E395 | This Week in European Tech with Dan Bowyer, Mads Jensen, Lomax Ward and Andrew J. Scott

0:00Okay, welcome to Upside. I'm Dan Boya and with Mads Jensen, we're both VCs at Superseed where we back early stage technical founders, transforming how the world works. Yeah, we meet every week to unpack what's happening at the intersection of tech, venture, and business in Europe. So whether you're a founder or investor or really just curious about the forces driving our future, you're in the right place. Just remember, this is not investment advice.

0:28Tear down this wall. It's more than just an alliance. This is a union of values. Let's start acting. This show is not investment advice and the hosts of this episode may be invested in the funds and companies featured. And this week we are looking at 2025. We've got two super chaps with us. We've got Andrew from 7 % and Lomax from Outsized. And between us we're going to crystal ball. We're going to have a look and have a ponder. Hopefully some of the things that aren't so popular. A bit more controversial perhaps. and have a look behind the scenes at what is going on in venture startup with a focus on Europe.

1:12And to start with, we're going to start with one of mine, which is I reckon Apple is going to win AI on the consumer side. I reckon they're the dark horse. And the reason I say that, just to unpack this a little bit, is that they've been at AI for a while. I remember talking to a

1:35Obviously, this is pre-chat GPT. And they were very, very mindful of what was going on. And they're always so good at being the last to market with these things. And because they have such an incredible ecosystem and they own that ecosystem and it's not fragmented like Android. And it's so interconnected with across your Macs, across your iPads and across your iPhone. I think with such a loyal fan base, with so much money, so much resource, they brought over the head of Google last year. to run the AI department. 87 % of US teens own an iPhone. I think they might well crack it. Now, Mads, you're not a massive Apple fanboy.

2:17What do you reckon? No, I don't mind. I just think it's a while since they've been really innovative. And I think your argument is, look, there's not really innovation that's needed here. It's execution. We got all the component parts. They can stitch together stuff from elsewhere and then integrate it beautifully yes they often do i saw just as a side gig i saw their vision pro now they're discontinuing production yes it's quite quick so that was that was always on the on the writing was on the wall or the eyeballs were on the wall yeah i just but i just think it's symptomatic of a company that's lost its way a little bit in terms of they used to be innovative and it seems like they've lost some of that and i think your argument i think jobs he was innovating in the way he looked at the world?

2:59I think so. But I mean, you could argue there's no innovation. Everything is just stitching together different things, you know, old things in new ways. I don't know. I mean, I'd say Apple was a hugely innovative company, but I just want to ask me, what are we talking about here, Dan? Is that what do we mean by, you know, wins consumer AI? I'd love to actually unpack that a bit because when I thought about it, when I think about it, I'm like, is this a winner takes all? What are we actually talking about? I mean, this device will probably sort of remain this device and, you know, the installed base of it will continue to grow, I'm sure.

3:29But what do we mean by winning consumer AI? For many people, I think Siri and voice will become useful. And the back thinking was around, because they own the ecosystem, if they get this right, which I think they have a fighting chance of doing with all the resource and with their partnership with OpenAI, if you can use voice to command apps on your phone to interact with other apps because they control the entire ecosystem Apple so my thinking is that it will become almost like chat GPT has become the kind of the known consumer standard I think Siri will become the known consumer standard for AI in the consumer world that's what I meant by Apple will win consumer AI so I think that they will own Apple intelligence which i hated the name of when i first saw it and i think they will end up owning it and i think they will end up i hope and i think siri will be useful i mads always shouts me for having wishful thinking and by i think there are enough data points here there there's enough infrastructural data points here where i think they can if they get this right and let everyone else fight on the foundational side i mean we'll talk about you know we'll open ai take it to the finish line we'll talk about that later but i think on the consumer side i think Apple will, I think Apple could become the de facto standard for how consumers interact with AI just by using voice.

4:57One thing I will say is I think it's a really good point is that the modality is I think going to change. So we're looking at, we've been focused last year on text. I think voice will be the modality that everyone focuses on and will become the use case, the most prevalent use case. I also think, I know we talked about vision probing discontinued and there's actually a chance that oculus might get discontinued but i still think we're moving to a different form factor of device and voice is going to be so important for mixed reality so actually i think the kind of the trend is in in that direction and so maybe maybe you're right dan in that case i thought the question of is apple an innovative company was a really interesting one because what is innovation i don't think innovation is the same as invention i mean apple did not invent the first mp3 player uh there were a bunch of mp3 players before the ipod came along yeah if you ask well i mean if you ask the average person who invented the portable music player they probably will say oh apple it was the ipod because that's the one we remember so for me innovations a collision of the right timing with a product which is sufficiently better that you know it grabs the imagination and that feels like innovation so i think it's too early with apple um and sorry with ai to say who who's going to win because dan i'm what i'm hearing for you dan is that basically apple will win because they've got access to the customer and they own the customer yet iphone sales are flattening and to your point mads about you know the company in general i mean the apple car the apple car is nowhere to be seen the vision pro definitely didn't take off you know yes they've got a market cap of like what is it three three four trillion so they're not going anywhere but i think the vision pro really suffered from a lack of killer app and i think i think the last stats i saw was like it got like three percent market share in vr so they've definitely failed last year to come out with anything which is could be called a win in the market perhaps that's a case of the company trying to do too much and they need to start to decide like what's really important are they going to try and win ai are they going to try and defend their iphone market share is there still a car being built somewhere we don't know about i don't think they're going anywhere because there's so much intelligence inside that company right there's so much such smart people work there it's way way too early to say that apple or open ai will win the the ai race this is the i'm talking consumer here i'm not i'm not talking business i'm talking consumer but that but that differential becomes less and less significant right the sort of socialization of enterprise software um even the way now you capture enterprise users if you think about slack and other people where you go bottom up you capture the individual you know which is essentially a d2c sale um and then you you go upwards into the enterprise so i'm not sure whether that differentiation even makes much sense anymore i think with apple i think with apple it might do i don't think apple is ever going to play that game okay so next up we've OpenAI will be dethroned as the AI leader.

