SaaS was dead last week, but now it's back.

Wrappers are dead. LLMs are dead. SaaS is dead. The application layer is dead.

It's moving pretty fast still in AI, and we thought it was time to take stock of what we know right now.

• (00:00) - Introduction
• (01:03) - OpenAI's Superbowl Ad Breakdown
• (02:29) - The Role of AI in Modern Life
• (03:16) - Teasing Future Content
• (03:50) - The Fast-Paced World of AI
• (05:35) - Understanding GPT Wrappers
• (07:40) - Hype vs. Reality in AI
• (10:04) - The Evolution of Solutions
• (14:22) - Consumer Perception of AI
• (16:25) - AI's Role in Problem Solving
• (17:47) - The Commoditization of AI
• (18:50) - Cloud Providers and AI Models
• (20:39) - The Future of AI Investment
• (22:31) - Challenges in AI Implementation
• (28:51) - Skepticism and Realities of AI

This episode is brought to you by Everstage - the highest rated Sales Commissions Platform on G2, Gartner Peer Insights, Trustradius with over 2,000+ customer reviews. Some of their customers include leading brands like Diligent, Wiley, Trimble, Postman, Chargebee etc.,

You can go to https://www.everstage.com/revenue-formula to check out Everstage and mention Revenue Formula to unlock a personalized Sales Compensation Strategy Session with Everstage’s RevOps experts—crafted for enterprise teams to maximize performance.

Never miss a new episode, join our newsletter on revenueformula.substack.com

Creators and Guests

Host

Mikkel Plaehn

Marketing leader & b2b saas nerd

Host

Toni Hohlbein

2x exited CRO | 1x Founder | Podcast Host

What is The Revenue Formula?

This podcast is about scaling tech startups.

Hosted by Toni Hohlbein & Raul Porojan, together they look at the full funnel.

With a combined 20 years of experience in B2B SaaS and 3 exits, they discuss growing pains, challenges and opportunities they’ve faced. Whether you're working in RevOps, sales, operations, finance or marketing - if you care about revenue, you'll care about this podcast.

If there’s one thing they hate, it’s talk. We know, it’s a bit of an oxymoron. But execution and focus is the key - that’s why each episode is designed to give 1-2 very concrete takeaways.

