  

Tracy Osborn:
I have to be assured that it’s thinking about things the way I want to think about it. That’s the thing that like, okay, cool. I can have it look at my insecure stuff as long as I know that it’s going to be doing what I want to be doing and not going off and it’s doing its own thing.  

Einar Vollset:
Sometimes I’m really glad that it’s not doing what I’m thinking because it’s better.  

Rob Walling:
So you want it to be better. Fair enough. Folks want to keep up with Tracy makes Osborne. You are not on Twitter, but you did bet Aina Volls did a thousand dollars and a sushi dinner for me that I just kept on this sushi dinner That Blue Sky would be at 100 million signups, not active users. At some point, it’s like a year and a half, two years from now.  

Einar Vollset:
You might as well just pay me now.  

Rob Walling:
What is your handle on Blue Sky? Is it Tracyosborne.com?  

Tracy Osborn:
It is Tracyosborne.com, yes. I need to get TracyMakes.com at this point. What am I doing? What’s wrong with me?  

Rob Walling:
Yeah. Oh, I’m about to go out and register that. And you these days are working on SaaS Institute. We working on several things, but SaaS Institute, which is our premium coaching for seven and eight figure ARR SaaS founders.  

Tracy Osborn:
It’s our new program too, so we can work with more founders that are at one million up and want to have a tiny Seedlike experience, but aren’t the right fit for the accelerator. Ainar, share it more. I see you talk about the accelerator a lot. Talk about SaaS Institute.  

Einar Vollset:
Okay.  

Rob Walling:
All right, publicly calling me out. Calling you up, Polly. You  

Einar Vollset:
Pitched the guide and I’ll pitch the institute.  

Rob Walling:
Okay. Yes, I already have. Einar things. I don’t like it when mommy and daddy fight. Ainar Vollset on Twitter, E-I-N-A-R-V-O-L-L-S-E-T, tweeting about the giants and AI and Norwegian fiscal policy, it seems.  

Einar Vollset:
Fiscal policy and like Epstein stuff now. Apparently the Epstein thing was a big thing for Norway.  

Rob Walling:
I know.  

Einar Vollset:
No, seriously. Turns out that the Crown Princess was like buddy buddy with Epstein and like a bunch of folks are getting arrested. It’s all good. It’s good times in the Norwegian media. I’ll tell you that much.  

Rob Walling:
That’s crazy. Ainar is originally from Norway for those who don’t know. And you’re also working, as we said, discretioncapital.com. And if people want that guide, they can just hit slash guide, get it for free on the internet. Tracy and Einar, thanks again so much for joining me for this week’s episode.  

Einar Vollset:
Happy to be here. Thanks for having me.  

Rob Walling:
Thanks again to Tracy and Einar for joining me this week and thanks to you for subscribing and rating this show. If you are listening in Spotify or Apple podcasts or whatever app you’re listening, I would appreciate a five star rating or a thumbs up or whatever mechanism your podcaster of choice is using. It helps new folks find the show. Thanks for listening this week and every week. This is Rob Walling signing off for episode 823.