80. Crossing the Vibe Chasm: The Most Important Tech Skill of the Next Decade

0:00Welcome to The Difference Engine, the show for tech founders, investors and innovators.

0:09hey Paul what's coming up today we're going to give our review of London Tech Week and see what this annual schmooze fest actually tells us about the future of UK tech success but first we return to our concept of the vibe chasm we know it exists so how can we cross it

0:32so the last episode of the difference engine uh we introduced uh the idea of the vibe chasm this is a phenomenon which emerges as a new type of company is built these days you know the ones where a single founder armed with ai tools and vibe coding can build products acquire customers automate operations and grow revenues to unbelievable levels that previously required entire teams. So just to recap once again the vibe chasm analogy we're using is actually borrowed from Geoffrey Moore's Crossing the Chasm analysis which he published back in 1991. Can you believe it so long ago Paul? Nearly four decades.

1:09Of course what was perfect years before the iPhone and months after Mythos needs a juju and that's where the vibe chasm comes in. Now to be clear These vibe-chasm-crossing companies are not theoretical anymore. They're already appearing across software, media, commerce, consulting, education, and indeed niche SaaS. But they are far from being bulletproof. Yeah, you're right. The pattern's increasingly clear. You've got one founder, tiny operational overheads and supporting teams, all AI native workflows, There's a multimillion dollar turnover, preferably, and extraordinary profitability. Yeah, but you know what?

1:55What? Then something happens. It's a sort of luxury problem. And it's because of the leverage that AI gives the founder, eventually, if they're being successful, the customers will become enterprise grade. Now, Paul, that definitely means something else. For sure. Because that means reliability expectations arise. compliance actually matters, security actually matters, uptime actually matters. Oh, and financial controls. Think about them. Hiring becomes important and governance becomes essential. So if you think about that, the jeopardy to a growing firm could be worse in the world of AI-powered apps because switching costs are lower.

2:42Just look at the furor from the Mythos Fables switch off to Barclay. Yeah, one issue for an enterprise in dealing with an AI app supplier and your competitor could be in light Flynn. And answered calls and responsive support chat next. So back at AI Central, the move fast and break things in formal AI narrative operating model collides with the realities of operating at true organizational scale for enterprise customers. Now, what happens then, Paul? Want to give it a guess? Well, sadly, the founder becomes the bottleneck. And reading careless people at the moment, I think we know just how real that could be if your name was Mark Zuckerberg.

3:24Your growth begins to tail off a little bit, which disappoints the investors. Meanwhile, the complexity is compounding and the good times are gone because your systems break. That means what? The vibe chasm is open. Now, bridging that paradigmatic shift means crossing the vibe chasm because for years, the defining startup skill was building, building itself. Today, AI increasingly commoditizes building. So, the emerging differentiator will instead become knowing when to transition from individual free improvisation, using a bit of jazz analogy here, to agile orchestrated systems. Now, we believe that means the winners of the next decade may not be the founders who use AI the fastest, nor will be the companies with the fewest employees.

4:17But they will be the organizations that cross the vibe chasm effectively. This is the old school story of best of both worlds in a changing situation. So you want to preserve as much as possible your speed, your creativity and AI leverage, or perhaps more accurately, AI agility, especially when things change as they did when Mythos and Fable got switched off. At the same time, you want to introduce some really boring sounding operational rigour, scalable architecture and sustainable management systems and processes. Right. So in other words, that means the future may belong to companies that combine the fluidity of AI native creation with the discipline of enduring institutions.

5:02Now, you know, are these things completely at odds? It might work, but it starts by consciously defining a new corporate life cycle. So that's got major implications for the nature of entrepreneurialism and timing and the purpose of funding. Right. So just to put this into context, Paul and I have been around a long, long time working with traditional startups. And, you know, they tend to follow a predictable sequence, don't they, Paul? Yeah. Go after the money, raise that capital, first order of business, then hire aggressively. Don't worry, you can always sack people later on. Build that product.

5:43Build, build, build. Drill, baby, drill. Cross the chasm. Then you think, oh, I need to seek a bit of scale. Then I might need to optimize for efficiency. And only when all seven of these steps are done can I dominate my category. That's all changed, hasn't it? If everybody's been listening, they'll realize that AI native individuals and AI native companies actually invert that life cycle. So what does it look like then, Paul? Well, in this case, you just build as your first order of business. It's even faster than minimal viable product. It's just product. Then you automate aggressively the building of that because you can with AI.

6:20Delay your hiring as much as possible because you've got AI agents. Reach some sort of profitability quite early. Then go after the capital. Then introduce the necessary systems. Then try and cross the vibe chasm. Then seek scale and then hopefully dominate the category. Very different. Yeah, and it does create a fundamentally different type of business evolution. So the early company resembles a highly intelligent organism. The latter company increasingly resembles a structured institution. Now, it's the transition between those states where the hardest problems emerge. And this, you know, Paul, you were talking to me the other day about some of these issues around anthropic.