[00:00:00] Toni: Hey everyone. This is Toni Hohlbein. You are listening to the revenue formula with Mikkel and Toni. In today's episode, we talk about the recent developments in AI, what we really know about it and why everyone should just calm a bit down.
[00:00:16] Before we jump into the show, today's is to you by EverStage, the top sales commissions platform on G2, Gartner, Peer Insights,
[00:00:26] and Trust Radius with more 2000 reviews from customers like Diligent, Wiley,
[00:00:32] Trimble, and more.
[00:00:34] Visit everstage. com and mention Revenue Formula unlock a personalized sales compensation strategy session with one of EverStage's RevOps experts.
[00:00:45] And now, enjoy the show,
[00:00:47] Mikkel: so last night uh, it was Superbowl Superbowl. I mean, like, I guess many, most Europeans, no, no, no. And I guess like most Europeans, you're like, oh, that was yesterday. Oh, cool. I didn't know.
[00:01:00] It's like, I didn't know it was in February. But that was yesterday.
[00:01:03] And one of the companies that spent some money. On the infamous ads there, which by the way, it costs around 8 million us depending on how you place them was open AI, open AI. They placed an app there and you would assume that they created with AI, right?
[00:01:20] You would assume, did they?
[00:01:23] Toni: Of course they didn't.
[00:01:25] Mikkel: Of course they didn't, they used it for prototyping, but they were like, ah, this very important stuff of actually getting that 8 million ad to succeed. We'll leave that
[00:01:36] with the
[00:01:36] Toni: What was the ad about? I didn't see it.
[00:01:38] Mikkel: No, I just watched it actually before we started recording. So the ad is 60 seconds and it's basically 45 seconds.
[00:01:45] Of weirdly animated things showing human invention which is also hilarious, by the way, showing human inventions, like the light bulb, the airplane finding out the discovery of the DNA, getting a man on moon, all that stuff. Like the, the classic, classic
[00:02:00] stuff.
[00:02:01] And the, yeah, yeah, no post. And then all of a sudden it's like, turns into summarize this for me and blah, blah, blah, blah, blah, that for me.
[00:02:08] And then chat GPT, right? I was just like,
[00:02:11] Toni: It's, it's so funny, right? Because, the invention of the lightbulb, the invention of fire, the invention of electricity, the internet you know, DNA, and then
[00:02:21] Summarize this email for me. It's like, oh wow. Yeah. No, you deserve to be in that chain. A hundred percent.
[00:02:29] Mikkel: I think they're playing with the things that have made an impact on our life. And could you imagine not having AI?
[00:02:37] Toni: Could you imagine still needing to read actual emails instead of the just summarize the shit for me button?
[00:02:42] Mikkel: Or like actually hiring SDRs to make calls versus just having an AI SDR.
[00:02:48] Honestly, Jesus, Jesus, Jesus.
[00:02:51] Toni: it's right up there with imagining not having the internet or not having a list.
[00:02:57] Mikkel: Oh man. Oh man. Oh, so much is happening. You also kind of called it and said, you know what? I have an intro today. I've done my homework and it's, is it about kids today or no?
[00:03:07] Toni: So, I mean, I think this was the perfect intro, actually. Maybe you can save this for the next episode. It was just like a, like a, you know, mini story I wanted to say,
[00:03:15] but, but you know what?
[00:03:16] This is, this is why you need to hit the subscribe button. If you really want to hear this story, you need to listen to it next time.
[00:03:22] Sorry.
[00:03:22] Mikkel: was just going to say, it's like those good old shows that really understands how to hook people. This was like, Oh, cliffhanger. Dang it. Now I need, now I need the next one.
[00:03:32] Toni: and I think we talked to enough people to know that actually they shut off after the first five minutes. Once the intro is over, it's like, Oh, okay. I don't do, you know, AM, AE, whatever. I don't
[00:03:42] Mikkel: So thanks for listening. Uh, it's been a pleasure. talk to you next time.
[00:03:45] Toni: bye
[00:03:48] Mikkel: No, so the energy is, is kind of returning.
[00:03:50] I can tell. one of the things we wanted to talk a bit about today is Sass. Is that, or is it, obviously. A lot of things is happening in the AI space. It's moving incredibly fast. And one of the things you wrote about the other day was, funny enough, open AI, presenting in Japan.
[00:04:07] What was it they were presenting there?
[00:04:10] Toni: I think they were actually presenting to like a business crowd. I think they were trying to sell open enterprise. And I think for many people, it's really hard to imagine, well, what can you do with this thing besides of yes, summarizing an email and telling me what, you know, I should be doing in my travel destination and stuff.
[00:04:27] And what they did there, and I think it's either internal tooling of open AI. Or it's really just some developers whipped this up over like the weekend. They basically did a demo of I think precisely an inbound SDR workflow. So someone fills in a form, then something gets researched and qualified, then there's an outreach email going out, you know, and so forth.
[00:04:47] Right. And and I think what was stunning there. Well, the, the, the, again, the purpose was, Hey, listen, this is what he can do with AI kind of, this is just one idea and there are hundreds of others. But what I felt was stunning there was like, Oh, geez, man, this, you know, this used to be, you know, someone building a slide deck for two weeks, trying to, you know, stitch a demo together, or maybe a figma in more recent years.
