20VC: Why Most AI Investments Will Do Worse than the S&P 500 | Why Early Stage VC is F******* | The Danger of Kamala Harris and Why Trump and Vance are Best | Freedom of Speech, Censorship and Government Control with Eoghan McCabe @ Intercom

0:00Most VC money is going to support a lot of excitement around AI and actually not do better than the S &P 500. I think that building your own AI agent platforms is going to be as smart as in the previous generation building your own SAS workflow. The terrible fucking idea. I'm a big believer in individual liberty and peace or in a time where both of those are being threatened. It's crystal clear that Trump and fans are the people who are promoting the most personal liberties and freedoms. I think it's deeply evil to decide what can or cannot be spoken about. It's really dangerous. Governments become insanely and dangerously powerful when their citizens cannot say what they believe is true or not.

0:41I mean this show has more peri -peri -sourced than a Nando's extra hot chicken. This is one of the spiciest shows we've done in a while. Also such range of discussion for a short show AI politics freedom of speech. Oh in McCabe co -founder of Intercom is incredible here. If you want to watch the full episode, you can find it on YouTube by searching for 20VC. That's 2 -0VC. Before we begin, I want to share something interesting about the private market. There are many ways to buy and sell private stock today. But Hive is the only platform that provides a one -stop liquidity solution for private companies and their shareholders.

1:15It not only provides access to a deep pool of investors, but also allows companies and boards to control key elements like buyer approvals, seller eligibility and even pricing, offering liquidity on their own terms. What's more, Hive's comprehensive transaction execution service and centralized transaction tracking allows companies to stay on top of all past pending and upcoming transfer requests, so easily and at no cost. Visit hive .com, that's h -i -i -v -e with two eyes, and see why they're the fastest growing liquidity platform in the world. Hive markets limited is a broker dealer and member of FINRA, SIPC, investments in private stock are risky, only accredited investors may buy through Hive.

2:03And just like Hive transforms liquidity, TIGUS transforms how we gather insights for investing. When I invest in software today, I think software is valuable when it saves time and makes something easier, but when it enables the user to do something they simply could not do without it. Then it's insanely powerful and valuable. That's the case with Tegas. I literally could not live without them. When I'm investing in a company, Tegas lets me see the most incredible reference calls with experts in the space who work at a competitive company or incumbents. These calls and data provide the most incredible insight that really shapes my opinions and thoughts and investment decisions.

2:39And I can then share snippets from calls with team members. This is a tool that will change how you work. Honestly, you cannot imagine life without it once you've experienced how powerful it is. And you can check them out today by heading over to tegas .com. You have now arrived at your destination. Owen, I am so excited for this dude. I always remember our last show. It's one of my favorites that we've done. So thank you so much for joining me, stay man. Thank you for having me back. Not at all, but we always start with the context and you've been asked 10 ,000 fucking times how many times you found it in to come.

3:10And so we're just going to skip it. I want to dive right in. We are in a very interesting time of AI. A lot of people are saying that customer service is the most impacted and will be removed with Intercom's position. Why are you not fucked by AI? Sad with love. I appreciate that. I mean, everyone's fucked by AI in the fullness of time. I mean... No, that's just the title for the show, is it not? That's the intro, that's the segment. Unfortunately, I'm fortunate, I regret that already. Yeah, in the fullness of time, increasingly powerful and abstract AI will do away with all of the work we need to do today to deliver software experiences.

3:48That has always been the case with technology. There is no new technology that won't be obsolete in some period of time. And I wouldn't be working my ass off. And Edricom, if I thought that it was about to die and be crushed by AI in a year or two. So I think we've got a decade, two decades. So the conventional wisdom in AI circles now is that actually customer service and customer the sport is the low hanging fruit and that's going to be picked off in the next year or two. What position does Intercom play in that landscape moving forwards or do you disagree that is not the time frame and it will be a decade plus?

