This is an AI Transcription. It’s pretty good, but please forgive any errors.
[00:00:00] Jonathan: Welcome to The Difference Engine, the show for founders, investors, and tech innovators.
We've got some big topics coming up today. We'll take a look at how Apple is redefining privacy and why London Tech Week needs to take a long, hard look in the mirror. But first of all, we want to take you back in time to the start of June. Paul and I have just returned from this event and we have a burning question on our mind.
[00:00:35] Paul: We've been at SASTA. It brought up a question that seems we ask time and again. Can Europe create an AI category leader? And did you find it when you were there?
[00:00:45] Jonathan: I think I may have done. One of the sessions that was really worth sitting through was between Jarek Kutylowski, who's the founder of Cologne, that's Kuhn, for those better acquainted with German.
Um, and his company is called DeepL, D E P, Big L. And he was doing a little farsighted chat with a gentleman called Johan Butting, who's a partner in Visionaries Club. Now, DeepL is, uh, an AI player specializing in language translation, and Visionaries Club is an early stage VC fund with offices in London and Berlin.
Butting is the classic done well in the corporate world, ex Google, Dropbox, Slack, et cetera. But he's been advising DeepL. On its GTM strategy for the last three years. GTM go to market, right? Go to market. And I thought, as I sat down, This is going to be easy balls to hit, isn't it? But it actually didn't turn into some sort of hagiography as it might.
I mean, instead it actually offers some genuine insight to the journey so far of what I think is an AI category contender and why it might have a chance in a world of the U S based AI behemoth, such as. Microsoft and Google.
[00:01:56] Paul: You're right, because not a lot of people would bet on that right now.
[00:01:58] Jonathan: Yeah, Jarek warmed his audience up really nicely, because he said to them, everything is possible in Europe.
We just have to believe it. Imagine that said in an earnest German accent. It was, it was really encouraging.
[00:02:08] Paul: Sounds like an Adidas advert.
[00:02:10] Jonathan: Absolutely. It certainly got the audience's attention. And it turns out that DeepL is six years old, started as, as what Jarek described them as a super researcher and tech company.
There was a lot of use of the word super. And he did his PhD around machine translation, and he thought there has to be a commercial market for it. I also think there was a natural advantage, as he pointed out, that as a European, there's a natural understanding of the need to work in a world of multiple languages.
After all, you know, if you are based in Kuhn, if you drove about an hour, You're going to be speaking Dutch, French, Flemish, and, and of course the international language of English. Yarek started to talk about the early days of the company, um, and when they launched in 2017, it was clear that what they had was a proposition for business.
Um, there was a lot of interest in what they had becoming a PLG product, but the compelling use cases were absolutely in SLG and particularly the problem of developing and maintaining online presence in multiple languages around the world. In category creation terms, that is an important early realization.
It's the customer problem. All they had to do was figure out a different way of meeting it.
[00:03:23] Paul: And I wonder if they went with this product led growth strategy, that it would have demeaned the value of the product.
[00:03:29] Jonathan: I think so, um, and the margins, certainly. So I think the, the power of the product was the fact that they had, I, I, and it wasn't quite clear whether this happened by design or accident, Hit a customer need that could be reframed, you know, you don't have to do loads of in house translators retain an agency on retainer or suffer people churn rather use AI to do it and gain a lot more advantages, economy, accuracy, speed of turnaround, basically the ability to do more constantly with less.
I mean, that proposition was so strong, it meant they didn't need to spend money on marketing. Word of mouth kept the early adopters rolling in. So the focus was on enterprise, but the reality that wasn't going to go away. Was that each enterprise had different demands. So that meant they had quickly hire a sales force that could deal with this.
So the early adopters were not disappointed and became very, very.
[00:04:23] Paul: And I think that answers the previous question of product led growth. You can't sell this stuff off of the website. The customers are too different. They need a lot of hand holding.
[00:04:31] Jonathan: And Jarek explained that his challenge early on was that the company had to become both excellent at building and excellent at selling SaaS.
And he said, if you don't get that right, the likes of Microsoft who have maybe an inferior product, but with excellent GTM strategy, you'll just get crushed. So I thought. Sitting there going, you know, is this a potential category leader from a category point of view that fundamental insight into the reality of competitive ecosystems all goes really well for the future.