7:02Yeah, I think, you know, there's been a shot across the bowels on this whole AI scaling ethos with the discontinuity we've just seen with Fable and Mythos. So as we said earlier, there's a question here of whether you can go from massive autonomy and how much agency you have and how much control you have over your future. Of course, in the world of vibe coding, when the vibe coding stops, as it recently did, due to the actions respectively of vendor, Anthropic declaring mythos is too much for ordinary people, and then the US government saying that Anthropic's not a friend of us and we need to stop it and call it a nasty supplier.

7:39When that happens, the chasm grows. We are in very interesting and testing times at the moment. But, you know, let's assume that you can successfully cross the vibe chasm. The likelihood is you'll become extraordinarily powerful. But I think it's if you can build the resilience to separate category build from the vibe stage. Now, why might that be? It's really because they begin with advantages traditional startups never had. then what might they be paul well yeah so this is a really a question of best of the new best of the old so best of the new ultra low operational costs forget buying servers forget renting lots of amazon compute why not just buy tokens and you've got ultra low operational costs that also means you have deeply automated infrastructure um the culture of all your folks is very ai native and therefore very us and them, the decision-making can be quick and you can be profitable if we disregard the token economics per se.

8:46You can be profitable from an early stage. But that's great. That's when you're growing like crazy. But other stuff is going to have to come in. It might not be too palatable to the AI native culture. But you and I, Paul, having been around the block a few times, know that there's things that grown-ups have to do and what might they be? Yeah, you're going to want some elite operators. Folks who've been there done it. You can't just vibe your way through certain things, especially if you're in a regulated industry, for instance. You're going to need systems that truly scale. Not scale when the LLM is up.

9:23Not scale when the LLM is not sanctioned by the US government. Not scale when the LLM has decided it's going to downgrade the model you're on. You need scalable systems. and that requires organizational discipline and governance and all of that much as it pains folks to find out requires money so you need the finance to support all of that and that's that's what's really interesting because the critical time for funding in companies going forward is not going to be right at the beginning and it's not going to be the a and b rounds it's going to be how do we transition this company to cross the vibe chasm.

10:04Now, if you can do that, these companies will achieve levels of efficiency that traditional companies cannot match, simply cannot match. And this can create a new category of firm, small in headcount, massive in output. So it's not a startup. It's not a traditional company. It's something in between. And this is what we call the AI amplified enterprise. Some of those AI amplified enterprises are going to be revenge startups from all the layoffs at Oracle, Microsoft, etc. That's where these AI amplified enterprises are going to come and God bless them. But as we've said earlier, we've learned a lot here.

10:45We're trying to teach some new lessons to folks who may not have seen it before. As we said before, for decades, technology just did one thing, reduced the cost of distribution by automation. Now AI is reducing the cost of the organization you have to build itself. Right, that's got profound consequences, hasn't it? One person can now build what previously required hundreds, but, and this is the big lesson of crossing the Weidkazen, scale still introduces complexity and complexity eventually demands structure. And the defining challenge for the next generation of founders will not simply be building with AI.

11:28It will be recognizing when they've reached the edge of solo scalability and learning how to cross the vibe chasm before growth collapses under its own weight. Perhaps this is stretching the analogy a little bit, but if you're like Roadrunner in the famous cartoons and you are vibing your way to profitability and vibing your way up the user numbers nice and fast, it's very hard to see where that chasm starts. Right. So the way to think about this and the journey is building with AI is great for proof of concepts and minimum viable products and even early stage sales. But scaling and realizing the full value of the idea through category domination requires a different approach based on additional skills and experience.

12:21so another london tech week has wrapped up a big stage big speeches big numbers do you think it actually moved the needle for uk tech honestly no i mean it does feel at times as ctm uh said um it's complacency in conference form I've not been a fan for so long. Even though it's just down the road from us here at Olympia, so it's sort of coming home, so to speak, I have to say, what is this thing about? Well, we roll out the same unicorn statistics, celebrate funding totals, give ourselves a big round of applause, and then go home acting like the job is done. Well, because it's on to the next by-election or the next government crisis.

13:08Yeah, and that really begs the question, who is actually behind London Tech Week? who is dumping up the bills for all this jollity. Yeah, yeah. So what do we see? Major announcements, AI infrastructure investments, partnerships, government engagement, right? But you just ask the founders building companies day to day whether those announcements change anything meaningful for them. Yeah, most won't see the difference. And while politicians were making speeches at Olympia, I actually think a huge proportion of Europe's serious VCs and LPs were across the channel in Berlin at Super Return. Now, that would be ironic, especially, and we'll come on to it, when we figure out who's sponsoring the Tech Week.