4:20You know the reality is that adoption and new technology takes far longer than we all always imagine. We were the first to come out with an AI agent that would do customer support for you like actual conversations with customers and closing tickets. it's we now have, forgive the fucking pitch. We now have the highest performing AI customer service, but you look at everything that's on the market. They can't resolve tickets at the rate we can, the average resolution rate of our bodies, now approaching 50%. How is that, Owen? You leverage open AI and alternative models to draw that efficiency? Yeah.

4:55Just help me understand, I don't actually understand, is it the existing data sources that you have which allow you to do that better? Yeah, it's all the existing data sources is with the intercom it's your historical conversations that your human reps are also having. But what's really interesting and what's key to understanding how the AI is going to develop and we can go back to your questions if I manage to remember is that building domain specific AI systems requires a phenomenal amount of work on top of these LLMs. Of course, yes, over a decade or two of the LLMs will get powerful enough that they will disrupt away all of these detailed intricacies that we've built.

5:30But FIN, our AI agent, for example, is the product of over 100 unique experiments to figure out how exactly to understand customer questions and resolve tickets. There's patented components to it. I think there are somewhere around seven different LLMs intricately connected together. This homegrown AI and just a lot of very detailed prompt engineering massaging to make it highly effective at closing tickets. When we first released our AI agent, the average resolution rate was in the high 20s. So that's the rate at which it successfully closes tickets when it gets involved in the conversation. And now, like I said, it's approaching 50%.

6:09I think Zendesk is still high 20s. And so it just takes a long period of time and a lot of work to build all of the functionality on top of the LLMs to make them really, really great at these domain specific things. That's why these AI application companies are going to win. That's why they're going to become so big because I don't think at least in the next five 10 years that any of the AI labs are going to build all the domains specific stuff. I interviewed Ethan Molyck, he's the head of ML or AI at Warton and he said you cannot believe in AGI and also be a SaaS investor. Right. That's right.

6:45If you believe in AGI, then every SaaS product will be killed. Straight off. Yeah. I think that's right in the fullness of time. And again, if you held on to your stock in facts companies, you'd be crushed. But for the first 10 years, you would have made a fucking fortune. So in technology, it's all about timing and it's all about waving success of growth cycles and riding the trends in that moment in time. And then you must reinvent yourself and kill yourself. You asked an interesting question at the start. We've already forgotten what it is. Well, I also have a question which is like, why are you not thought by AI?

7:19AI. And that's the answer. It's because, sure, AGI crushes all of software in the fullness of time, but it takes a really long period of time to pick up these technologies in the question you asked. Do you think that we are overestimating technological advancements in AI? I mean, people are now thinking that actually with scaling wars and with bluntly the amount that incumbents are spending, we could actually get there so much quicker than we ever thought. And we are a scheme reaching a scheme. Do you think we're not actually, and that is over hyped? There's various aspects of it. The rate at which AI is going to change the whole world is a little overhyped.

7:53The raw technologies will be built long before they have a real impact on the world. That said, even the technology community at large has not yet made contact with how valuable AI already is. There are 100, 200, maybe there are 500 companies globally that have built AI first, products that are blowing up. I saw some statistics that showed that the average time it took for a SaaS company to reach 30 million an hour or from a specific investor who was picking great SaaS companies was something like four or five years. And the average time for these AI companies to reach 30 million an hour or was less than two years.

8:35AI application companies are currently blowing up in a way that people don't realize. They're currently delivering a lot of value. This is my hardest thing as an ambassador, which is to termly sugar high revenues first is actually long enduring sustainable revenues because you're absolutely right, I invest for a living day today. I see 10 million ARR, the whole freaking time from AI companies. Yeah. Before that would in 18 months have been the gold standard. My challenge is the quality of revenues much lower, often their POCs and pilots if they're enterprise at all. Sure. It's the most exciting time to be investing and it's also the most challenging.