I think even more so was the realization that the entire company had to be aligned with this reframed customer need and this in turn made Jarek realize that the entire team. Had to understand why selling to enterprises was so different to getting PLG customers to order. And you can just imagine what's going on, uh, in the development rooms.
Those irritating checkbox requests, the long sales cycles that tech people hate. We're all part and parcel of selling big deals to big companies to get big revenues and margins.
[00:05:39] Paul: Open source has got something to blame here. You know, when you can, um, Work on a project as a sidebar to your day job, you can post little bits of code and watch the community take it up and feel all fulfilled and get the dopamine hit you get from developing small things very quickly.
[00:05:55] Jonathan: Yeah,
[00:05:55] Paul: that sort of developer will struggle in this sort of culture you're talking about. You
[00:05:59] Jonathan: need patience, you need people who are patient and can see the big picture. One thing Jarek said was that if we don't go after the enterprises now. Somebody else will, and it will quickly become too late to course correct.
[00:06:12] Paul: To be frank, we've worked with a couple of companies that have struggled to make that move.
[00:06:17] Jonathan: So, again, from a category point of view, I was, you know, having my boxes ticked, that Yarek clearly understood the alignment across the company on what it takes to be a leader. In emerging area was paramount, and he also emphasized that while you're doing this, he's having to deal with an environment where the translation business per se globally has transformed completely in the last two years.
So, you know, can you remain true to your category? Despite the fact that you need to transform to stay ahead and and we know the answer is yes You simply change your point of view and I think in the current wild west of ai You need to be prepared to transform to stay in the game. And of course that Requires a very grown up appreciation of ecosystems change.
It's clear that customers are struggling to adopt AI based solution. And here was another piece of genius from Jarek, I thought. It's very difficult to solve Currently, the sort of daily problems that we have in our desktop environment in a reliable way that employees can use effectively. To me, that's the beauty of their focus on translation.
It's a specialist solution to a specialist high value problem.
[00:07:41] Paul: This is an area that AI. It's making all sorts of promises with the GPT 4. 0 release talks about simultaneous translation, but AI will be adopted at different speeds by different people. And presumably, these guys are talking about catering to those different speeds that people want to adopt the translation services.
I think
[00:07:59] Jonathan: absolutely. And as Jarek said, AI translation has a long way to go. They're absolutely cognizant that just at the beginning, but my view would be that if they focus on translation. They can grow a big business there. And the learning of the potential AI issues and solutions can start to move out beyond the translation department.
So I think they've created a great bridgehead, but this is not an easy ride because the cost of investment is still very, very high for sure. He was very happy that even now the product is still selling itself and the money they can spend on R and D. Rather than marketing. And he says there is an advantage it makes for great company DNA as the enthusiastic early adopters give great customer feedback to developers.
And that's really highly motivating for the team. So again, thinking about Jason's point about you need people a hundred percent on it, this is one way to do that. And. The other thing which just kept coming forward was, we talk about AI like it's here, it's absolutely not. It's very early stage and he said, they have to spend enormous amounts of money, I think about 10 percent of their investment on research.
Because nobody really knows what's going on. Because they're operating on so few data points and very little experience, you know, they are inventing the future. And I thought, again, because this wasn't a hagiography. He was questioned about what I thought was the elephant in the room, which eventually emerged, which was, you know, do AI startups have a chance against Microsoft and Google?
And he answered it really honestly and rather well. He said, there's always room to compete. With DeepL, it's vertical integration. Businesses are super confused about how to adopt AI, and that's the opportunity. The opportunity is to produce a solution for a particular use case. And again, I thought, great category focus.
[00:09:52] Paul: Let's not forget, these guys that seem to be steaming ahead, Microsoft, Google, Meta, On AI, they are going to face a few regulatory hurdles.
[00:10:00] Jonathan: That was one of the points that Jarrett made again and again and again, which is that you've got to evolve all the time. And as a founder, you've got to think about it 24 7.
There's simply no other way. And when you're doing that and having to You know, maybe deal with issues such as evolving regulations. You've got to realize that you, as a founder play to strength and you need a complete team so others can compliment those strengths and all part of going into the unknown is that you need to work with people from different backgrounds that you're not used to working to, so be prepared not to know and to learn.