13:55Our issue isn't just the event itself, right? Isn't one of the issues here that London Tech Week actually puts possibly two deeper problems on display, right? So first is Britain's relationship with its own success. and let's face it, tech is not immune here. We've got an incredibly negative narrative around our economy and our tech sector. We spend years talking about what we haven't achieved instead of celebrating what we have. As an example, take DeepMind or Arm. For a decade, we've been lamenting their acquisition by Google and SoftBank instead of asking how we build the next category leaders just like them.

14:37It depends who you listen to, but the former head of MI5 was like, we should build our own Google. And, you know, we've got people who've never had jobs in commerce, never mind in IT, trying to big up tech. It just sticks in the craw. The problem is when criticism becomes the dominant story. I mean, look at political leadership. The prime minister for the next few weeks, at least, is called Keir Starmer. um you know his most visible engagement with technology this year has been around online safety and children's access to social media now that that is an important issue um but if the primary message from government is that technology is a source of risk and harm that's the signal you know uh investors founders talent um anybody outside the uk to receives talk Work is cheap and picking on something that tugs at the heartstrings is great, but the cold hard facts of what this government's doing to the EIS investment scheme and adding to the cost of taking on new talent, I mean, it goes directly at odds.

15:50We need leaders who can hold two thoughts in their brains simultaneously, right? Blown it, you know, walking and chewing gum. In this case, protecting citizens and building a globally competitive technology infrastructure and ecosystem. This is what's so frustrating. We have the world's third largest tech ecosystem, yet many investors still feel compelled to, and I've seen it over and over again, to begin presentations with slides explaining why the UK is worth investing in and why London is a good place to build a startup. That's a remarkable Stalmer impersonation there. Yeah, it's exactly right.

16:23It's meant to be a general nerd. nobody's opening investor decks in singapore nairobi so paulo wherever to say this is why this city matters you know if it's got the talent it matters yet in london we still need the feel we still need to feel the the need to justify ourselves despite the fact we've got the deepest concentrations of finance the insurance industry is basically based out of london we've got amazing technical schools as we've heard on this pod before and now offering real money as the u.s coming off of real salaries here. So that's negativity nailed, Dan, right? And there's another one, another word I hate, another action I hate.

17:01Complacency. Yeah, right. You got it. You've heard me going on about it many times. You would also not be surprised, I think, that's the more dangerous one because London Tech Week exemplifies this. It looks backwards too often. It's the same cast of characters celebrating outcomes that were created by conditions that no longer exist. You know, we won much of the 2010s thanks to a generation of founder-ready talent emerging from the financial crisis and a regulatory environment that actually competed for innovation. Right, totally. And let's have a little look, shall we, at where this allegedly, I mean, to me, London Tech Week is just three words that politicians like to say these days.

17:43I know because I'm in a WhatsApp group which was founded from a previous London Tech Week, hat tip to the potatoes group. But these guys are all, it's a collection of genuine innovators and startup founders, etc. They can't afford, and we're bitching about how much it costs for an average punter to get in. Meanwhile, let's have a look, shall we, at who the sponsors are for this year's Boondoggle. See if you recognize any of these British startups. Microsoft, Amazon Web Services, Accenture, Aviva, Nscale, HSBC, and of course our Chinese brethren, BYD, and I've never heard of this company, I've got to be honest, Denza, D-E-N-Z-A.

18:26I think you'll find they're also from China. They are indeed. So does that smack of leading-edge innovation, or is it more like a public affairs outreach? We don't seem to have replaced that sort of gung-ho 2020 vibe with anything equivalent. And I would contest that our early stage ecosystem is weaker than many people realise. We do produce lots of startups, people I speak with every day, but only a very small percentage make it to serious A. You know, quantity isn't the same thing as quality. I would also argue that it's not just about quality. Early stage funding is still a mess in the UK. And too many early stage companies have to spend too much time running around trying to raise small bits of cash.

19:22Which makes you think, again, who benefits? Qui bono? Who are the folks behind London Tech Week? Well, it may or may not surprise you to find out in the main, from what we can see, there are three folks behind it. There's Founders Forum. That's a friend from last minute, Brent Hoberman. And Martha Lane Fox's little sister, a little brother, rather, excuse me, Henry. There is Informa. So a very much a part of the tech ecosystem. Been going around forever. Owns a bunch of analyst firms, owns a bunch of conference companies. Nothing to see there. How surprising. And then the third partner, and we couldn't dig up much in terms of the finances of London Tech Week.