9:06But of course, when anything new happens, any new trend, there's going to be so much noise. And it's your job to be one of the few investors who can spot the signal in the noise and beat the market. Most VC money is going to support a lot of excitement around AI and actually not do better than the S &P 500. There's a brilliant stat, which is like, in the two years following the family of Netscape, only 1 % of internet enterprise value was created. Yeah. Now, they were the very early days. And still, this has to be the very early days of AI, and there will be a lot of AI companies that will totally flame out.

9:42I stand by the assertion that there are AI companies delivering really tangible value, and the things you pointed to are exactly right. Usage and retention is how you spot the difference. So there is definitely a lot of revenue that's driven by just excitement and people kicking tires, but there are AI businesses out there where they retain customers. Those customers have totally changed the way they've worked. They're expanding on the product. They're not going anywhere. They've forever changed. Clana announced that they were stopping their customer contracts with that work day and with service, with Salesforce.

10:13And it led to this wave of what? This is the future. Companies will build their own SaaS tools. I'm interested. How do you feel about that? And how do you respond to that question and concern? Well, I have no choice but to say that that's a terrible fucking idea. And just like Intercom is eventually fucked by AGI, we'd also be fucked by companies building their own AI systems. The reality is, hand on heart, independent of my association with Intercom, I think that building your own AI agent platforms is going to be as smart as in the previous generation building your own SAS workflow tools. These things are super deep.

10:49Like I said, with FIN, our AI agent, it's the product of over 100 different experiments, many patented parts. We've got 30 ML engineers and counting, 100 other engineers working on all the application components. You're just not going to build a highly performant rich AI agent on the side. Not going to be done. And I think any attempt to do so may yield some short -term excitement. But even clarinet long -term, I'm pretty sure we'll move on to a professional purpose -built system. I completely agree. It's one of the most ridiculous misnomers in the valley right now that everyone will build their own individual tools.

11:23Next no sense. There's two worlds. One whereby It continues to cost a lot of money to provide such great high -quality AI products and you do a Canva, an increased pricing by 300X. Well, I'm Canva, love those guys. Or actually, we see costs go way down as they have done so far and actually pricing remains relatively stable, but it's the cost that really reduces. What world do you think is more likely? And why? Technology is deflationary. Technology gets cheaper and cheaper and cheaper over time. Look at these statistics where they show the real prices of various products and goods, and even as currency inflates, the price of, say, TVs, flatscreen TVs goes down.

12:09And so these AI agents will be way cheaper than humans while they're also better. That said, what you're going to see people do is augment their savings they make in their budgets or use the savings they make in their human budgets to spend on as much AI as they can until they're now totally over -serving their customer. I think people are going to start to find that if, for example, in our e -commerce story, these companies can generate substantially better sales through providing these epic AI assistance that will happily spend more money for them. But when those themselves become commodities, then it'll get cheaper over time too.

12:47Do you buy the reduction in team sizing that AI has? I think if you look back to the 70s, people would say, God, if we had computers that did all these things, we wouldn't need anyone here. Yet there's still companies with hundreds of thousands of people. How do you feel about the reductive nature of AI on team size? Valuable businesses have got smaller with technology. The size of Google or Facebook when they were worth $100 billion in employee size was substantially smaller than and generally that electric or Ford. The long arc of technology and automation requires less humans. So it'll perpetuate that arc.

13:23You're not gonna see mass layoffs. At Intercom, for example, we've been using our this AI agent fin for nearly two years, and we've not let anyone go, but we've stopped growing the size of our support team. I think any future growth is gonna be catered to by the AI agents. The AI agents will first do a lot of low level and simpler work. and the humans will do the higher level work. Now when we hire for customer service, we're hiring a more talented experienced person. You can do the creative things that the AI agents can't. Eventually, as these teams get smaller and through attrition, the AI will be doing more and more of them and there just will be less humans in the mix.

14:02I am asking a spicy one here. Is Sebron then to fire 700 people in Clawna's customer service team? If they could do it, capitalism dictates that they should. But if they do, the way you do it, like why go on average? They must be something else. Certain use cases where the distribution of the types of questions is really tight, rather than this big, long, normal distribution of so many different types of questions, will be able to create these tight custom experiences. If you're a clarina, for example, you might only get three questions that cover 80%, three types of questions that cover 80 % of your inbound.