Even if what people are doing and saying is really alien to you. The final question, which I loved and obviously Eric hated was, where will the company be in five years? He was gracious to say, you know, why he hates the question. It's because it's so hard to answer, given the nature of the AI business. But what he was prepared to predict was that the business would continue.
To be research heavy and it will continue to focus on specific applications and it will need to become ever better educating customers. And that means having an even better sales force that can do it. And even that they may have to completely rethink the company to remain competitive. What was really nice was that they actually asked.
Are there DeepL users in the audience? And I reckon about a third of the audience were actually DeepL users. And this wasn't a load of people that had been bused in to look good. And what was really interesting is somebody shouted out, It's way better than Google. So that's, that's the sort of traction they're getting.
You know, let's hope DeepL can avoid acquisition and go all the way and be Germany's second
[00:11:44] Paul: SAP. Does sound as though we've got Uh, potentially Europe's category leader.
All right. So we'd like to tell you things ahead of time. We're very future looking on this part. And, uh, we'd like you to cast your mind back to February 21st of this year, episode 15, go find it in the backlinks and, uh, let's just play the clip.
[00:12:10] Anuj: I predict that later this summer. Once the presidential race has been set and Trump has been appointed as the Republican candidate, deep in, Biden's going to announce very suddenly that he's not running and they're going to put in a younger Democrat, who is very electable, to kind of switch up the game.
So once they've already set this old, you know, battle of the oldies, they're going to come up with a vibrant young candidate, which is going to switch the race on its head.
[00:12:35] Jonathan: So that was our esteemed guest Anuj Nair, and what was he saying? He was predicting there'd be a change of democratic candidate for the US presidency six months before it actually happened.
[00:12:48] Paul: You heard it here first.
[00:12:54] Jonathan: What does the future hold? Let's look into our crystal ball. So Paul, what have you been
[00:13:02] Paul: seeing in your crystal ball this week? Oh, the vapours, the vapours are lifting. I'm getting some clarity. And I want to talk about the not so originally branded AI launch, much feted AI launch by our friends at Apple, with something that they've called Apple Intelligence.
[00:13:17] Jonathan: Oh, that's sort of interesting because hopefully listeners will have listened to our, uh, quite long piece about the history of Apple and whether it is the end of its category creation journey. Will, will this help? Do you
[00:13:28] Paul: think? I'm not going to commit to that. We know somebody, um, who's very closely connected with Apple's history, let's just say, who thinks it's, uh, this is a very good idea.
I say the jury's out a little bit. It is a move to improve Apple's relevancy in the AI revolution. And others have thrived from that. So IBM's got a bit more of a spring in its step. NVIDIA, many thought was, um, was out. But now, uh, the GOAT, greatest of all time, category king, uh, Apple has been derided and is now making its moves.
And I don't know, uh, to your point, if this is going to reinvent, reinvigorate the company.
[00:14:03] Jonathan: If we know anything about Apple's history, we know it's a great fast follower. So is it about to repeat that rather good piece of history? It would be nice
[00:14:10] Paul: to think so. Because I think the guys at Cupertino said, Not so fast when everybody wrote them off.
And now they've got, um, the crowbarring in the tech from Microsoft, courtesy of the OpenAI, uh, move, and that's all very new and very exciting. And they might think, we're all set for now.
[00:14:26] Jonathan: So where does Elon Musk sit in all of this? Well, I mean, he's just He helped seed fund, uh, OpenAI. He did
[00:14:32] Paul: indeed, he did indeed, uh, you know, came back to haunt him a little bit, and he's, uh, certainly not pleased about the Apple intelligence move, because he's, he's, he's Literally said that you bring an iPhone to any of his workplaces, SpaceX, Tesla, or even Twitter.
Let's remember they got to stay outside if this goes through and they will be unconnected from the internet in a nice Faraday cage doing it's doing its own thing. But, you know, Elon's just had a nice little payday, so God knows where he could take things if he wanted to.
[00:15:00] Jonathan: But it's usually it's been a bit punchy, hasn't it?
He's been putting his opinions out on X. I think he said something along the lines of. Apple has no clue what's actually going on once they hand over your data to OpenAI. If that wasn't bad enough, he finished it by saying they're selling you down the river.
[00:15:21] Paul: Wow. So, uh, Elon, the customer advocate, it's interesting, uh, famously a guy who does his own thing.