[00:05:09] And now it's, this looked like legit software. This looked like, you know, this was not kind of a mock up. This was the real thing. And, and that they could just pull us off for like a like an event like this, pretty stunning.
[00:05:22] Mikkel: Yeah, no, definitely. I think it's also just hilarious because it, it also made me reflect on one other thing, which is this whole people talking about GPT wrappers, that that's not going to last. It's not going to work out.
[00:05:35] Do we want to maybe just, because this is a GPT wrapper, by the way, it is a packaging of chat TPT into something else.
[00:05:41] Do we just want to, because I said, like before we even recorded, like, what is a GPT wrapper even let's maybe just reposition that a little
[00:05:49] Toni: Yeah. So what is a wrapper? Basically it's the idea of a very thin and, and insinuating with it also very low value. Interface on top of the actual value, which is the AI and the model and so forth. Right. And, and in the early days someone like copy AI was you know, considered the GPT wrapper with just a very simple interface.
[00:06:12] And then doing tons of AI calls in the back to you know, to do the work that they need to do. Right. And I want to say, this is now two, three weeks old of time of recording. There was a wave of people basically saying. All of those rappers are dead. I forgot why that was actually, I already forgot the context,
[00:06:30] Mikkel: Yeah.
[00:06:30] Toni: I, oh, I think it was, I think it was open end came out with scheduled. So
[00:06:36] Mikkel: Oh, the task. Yeah. Yeah.
[00:06:38] Toni: I mean, literally this feels like a
[00:06:40] Trump year new
[00:06:41] Mikkel: And operator as
[00:06:42] Toni: this is only two weeks ago. And then this operator thing on top, right. And basically everyone was like, oh no, this is, you know, now forget, forget all the interface stuff you build.
[00:06:50] You, you know, you can trash it now. Then, you know, later that week the deep seek stuff came out and people were like, Oh, wait a minute. Oh, wait a minute. Wait, wait, wait. I think, I think those API calls, they're going to get a lot cheaper. So I think the actual value. It's in the thin wrapping that we used to poop on just yesterday.
[00:07:10] So let's just change course. Let's just go 180 in the other direction. And I think around this time, I think a little bit previous to that Satya Nantella from Microsoft basically had an interview and he was talking about, Hey, agents and agentic workflow is going to replace lots of the application stuff.
[00:07:25] And there was a whole maelstrom around. I mean, we did a show with Jonathan Moss on you know, AI is eating software. Et cetera, et cetera, et cetera, et cetera, right? So I think lots of, you know, killing, dead, eating you know, stuff out there.
[00:07:40] And and I think this is what we wanted to talk about today, to clean this up a little bit, right?
[00:07:44] Because there's a lot of hysteria, there's a lot of hype. And I think some of the hype is justified. But then there's also a lot of hysteria around this. You and I sometimes stoking that hysteria
[00:07:54] Mikkel: just gonna
[00:07:55] Toni: not on purpose, but I think because we are hysterical ourselves, right? Kind of, we're just, we're just a reflection of what's going on here.
[00:08:02] But we just wanted to take a step back and like, Hey, you know, what is it we actually know about this whole thing? What is it we don't know and what should we take away from it?
[00:08:09] Mikkel: Yeah. I think it's also kind of our obligation, right? We, you don't definitely want to be in our hysteria when we're not recording. It's just too crazy for everyone to even, to even publish, to be honest. Right. But I think it's fair to say things are moving incredibly fast, right? Things are moving incredibly fast.
[00:08:23] And I think also one of the telltale signs of how crazy things are getting is. When you search for things on YouTube and you get a random YouTuber pro dude, who is like an affiliate, person, and it feels like everyone is pushing the same N8N workflow automation type things that you can build, pay me X thousand dollars for a course, or I think there's actually a lot of fraud as well happening in this whole kind of bubble.
[00:08:49] And I think it's just time for us to say, calm down. Let's, you know, let's calm down, take a step back and just assess. What are the things we actually know to be true right now? Because things are moving incredibly fast and One of the things you and I just, very quickly arrive at is like, are there really new problems out there to be solved, how they're truly that many new problems to be solved.
[00:09:13] Right. And, it's almost like I was thinking about the, How many companies actually die when new technology like this happen? And usually it's like one or two big ones. You're going to remember right now we've discussed check. so the online tutoring company that's publicly
[00:09:27] Toni: That no one knows by the
[00:09:28] way.
[00:09:28] Mikkel: that no one knows, but they were like a billion dollar company. now they're, well, actually they were maybe 50 billion or something. Now they're just 1 billion, you know? Too,
[00:09:37] Toni: still a billion dollar company, but
[00:09:39] Mikkel: Yeah, but maybe it's going to end up being zero. And it's like, so last time this stuff happened, uh, with, uh, smartphones, it was Nokia, that kind of, they still kind of exist.
[00:09:47] I think they were even acquired by