14:41And therefore, you can actually create AI or AI -ish experiences to do big swaths of your customer service. Most businesses are not like that. What extent do you think foundation model layer companies, your open AI, your anthropics, subsumed the core application layer companies? Customer service may be being one of them, language translation may be being one of them. To what extent do you think that's a reality? Again, in the fullness of time, if they manage to become ever closer to something we call AGI, they will. Storytellers and visionaries are very different from builders and business people.

15:19These are just different types of disciplines and different individuals. So the people that tell these big, visionary stories about how LLM companies are AGI, it's going to crush all of software. They're not actual business people. They're not actual builders. and they don't realize that businesses, consumers have real practical problems that need solving and it takes them a really, really long time to adopt these technologies. Can I ask you, what do you believe will happen in the next 10 years that can't need today is considered crazy or nuts? I think we're going to see a giant rise of robotics in many different areas.

15:56A lot of blue color work will probably be done by robotics. The one thing that's important to say while I'm celebrating this human work being done is that technology has always put humans out of work when the work has been dangerous shit and demeaning. Whether it's individuals on factory floors where they lose limbs, humans dying in mines, so much of these types of work that humans fucking hated that actually killed them has been replaced by technology. and all the while the population of the world has increased, GDP has increased, longevity has increased. We have no reason to believe that we're not still on that same trajectory.

16:38How do you feel about the photographer who's gonna have their spend cut in half because we can do AI backgrounds so much easier, we can reduce spend on a photo. I literally just came from a photo shoot and I was like, dude, this is gonna be mostly AI in three years. And he was like, I know. They'll learn new things over time. When I got into the technology industry, I was making crappy websites, writing HTML by hand. Now we have so many different ways to generate websites. You've certainly have seen so much web designer work disrupted by wicks and squares base and Shopify and all of these tools and technologies and frameworks.

17:18But actually, these people were never put out at work. The demand, the greater demand that these people could not serve was catered to, and those people got more specialized. It's going to continue to be the case. Something I find really challenging on is you look at the latest YC batch. All the companies are relatively interesting AI consumer application layer companies in large part. And actually when you dig deeper a lot of them is kind of 11 labs, tag to an open AI model, and it's really just bringing together different foundation model companies. And the core thing comes back to, well it's not really that much defensibility, it's just kind of awkward, etch, you start together.

17:55How do you feel about defensibility in startups? I know you've invested too in that. There is no such thing. There's no such thing. Even Apple don't have it. Apple very famously championed their patented iPhone technologies and yet Samsung came and actually beat them on many different criteria. Technology is a kind and most game. It's about who builds it first and brings it to market, helps customers understand and adopt it, always being the case. And so talking about modes, it's kind of silly. Now that said, some of the greatest technology companies of our generation then go on to build networks and other things that do become defensible, whether it's Apple's eye messages, whether ecosystem, whether it's Google's data, whether it's Facebook's network.

18:34At some point you can build some degree of defensibility, but software companies and technology companies are particular. They don't tend to have any defensability other than the fact that they can build better and faster than anyone else. And that's OK. It speed the single most important thing in going from zero to one. To one of them. To one of them. Working your ass off, working very hard. Can I ask you, if we think about Silicon Valley Central casting, PG's essay on Founder Mode when incredibly viral obviously in the last month, how did you feel and think about it when you read it? I think it was really acute and fun, branding for a thing that anyone who's built a company has deployed since the start of companies.

19:11There's nothing new about it. There's countless books about Steve Jobs and how he operated. Maybe it gives permission now to founders to trust their good and instinct, which I think is a really, really good thing. But business is very much an art, particularly starting a business from scratch, as a deeply creative endeavor. The worst type of art is art produced by committee. I would say it's just not art if you're trying to find the lowest common denominator and something that satisfies everyone. And so highly agentic individuals and acting on their own intuition, I think it's how greatness has always been made.