Now he's, you know, all on the side of us all. Great. Good to see. Now, other commentators have talked about this privacy issue with, um, Apple Intelligence. I'm just. To characterize it from what we know so far, this means that all of your work and thoughts, wishes and dreams that you put into your device or things that maybe allegedly you utter within earshot of the device, that's going to be put into some sort of AI.
And, uh, act as a personal assistant with clear potential issues around privacy. Other commentators have been generous about this. They don't see it as an issue. And I am on the fence. I was very cynical about, you know, people using Dropbox and, and, and cloud storage and the issues around that. They, those issues seem to have receded a little bit.
People have got used to it. And I can really clearly recall Aaron Levy, uh, the CEO of Box, making the claim when asked about what people are doing around data privacy with cloud storage products, saying the definition. Or privacy will change.
[00:16:22] Jonathan: It's, it's interesting how, where human beings find value, uh, changes.
I mean, it's not that long ago that if you said to somebody, Oh, I'm going to buy a product through my computer from somebody I've never met, and I won't see the product until it actually arrives. They'd think you were absolutely mad. Now they think you're absolutely mad not to do it.
[00:16:39] Paul: Pretty interesting how these definitions change over time.
So therefore, I am looking in my crystal ball and thinking that Apple might just have a shot of redefining privacy once again. With this Apple intelligence.
[00:16:52] Jonathan: It may be showing it's very deft understanding of what the consumer wants and why the consumer sees value.
You know what? This really grinds my gears. So Paul, what's
[00:17:14] Paul: been grinding your gears this week? Well, London tech week. What is the point of London tech week? It's been a feature in London for what seems like for a decade. ever. And however noble its intentions, however fevered its sponsors become, however frothy the BBC and others become about this, I find it a dull non event and it's been attended by all the hangers on, the tech tourists and wannabes we don't need in tech.
So London Tech Week was originally associated with something equally dumbass, which was the Silicon Roundabout attempt to rival Silicon Valley. As if, right? And that Silicon Roundabout, which was Old Street Roundabout, I believe, has now been magically transformed over to the hideous carbuncle that is, uh, Westfield Stratford, and with Mr.
Khan's white elephant Tech City East over there, it really is just a complete white elephant. And back in the go go days before 2021, and things, the wheels started to come off a little bit in tech, the parties at London Tech Week were, uh, Exclusive and were fated. We had celebrity attendees. We had flowing refreshments.
We had, have you got a ticket? It's really hard to get. I can help you out. And the sponsors were legendary. However, so we know celebrities aren't everything, but let's zoom forward to this year. Let's see what exciting celebs and important people from the industry we've got this year. And I have. Just the four mentioned, Nicola Hodgson, bless her, chief executive of IBM UK and Ireland.
A bit of corporate capture there, I'd say. That's a bit of corporate capture for me. Um, Claire Barkley, CEO of Microsoft UK. Oh, save a thing, I think. A little bit of corporate capture there. These are not, uh, brilliant companies though they are, amazing legends. They are not the cutting edge. Kate Ryder, founder and CEO of Maven.
Who? And then, uh, John Serafini, CEO of HoneyBook. Um, sorry, John. Who? Yeah. So, you know, obviously, even judging by the caliber and the sort of makeup of the folks that are there, this is not, on the face of it, the most exciting event.
[00:19:11] Jonathan: It does seem to be a sea of civil servants. As usual. And government funded pressure bodies of various sorts, it seems.
[00:19:17] Paul: Indeed. Yeah, a sea of civil servants. That's a decent group
[00:19:19] Jonathan: now, I think. What I'd noticed this year was just, The terrible lack of ambition that seemed to be there, you know, when, when Nicola Hodgson did her opening speech, it was the usual place for government to help drive economic growth through inward investment.
I think in this case, I bung some money to IBM so they can agree to do more AI research in the UK rather than somewhere else. Um, help startups become scale ups. This was the classic and picking some winners in terms of technology like. AI, or maybe quantum. Do you know what? This internet might take off.
It might.
[00:19:55] Paul: And what's this got to do with the government anyway? I mean, these days, London Tech Week is actually owned by, um, billionaire Mike Danson's, um, mega publisher, Informa Group. So it's a properly commercial event. Yeah. That might be good. So maybe there's less, there's less, um, public funds being thrown at it.