[00:09:48] Toni: before that was Kodak and kind of, there's always those, always those stories.
[00:09:52] Mikkel: there's always those stories, but it, so is it like an entire sector? Does that ever die? And I'm not, I'm just not so sure because those problems they've, to have by and large already been solved, right?
[00:10:04] Toni: So I think, you know, when we were kind of, shit chatting about this previously, I think you had a good example there with like, so, you know, the horse carriage to the car, to the electric car, it's like. I mean, the problem was always the same, like get me from A to B without me needing to move my legs, like, please do that.
[00:10:22] Right. And when we had the horses and the horse carriage, I mean, car carriage, it's the same, by the way, you know, it comes from one another.
[00:10:29] Mikkel: Yeah.
[00:10:30] Toni: you were, you were sitting there and going from A to B now you're in the car, you need to do the steering yourself. And, you know, soon there's maybe, you know, steering not, and the EV part is, is there's really convenient, but ultimately.
[00:10:42] Ultimately the problem stayed the same, the solution evolved a little bit, but the solution still at its core was getting you from A to B,
[00:10:51] right? And, and and really, so frankly, as, as humans, you kind of care little about how this is sometimes being solved. If we had a, if we had an awesome solution with a horse in front.
[00:11:04] Would, would you care? I mean, , it's like yes.
[00:11:07] Mikkel: if it's faster than your car.
[00:11:08] Toni: work Exactly, yes. If that was still to work and kind of would be better and cheaper and so forth, then you would still do that. Right. But the the, and then I think the same thing. You know, when we take a step from all the, all the LinkedIn and Twitter and, and, and YouTube stuff that's coming at us, it's like, what?
[00:11:25] You know, what problems? You know, what new problems do we have? And the answer is very, very few new problems, very few new problems. So really it's about like maybe improving some, some solutions, like in a new way with AI, but is that, is that solution going to be fundamentally different in all cases? And I, and I think the answer is, I don't know, probably no, no, does, does everything now need to be a chat interface?
[00:11:54] Am I going to prompt my way through, I don't know, setting up a Marketo workflow, like, I'm not sure. I'm not sure that this is the most efficient way to do some of these things. Right. And I think it's really healthy if I want to just like, take a step back and realize what really has changed and most importantly, what hasn't changed.
[00:12:13] And what you will find is that 99 percent of things probably haven't actually
[00:12:17] changed.
[00:12:18] Mikkel: I think it's also like with the example of the car, right? How often do you think people go out and they just pop the hood and look at the engine and go like, Oh my God, I totally made the right decision. Getting a car because it has this engine that looks exactly, it's like, you don't buy it because of the technology.
[00:12:39] I'm sorry. You buy it because of a million, like you buy it maybe because of how the engine is performing, by the way, like the mileage and blah, blah, blah, blah, blah. But you don't care that it's a, it's an electric or a petrol or a million other things, unless you have like preference for environmental friendly stuff, et cetera.
[00:12:53] So just to say the solutions are there. And I think the point, I think also to be made here is like, we already have a CRM. We have a CRM, it's implemented, it's working, you can hate it for all its flaws and imperfections, but is an AI solution really going to come in and replace that? I'm not so sure.
[00:13:13] I think on the flip side, there might be some problems that we've lived with. That could potentially get solved a little bit better now, or could actually get solved. and I think specifically stuff like the repetitive task and the manual task, that will be some, that will get replaced. I think we've seen the company called, Harvey AI that does legal research.
[00:13:34] Super helpful. You will still need a human in there, but super helpful. It's probably replacing quite a few people, in doing the research. The question is, were there people doing that amount of research before? Question mark,
[00:13:44] Toni: Yeah, they were.
[00:13:45] Mikkel: Yeah,
[00:13:46] Toni: I think this is one of those examples. It's
[00:13:47] like, oh no, no, this makes total sense. actually, it makes total
[00:13:50] Mikkel: yeah,
[00:13:50] that lasted, but, and I think that there are going to be some of those, let's be fair, we know there will be some of those tasks out there, I think that's why you're also seeing a race to try and replace, SDRs because it's fairly standardized work.
[00:14:03] To be honest, and, for whatever reason, they've just not been able to nail it yet. Probably in a couple of years, we'll sit on this show and talk about, Hey, which one is actually performing best? Is this one or that one? Because all of them work, but you know, so I think, let's just conclude. There are only so many problems out there.
[00:14:17] There are only so many opportunities. calm down, calm down. Right. Yeah.
[00:14:22] Toni: thing that's kind of starting to, to happen is people don't really Care about that. AI is part of your product. And I think we can debate whether or not this is from a, Hey, I just need my problem solved. I don't care how it's being solved. Perspective, or if it comes from.