19:46Can I ask you, I don't know if it's public, but how much do you guys do in revenue today? It's in the hundreds of millions. So it'd be one of the larger private software companies. Why don't you go public? I always hear girly and you know, Brad and I'll tell him to turn their like more companies need to go public when they're in this range. Yeah, the valuation will take a minute. I mean, of course, a VC would say go public. Look, these VCs need liquidity. Founders who have a different mission and a longer time horizon and who actually need to operate the businesses unlike Gurley have different priorities.

20:19My goal is to build a giant, highly impactful, really fucking cool technology company. It means absolutely nothing to me whether or not my stock is traded on the public market with a private market. The reality is today. data for you in terms of rigged accountability, all of the reasons that they say. They're the things that the analysts want. They're the things that the investors want. I want a shit ton of customers, a shit ton of revenue. I want a really impactful organization. I want to have a really, really big amount of influence on technology at large. I don't need all of the regulatory compliance to do any of that.

20:58All the regulatory compliance gets in the way of that. But the reason that people take public companies private during crazy moments like this post -COVID now post AI Is so that they can make really big swings and big bets without all of the rig and world of being public We nearly went public in 2022 and I'm so grateful that we did not Talk to me about that. I didn't know that. Yeah, 2022. I think it was June 15 We had a stock ticker that I hated nothing cool was available We had written our S1 and the market's collapsed. We weren't not the only ones. How did people respond? Investors get excited.

21:35Equity coming. The right investors want value long -term and they know that in these moments a great change that the best value will be created behind closed doors. All the big swings and changes I made at Intercom that has dramatically accelerated our revenue growth. I don't think I could have made that if I was a pull -to -comedy CEO. I am an investor though. We need liquidity. and M &A, what you're going to do a Dylan and go through 18 months of hell and then get blocked. No, you're probably going to do a whiz and just say no thanks at the start. And then I POs, I'm with you on everything you just said.

22:09So what happens then? We just have this extended private window? There will be IPOs, there will be liquidity, maybe on private markets, M &A will warm up again. There's a lot of money in PE. It'll have to on stick itself, but admittedly we're currently in an awkward patch where a lot of people are asking themselves. Does this VC model even make sense? I mean, Intercom is a 13 -year -old company. When people were investing in Intercom, they thought maybe they'd get liquidity in what, eight years. And we are, by far, not the only company of that 10 years. So, the game has changed. Do you think the VC model still works?

22:45The amount of capital at an early stage available today suggests to me that that's got to be a crappy end of the market for VCs. It's just a raw commodity now, early stage VC. I don't know how you make a return there. You mentioned the revenue there and the seismic shifts that you've done made to cause the changes. Why did revenue, plateau, growth, plateau, what happened? We forced more and more customers to talk to sales. It certainly increased ACVs, but it actually cut out a phenomenal amount of growth in SMBs and small customers. And those small customers were actually the lifeblood for our sales customers.

23:24They just needed some time to grow. That was a really, really, really, really big mistake. Our total customer account started to shrink. That's one example. Oh, there's where we're pricing. We got greedy. We wanted people to sign big contracts upfront rather than allow them to expand into them. So, it was just going against all our intuition, bad commercialization. That was the main thing. Do we want to talk about the internal changes that intercom about restarting the startup? Yeah, the culture. Companies are this deeply interconnected system. You know, you can't change one thing without changing another.

23:56You can't pick and choose. It's not an Alcar menu. It's a deeply interconnected system. We needed to totally change our values, make clear that hard work and a focus on our mission was key, was crucial. I gave permission to people to act in quick and scrappy and messy ways. I rolled up my sleeves and gone evolve myself. I set this project called Project 52 where I gave us 52 weeks to reinvent the company, I spoke to the company every single week. I just created this very high degree of intensity. I rolled out these new performance management processes where people who didn't fit with the new values I wrote and the values were things like hard work and resilience and a focus on our mission, where respectfully moved out.