20:13Curiosity leads us to believe, does this thing even make a profit? But the third person in that triumvirate of folks behind London Tech Week is London and Partners. Now, you may remember these guys. this is a government quango set up with um mr khan and all the rest of the guys to talk about london being a great place to um do business we've obviously had silicon roundabout that was amazing success um and now we've got london tech week it seems to be of the same ilk let's um think about london partners london partners have 287 people uh working for it doing great work there average pay is 69 ,686 which even in London is a decent you can rub by on that and two of the directors I don't know who share 356 ,000 pounds or 178 ,000 pounds a pop but the important thing to remember about London Partners is this doesn't this isn't really a commercial enterprise this is something that we're paying for so we're paying for our own government to big up what they see as the vision of their success and then feed it back to us as if it's individual success.

21:21It's very beautifully circular. Indeed it is. I just wanted to return back to some of the basics that do surround our somewhat self-congratulatory attitude around London Tech Week. If you look at tax schemes, there's a real obsession in the political class in this country. it's been going on for years, about startups. And for some reason, everything seems to be about startups, almost to the extent that the scale-ups, which are truly going to succeed and are truly going to put us on the world stage and are going to truly create employment and a platform for further innovation, sort of get ignored.

22:04And that is definitely in the tax system at the moment. You know, I'm a big fan of EIS and some of the other systems. Should explain, maybe a little quick explainer on EIS? Well, we have in the UK, we have a whole load of tax incentives for people to invest in early stage businesses, where essentially, because of the tax rebates you get, there's really no risk. Well, that said, you could lose all of your money on the startup. So these are for people who've got a little bit of extra cash. the sorts of people who will take a punt on a startup. Yeah, quite possibly. But again, the tax benefits do make it very, very low risk in terms of your personal wealth.

22:49So what we are doing is celebrating early stage startup creation. And I think we pay less attention to scaling those companies who are most likely to succeed. And then, of course, we have the perennial issue, which we do not seem to be able to get a straight answer on. which is talent and it's still clear that our you know for the talent we need to build to build our technology base because we don't produce enough people of our own is that our visa system is too expensive too complicated and too slow so meanwhile countries like france germany sweden are competing aggressively for the same founders engineers and operators so while we're busy congratulating ourselves on our ecosystem rankings, we are actively losing the key ingredients for future growth.

23:42Yeah, and maybe the conversation should be a little less government-centric. It should be a little less to the agenda of the corporate sponsors, all of whom have scaled up and some would argue are the reasons that we don't have so much growth here and some data sovereignty. So maybe a little less government involvement in the future might be a good thing for London tech week albeit you know the revenues such as they are may decline despite these criticisms we are we're still optimistic um very very optimistic in fact that you know that the fundamentals around the uk and particularly around the southeast are still exceptional um we have this very rare combination of finance research science and creative industries which do remain a genuine competitive advantage and the first generation of uk unicorns those are the real unicorns not the pretend ones who use it as a marketing exercise has produced thousands of experienced operators who understand how to scale a business and the good news is they are not just selling their first business or their share of their first business and retiring to the country they're playing it back they're going at it again and that's you know one of the secrets of the domination of Silicon Valley.

25:01So it is about getting the right conditions around those people because it's no good if you have prospered and exited and created wealth for the economy if the second time you go around the conditions have changed drastically and reduce your chances to succeed. Yeah and one of the things to think about when you hit wealth creators with taxes that make than think, and I know one personally who's renounced his UK citizenship to move abroad, bringing billions with him. You need to think about the joined up conditions for creating proper tech talent, not soundbites. Yeah. So the first thing I think we really do need is political leadership that actually understands technology.

25:50And it's a fractional minority that have any idea how tech ecosystems actually work and what technology is. And ones that engage with tech throughout the year, not just during a conference week when they can get on the Radio 4 Today program and bang on about tech, that they really don't understand. We do need to have some honest discussions about what's working and what isn't. We do need better support for the companies that are scaling, that are really nailing it. And God almighty, we need a visa system designed to attract top talent. And a recognition behind all of this, that economic growth and online safety aren't competing objectives.

26:38We need both. Yeah, let's not go, let's not even talk about the sphincter tightening presentation on the radio of what's going on with the social media ban. When asked if they were going to ban VPNs, literally the answer was, we'll get to that next week. It was horrible. Right. So what then is the real test, do we think, of a successful future London Tech Week? I think it's whether it sparks action rather than self-congratulation. The day we stop needing a why London slide in Investodex is the day will know we've tackled both negativity and complacency and i think what we're saying is sorry but this year's event didn't bring us closer to that no it didn't not nearly enough

27:23thank you for listening if you want to learn more about category design head to be categorical.com if you need help designing and dominating your category then get in touch contact details are in the show notes

27:39Thank you.

28:08and start building your own unassailable leadership position today. Working with the category design gurus at Categorical brings decades of differentiation expertise to your team. Book a consultation from our website today and we will send you our one-pager detailing how to start designing your own highly differentiated category.