[00:14:41] Just the expectation that AI is everywhere, like, you know, it's, is it as a hygiene thing or is it just
[00:14:46] Mikkel: It's like, oh, you don't
[00:14:47] Toni: thing, but, but some research now suggests that there's no difference between the AI level label being on your product versus, versus it being there from the perception of, of the buyer.
[00:14:57] Right. So very, very interesting. Also new fact that we starting to know and see that that doesn't potentially play a role. So kind of also around this, maybe calm, calm down.
[00:15:07] Mikkel: Yeah, I think it's also you, it almost makes me, makes me think about, David Ogilvy So when he would do an ad, for a airline company, he
[00:15:16] Toni: one knows who that person is,
[00:15:17] Mikkel: David Ogilvy Haven't they watched Mad Men?
[00:15:21] Toni: Okay, yeah, yeah, they
[00:15:22] Mikkel: Okay. So if you don't know, if you don't know, let's let Don Draper Yeah. If you don't know David Ogilvy it's Don Draper from Mad Men.
[00:15:29] This ad guru, he's written tons of amazing books. Go check them out. They're old, but they've, they still, they still work. One of the things he would do. he specifically did with an account. It was an airline. He went out there and got a walkthrough of, well, how does this thing actually, how does it work?
[00:15:43] How does the airplane work? What are some of the features of this airplane? And, they walked him through things like, oh, it has a seatbelt, you know, convenient. It has, these, insane filters on the air, to that, you know, basically removes any kind of bacteria, viruses, like insane. So all those features, he could go and say, well. It's fair to say it's a very safe airplane because of we have this filter we have a seat belt We have a few other pieces, right? He's not really talking heavily about the technology In those pieces and an airplane is, you know, at that point in time, it was actually a big achievement to fly that many people, from A to B.
[00:16:21] Right. And it is a very complicated solution, quite frankly.
[00:16:25] and I think it just goes to show AI is obviously part of deriving an outcome, but what people care about is can do the work I want it to do and give me the outcome I need. Right. So that's why the label people, I don't think to that degree care about.
[00:16:39] And it doesn't matter whether it's because they just assume it's there. Or not don't care. they care about their problem
[00:16:45] Toni: I think it's, I think it's a little bit like you know, for the people that are old enough in the late nineties, early twenties with Intel inside, probably in the early,
[00:16:53] probably in the two
[00:16:54] thousands. Right. And basically the, the reason why everyone was up in arms was like Intel and, you know, their stock wrought because of it was.
[00:17:02] They managed to be a boring technology provider that consumers suddenly cared about. Oh, it's the Intel inside because otherwise I think I can't, you know, and and I think what's happening now with AI is kind of the reverse, actually, which we usually see. That's the usual trend, right? You have a new technological breakthrough.
[00:17:20] Everyone's like, Oh, wow, this is absolutely great. You see a couple of companies rushing to become the. The shovels and pigs pickaxe survivors suppliers for organizations to use technology. And, you know, now we have OpenAir, we have Anthropic, we have you know, Mistral, maybe, I don't know, Le Chat.
[00:17:37] And then, you know, we have, you know, Facebook, we have DeepSea, we have a couple of other folks that are kind of doing that stuff.
[00:17:42] And I think this is, this is the other piece what's going to come out of this is like, number one.
[00:17:47] It's being commoditized. So what does commoditized actually mean? Kind of, I think it's a fancy word for it's going to get cheaper friends. Right. So it's really about now race to the bottom, you know, people don't care which, you know, model specifically, but all those models kind of, kind of the same.
[00:18:04] And we're, I think we're starting to lose overview of like. 0. 3, 0. 1, 0. 3. 1 mini and an 0. 1 mini is beating 0. 3 and it's like, you have no, you have no clue anymore. So kind of that, that stuff is just like, okay, now this has entered white noise space. You are starting to be like, oh wait, DeepSeek is better than OpenAI, but DeepSeek at 5 million, OpenAI is like hundreds of billions or whatever.
[00:18:28] Like, you know, how does, you know, all of your ways of trying to make sense of this stuff from a non technical perspective, they're getting diluted and washed away, so what's going to happen very soon is like, people are like, I act, you know, I don't care. Is it a model? Does it do the task? Yeah. Okay, cool.
[00:18:43] Let's buy that. And I think what's you know, what it's most, you know, comparable here is.
[00:18:50] Think about all the cloud providers, you know, it started with Amazon web services, right? And nobody even knows that word anymore. It's all AWS by now. Then Microsoft was, Oh, you know, dang, we missed, we missed the train on that.
[00:19:04] Let's jump on this with Azure. Google did the same thing. And now at the end of the day, it's now AWS, and that's it. And everyone, everything that, you know, a. Commercial person really knows about and thinks about. It's like, is it one of those three? Because all the other stuff I don't know. Right. And the same thing, the same thing is happening with those models right now.