Yeah. Why is it called London Tech Week? Yeah. This is not owned by London. This is owned by a private quoted company. Why isn't it called? Informer Tech Week. Yeah, a bit more honest, wouldn't it? Well,
[00:20:21] Jonathan: frankly, even UK Tech Week would be a little better bit better, given that we do actually have other clusters in London.
[00:20:27] Paul: Yeah, there are, believe it or not, other significant tech clusters are available. Oxford, Cambridge, Manchester, Bristol, and other cities. Why is London getting all the tech plaudits? And the irony was not lost on the BBC's tech correspondent Zoe Kleinman. She bemoaned packing for a long rail trip, taking away from a family family up in Glasgow, where she and others from the BBC were.
Encouraged to relocate and let's just say not all of them made it So, you know Why would we have it in London call it London Tech Week when it's a private tech event? With very little to do with London and London itself may not be the heart of the UK's tech industry I
[00:21:02] Jonathan: think there's a couple of real points that come out of this and I think hopefully listeners you're sensing We're a little disappointed about about this is that you know We we mentioned we'd been to Sester a few weeks ago and It seems that London Tech Week compared to the likes of Saster is actually irrelevant to the actual process of building tech business.
You really liked Saster. Why did you like it so much? Because it was just, it was so well organized and I've never seen such, such a structure that allows all the components of the ecosystem that are actually taking risks, that are actually putting money in, that are actually pushing things forward, get together.
The level of energy. was great. There was no words wasted. Everything was actually about doing business, not making policy, which seems to be the aim of London. You even
[00:21:53] Paul: liked the environs, the logistics. I mean, poor old Zoe. She was a little myth. There was no carpet on the show floor. They hadn't bothered with that.
Uh, times must be tough. And I have it on good authority. I was speaking to a close friend of mine, heads up a software services company up. He did walk to the floor. There was, uh, empty stands for things like tech Malta and several other companies obviously been sold exhibition stands thinking this thing was going to be throbbing and clearly wasn't.
And I have on good authority, the Tesco event, probably a couple of hundred seats, 40 people showed up. I mean, this thing is not what we want. Is this a meaningful show?
[00:22:29] Jonathan: I just don't, I don't see, um, the way London Tech Week has evolved as, as, as being in any way instrumental in helping us create the next generation of category leaders.
And the other thing I thought was a crying shame, um, given that what we think is London Tech Week's actual lack of ambition was while they were doing this, Meanwhile, in Dubai, you know, a new emerging economy, they just announced that they were appointing 22 chief artificial intelligence officers right across government.
Basically every government department serving citizens of Dubai will have a chief artificial intelligence officer, you know, just let that sink in virtually everybody charged with delivering public services and an entire country will have someone whose sole responsibility is to drive. AI. And that is sitting in a vision for future government work in this thing they call the Dubai Universal Blueprint for Artificial Intelligence, you know, DUB.
AI. And it
[00:23:31] Paul: sounds exciting, right?
[00:23:32] Jonathan: Yeah, I mean, it's ambitious. I mean, they're trying to cement their position as a global hub for AI governance and legislation.
[00:23:39] Paul: So the French have the tech show Paris, which is just two days, not a week, and it's not a week of after parties, right? It's, it's, it's Macron, it's, it's proper, it's beating the drum for France.
The Irish, okay. They lost web summit, but they lost it to the Portuguese. They're delighted. That thing's still rocking. The Catalan's got Barcelona, obviously, mobile world conference, four years, really, really focused on the industry, really providing value, still a highlight for most telco companies. And we've got this rudderless, sidelined, frankly, slightly sad and an overblown sideshow.
Yeah,
[00:24:11] Jonathan: I think. We wait over the next few weeks to see if there'll be any announcements about inward investment commitments that come out as a result of this, as the French managed to do very nicely recently. But I think London Tech Week needs a big rethink at
[00:24:27] Paul: best. It really does. We could do much better than this.
We don't want to be down on Brits. We definitely don't want to be down on British tech. But please, please, please, can someone rethink What the heck London Tech Week is all about.
Thank you for listening. If you want to learn more, go check our blog posts on becategorical. com.
[00:24:47] Jonathan: If you have a category issue, then we can help. Get in touch with us.
[00:24:50] Paul: And remember, don't be better, be different.