[00:19:25] And it's going to be all about the same thing with AWS and Azure and GCP is like, is it cheaper? Is it faster? You know, and, and then let's, let's move on. Right. And I think that's. That's frankly where this stuff belongs. I think from a very arrogant you know, commercial perspective, this stuff belongs in the background, push it and shove it down.
[00:19:47] And, you know, this, you know, I think open, I will take a hit number one or open AI will like, okay, you know what? All this modeling stuff is kind of commoditized. We now need to jump way more to the application layer, which is kind of almost another thing we could be discussing here, but you know, take away AI costs are going to go down, down, down, down, down.
[00:20:05] Mikkel: And I just want to double back to say, I'm so thankful that we have a European solution and that they chose to call it Le Chat. because I could not have lived with this being a German invention. And then it would have been called Das Chat.
[00:20:18] And it's just like, it's not going to work. It's not going to work. No, but I think you're right. So this thing will get commoditized. It will get cheaper. And I think what you will see is there's going to be some companies who are smart, who know that something right now is very costly to do with AI, but they also know that the cost is going to go down. so there is an opportunity there. if you can find it, write me, maybe I want to go build something there.
[00:20:39] The, the other thing, just to keep us moving along here, my friend is I saw, and this is like by now, probably two years ago, David Sachs on stage with Jason
[00:20:49] Lemkin at SaaStr I think
[00:20:51] in London, they were talking about AI from obviously a VC perspective.
[00:20:55] David Sachs is a VC. I don't, do you know what fund is it he's running?
[00:20:58] Toni: I forgot, but also he's, he's like, he's not cool anymore, at
[00:21:03] Mikkel: No, exactly. We'll cut it out just so I'll AI edit it to give the answer. So we sound smart, but they were basically talking about AI and, David Sachs made an interesting point, which was like, Hey, for now, there's no more SAS really to be built. But what AI is doing is that it's expanding the pool of opportunity, meaning there's more investment opportunities happening, right?
[00:21:26] And specifically also what we've seen to that end is that the market is expanding all of a sudden. There is AI budget out there and there has been for quite some time. people explicitly want to experiment with this technology and see whether they can reduce their cost base, improve efficiency, grow revenue or Improve customer retention, sorry, customer retention, employee happiness, or a million other things.
[00:21:48] Right. So there's a lot of experimentation happening out there and also a lot of learning and an interesting, I think, result of that as well is some companies are trying to build their own solution versus buy as well. so we're in this kind of weird scenario. And I think what is super interesting right now is a lot of companies, they're going to be able to grow fast, sell their AI technology into a lot of companies.
[00:22:13] But I just wonder what's going to happen in a couple of years when the companies are, okay, let's actually now go and evaluate all this AI budget and AI spend and AI tooling. Is it, is it actually working for us? I think that is going to be a very interesting point in time. that's, that's eventually going to come.
[00:22:30] Eventually it's going to happen.
[00:22:31] Toni: let me unpack kind of those two things you mentioned, you know, one is around the AI opportunity from an investor perspective, but off of a market perspective, then the other one is almost kind of connected to that kind of the budget around this. Right. I think on the AI opportunity side, I think a lot of people's imagination has been, you know, triggered.
[00:22:48] I don't know yet if it really has led to, you know, people being let, let go, by the way. You know, I still haven't really seen, um
[00:22:59] Uh, but no, yeah, but I haven't, I haven't seen at least anything that would would justify some of the spending. Let's just be really honest about that. Haven't, I haven't seen this.
[00:23:08] And then kind of our eco chamber is very much around the AI SDR and stuff. And although it doesn't work, so, but, but there might be other things where it does work out, but I haven't really seen companies scale down and like the Klarna guys kind of keep talking about this, but I would, I would tread carefully.
[00:23:24] I would try it carefully. I think there's also just a, Hey guys, we're kind of going IPO in like six to nine months. So, you know, let, let's add my profile and kind of, you know, so, so I kind of would be careful about this. Right. And then on the budgeting side, I think this is actually way more interesting because while a lot of people's imagination has been triggered and like, oh, wow, we could kind of say so much costs.
[00:23:45] And we all believe that that's probably true. You know, at some point in a probably in a year, maybe this year, maybe later this year you know, the CFOs will come back and be like guys, we, we just, you know, we keep running the numbers. It keeps not going in the right direction. I don't, I don't think this is working out here. Right. Someone will come and like, you know, call BS and, and add a reality check to this thing and was then, you know, the, the, the question then is going to be like, well, will that lead to all of the experimental budget that's been assigned to AI to be cut back?
[00:24:23] And it's like, yes, it will a thousand percent. It will. Right. Unless you really have solved the problem that led to, you know, people being let go. It's, it's very hard to imagine where, where this reality check isn't going to come anytime soon.
[00:24:36] Mikkel: This is like the, the gym membership you keep clinging to thinking you will definitely use it. You will. Tomorrow is the day.
[00:24:45] Toni: Yeah.
[00:24:46] Mikkel: No, no, no, no, no, no. I think the last piece, which by the way, probably also is a bit common. I think in a time like this, I think there is also this fallacy of. Hey, won't chat GPT just build that like, won't we just, or could we just not build it ourself, by the way, we'll also be honestly a thing on people's mind.
[00:25:10] And I think, let's not underestimate how difficult it can be to build and get, let's just say an AI SDR to work since they had that example. Right, right now it's, when I saw it, I was like. Okay. So it does something on email back and forth with the prospect on scheduling a meeting. Have you heard of Calendly and like the confirmation pages?
[00:25:35] So, you know, I don't know all the details here, but I think it's just suffice to say it's things are a bit more complicated on the surface. Once you want to build the, a new piece for a user and honestly, isn't there so much opportunity left in the current interface and solution they've built, I still just want to.
[00:25:54] See at some point them disrupting the search space, like completely there's so much, so much opportunity there to have, have ads. And I think, I, I heard from, Google's earnings call. They're like, yeah, with the experimentation they're doing with AI, it's actually better for their ads. and, also better for the users at the same time.
[00:26:14] So they're like trying to rush to roll this out now. Right. So I think there might be something happening there. And I also, it made me think of, was it Steve Jobs who said to the, it's Aaron Levy who, who founded Dropbox, like, we're going to steal your feature. And they kind of did, but you know what?
[00:26:30] Dropbox is also doing okay right now. They're, they're still an established business. I don't know. Like, I'm assuming so at least. Right. So I think I would just calm a bit down with the, with the
[00:26:41] fear of. ChatGPT and coming
[00:26:43] in and, you know, stealing Salesforce's CRM. Like, I'm not sure this is like the natural next step for them.
[00:26:50] Toni: So I think though, this is, this is in the realm of what we don't know yet. Right. And we're just, and I've been thinking about this a lot. Obviously kind of also building the company here that we're building. What is, what is an open AIs path next?
[00:27:05] Mikkel: Yeah.
[00:27:05] Toni: And we've already seen them, you know, I think wiping out as a strong word, but challenge some of the investments that were made in the, in the last two or three years, where someone was, yes, creating an, you know, an AI wrapper and helping on the workflow level.
[00:27:21] Right. And because it's fairly. You know, the wrapping is fairly thin, basically, it doesn't take that long for open AI to go there and kind of replace that thin layer of workflow that's been put on top. And, and therefore then challenging that company, that investment, that use case that was kind of built around, right.
[00:27:37] And it's very, it's not cool. I can tell you that. I don't think it's a cool situation for any startup to be in. To try and compete against an open AI, they're like insanely funded. There's still, you know, yes, there's seven or eight or nine years old now, but really they're two years old. Like that's, that's how you need to think about them.
[00:27:57] Right. And you know, it's, it's not like it's a, it's an old dusty incumbent you're going up against, no, it's like, this is a. You know, you, you just don't want to do that. That's what I think. You just don't want to do that. And if you're in the, and by the way, your investors also don't want to do that.
[00:28:11] So if you're in the path of them, that's kind of shitty. So really trying to figure out where they're going to go next or is, is, is super valuable and interesting, but we just won't know, right. If there's any prediction I have is if it's true that models are getting further and further commoditized.
[00:28:27] Then for them to maintain their, what 150 billion valuation they kind of need to then, you know, add another trick to the show, because otherwise it's not going to work out. And that, that might be the application layer. So kind of how far they're going to go into this, I don't know, but it's certainly something to, to be mindful of.
[00:28:45] I think if there's another thing that I'm kind of thinking about a lot, what is frankly unknown as like.
[00:28:51] How far will this AI thing actually go, right? We're starting to now live it for two or three years. We have, you know, for business purpose, for human purposes, I think it's already like extremely intelligent.
[00:29:02] You know, it's, it's a little bit unclear to me what artificial general intelligence actually means and how this is going to be useful and how's it going to be different and. Are they, are they going to become the Terminator and the Skynet that we all think? Or is it, is it just going to be a really cool tool in the end?
[00:29:18] Right? So again when people came out with a train, like the, you know, the actual train on rails, 150 years or something ago they were thinking their limbs are going to fall off if they go faster than 20 kilometers an hour right? It didn't happen. So it's like, there's a little bit, there's a little bit of this in me where I'm like, it might still go further.
[00:29:37] Maybe we're actually reading reaching a ceiling already now. But even if it does go further, what does that then actually mean? Is everything going to automatically integrate? Is everything going to be done by just an AI clicking button? I don't know, but there's, for me, there's starting to be a healthy skepticism of like, is it really going to get so much different?
[00:29:55] Have we kind of reached the Understanding what this thing is. And now it's all about incremental performance improvements and incremental costs, you know, reductions.
[00:30:03] Right.
[00:30:04] And then I think the last one we've touched on already is like, well, these experimental software AI budgets, how long are they going to last?
[00:30:13] Right. And, and I've now heard it from very kind of interesting source. Basically 11 access net retention rate and some months is 60%. So in what does that, what does that mean? It basically means, and maybe edit this out and say, AI SDR retention is 60% you know, in a couple of months. So what does that mean?
[00:30:33] It means that you a lot of those experimental budgets have, you know, are starting to pull back. People are starting to get to the point how this actually doesn't work for us. So, so we actually need to cut this thing again and then it's unclear. Does the CFO feel like, Oh, let's, let's roll the dice again
[00:30:48] on something else.
[00:30:49] Or,
[00:30:49] is it like, Oh no, actually kind of this hype. Let's just dial it back a little bit.
[00:30:53] Mikkel: Yeah, no, I agree. And I think it's also probably, so probably everyone is seizing the opportunity right now of building AI to be frank, because it matters for the VC landscape. If they want to raise another round, it matters for your ability to sell, to tap into those budgets. But I think where the struggle is real is on finding experiments that actually work, right? and, and it, and it must also be difficult to be on the recipient end saying, Hey, I'm going to give you some budget. You go figure out some AI stuff that's going to help us. It's like, well, okay. That's I think it's going to be very interesting to see once those conclusions on the extended proof of concepts, start coming in, what does that actually look like for business?
[00:31:32] Are they going to be still so bullish on, rebuilding everything with AI and slashing half their workforce? Because now AI is going to do it. Let's see, let's see, at least it's unclear, how that's going to play out. But I think. This was just important to get out there for, I think for everyone in the market, just calm a little bit down to, you know, cool, cool your jets. look at this through reason and logic of what's happening. Yes, it's going super fast, but try and look at what's happening and determine the best path forward for, for your function, for your teams, for your business, and don't, don't panic over this stuff. It's a, it's actually an amazing time we're living through right now. these major transformation. Only happen ever so often. Yeah.
[00:32:17] Toni: but we never talk about this on the show for obvious reasons, because, you know, that's not what we talk about, but there's a little bit of a, of a thing right now in the US politics that kind of is about like, don't, you know, don't listen to what they say, kind of listen to what they do.
[00:32:32] Right. Because there's a lot of like stuff being thrown up and things are being changed and so forth and everyone is like in shock. And I think the same is kind of true for this whole AI thing, kind of, instead of listening to all the LinkedIn influencers and you know, the YouTube, you know, stars and, you know, what's happening on X and Twitter, like start looking at what is actually, what is actually happening in those organizations, right.
[00:32:58] And, and what is happening in. You know, the, the, the, the, the SMBs of today, what is happening, the majority of business, what is happening in the, you know, the, the majority of the last corporates sure they have, you know, some arms that kind of do some interesting stuff, but 99 percent of their business is completely unaffected by this stuff.
[00:33:17] Right. So look at little bit, what really is going on instead of the hype. And I think that will give everyone a little bit of a balanced feeling. Does that change that, you know, Mikkel and I, I don't know, maybe you speak for yourself, extremely bullish on AI and thinking that this will, it will have a pretty strong impact.
[00:33:34] No, it doesn't. But it also means, you know, hype is hype and hysteria is hysteria, and we just sometimes need to kind of figure out how those two things play together. And I think for right now. Keep, keep an eye out on AI and how it's developing, but also don't don't, don't play into the, the hysteria keep calm and, and chill the fuck out.
[00:33:52] Mikkel: That's it. That's it. I also just want to, I owe you an apology, Toni because I kind of stole the intro from you. You're like, I have it today. And then I just swooped right in and stole it. So I just want to, preempt and say, uh, apology. And also I want to say to you, the dear listener, if you made it this far, it means you actually listened to the chunky middle, not just the intro and outro bit, really appreciate it. do hit subscribe, like follow if you haven't already or drop a review. It helps us grow this show. Toni, thanks so much.
[00:34:19] Toni: Thanks Michael. Thanks everyone for listening. Have a good one. Bye bye.
[00:34:22] Mikkel: Bye.