8:00Mads, I think this is one of yours. Do you want to tee it up? Look, they clearly came first out of the gate two years ago, launching ChatGPT. And I think they've had a phenomenal run and built up a really, really impressive business, sporting billions in annualized revenue now, growing very fast and having the vast majority of the consumer market, you know, back to consumers you were talking about before on the Apple piece. I mean, it's squarely in OpenAI's and ChatGPT's hands right now. What we saw last year was Google having a bit of a false start. In January, they came out with a Gemini that was so riddled with bucks that it was unusable.

8:36But they've really doubled down on AI now. They've got the whole founding team back in the lab. They've got the whole company orientated towards getting back to that leadership position in AI. And their models are really impressive. Gemini 2.0 is really impressive. You've got Elon Musk out-executing over at XAI. I think what he's done with Colossus is really, really impressive. You've got Anthropic building some super, super, just wonderful models. I think if you look even today, whether it comes to coding or whether it comes to prose and writing, the Claude model definitely beats the OpenAI models.

9:14Maybe in everything except for reasoning, where I think it's probably Google and OpenAI head-to-head today. But it just looks like OpenAI is being a sale from so many sides. I think also given the turmoil they've had in the company, it's going to be hard for them to sustain the leadership they build up. So I'm expecting that they will be, if not dethroned, then certainly assailed from many angles. One aspect here that's really interesting is what we saw here just at Christmastime, which was the launch of DeepSeek 3, which is this Chinese open source model that was released to the world. And effectively, it's a model that's been built on a fraction of the trading cost of the big Western LLMs.

10:00So whether it's Lama or GPT or Claude or Gemini, it's interesting that the Chinese team was able to build an open source model so inexpensively because it kind of changes the paradigm a little bit. And I don't think everybody's really figured out exactly what the implications are yet, whether we can borrow some of the innovation there. But what it certainly does suggest is this attempt there's been to hold China back in the AI race by restricting the export of chips has to some extent failed because they just found ways to out-innovate and create an amazing model even without that silicon. So super, super interesting angle to the development here.

10:43Before I open up to the floor, what does dethroned mean to you? Well, so today, OpenAI has got the leading share in market, and they've had the best models. And I think those are the two vectors, right? You can have the best technology, and you can have the most share. And I think certainly when it comes to models, I think they're going to be under attack, and I think they're going to be relegated to second place or third place. And market share, if not a catch-up, then there will be others eating into their share. I think Anthropic already is doing extremely well when you look at AIs and API workloads.

11:18I also wonder where we're at on kind of peak LLM. But Andrew, what do you reckon? My sense is you'll agree with this. Yeah, I don't know if it will be this year, but I think them winning is unlikely. There's a few reasons for that. First of all, I think you bet against Elon at your peril. And I'd probably bet against Elon part-time versus Sam full-time. And I don't know, Sam, but just outside looking in, I think any company that has that sort of leadership turmoil, when you're dealing in this amount of money and this sort of pace of change, you've got to question whether it can go the distance.

11:55Because there's sort of three fights going on here. One is it's a fundraising war. At the moment, in order to pursue the path that those Western LLMs are taking, you need vast amounts of money. So who can raise the most money? Secondly, it's a technology and talent war. So we've seen Marati left. John Shulman went to Anthropic, one of the early OpenAI guys. So there are people leaving a company. That's never a good sign for OpenAI either. So the smartest people inventing the smartest technology because this technology is super early still. I don't believe you're going to get AGI from the current Transformer model stack.

12:32So I almost feel like we're in an interim phase technology-wise. It's very early doors with this next-gen of AI. and that could at any point be usurped by the next great thing that comes out. And the third part of that is there could be an enabling technology that comes out at any moment, which means that suddenly training is a thousandth of the cost it is today. So that could be one of the new chip technologies, reversible computing, or one of the other low-energy silicon tech coming out. So I think it's just so early in the growth of both, not just the technology, but the whole industry itself.

13:15And it could go anyway. So I think OpenAI, in summary, is sitting on too much cash that it will be automatically dethroned in this year, 2025. But in the next few years, I think we'll just see some real, real left of field entrance to the market or different approaches or step changes in technology, either enabled by new silicon or just new AI models, but also because at the moment, the current funding is... We're in a completely new paradigm for valuations of companies so early and the amount of money they need. If you look back at other companies in the past, very few of them have approached these valuations at all.

13:59I mean, even now, you're talking about You've got SpaceX, I think SpaceX is now valued at north of 350 billion. You've got Palantir IPO at 20 billion, ByteDance, which is the backer of TikTok, I think it's a few hundred billion. You've got Stripe at 100 billion. So the numbers we're talking about and therefore even the model for investment returns that OpenAI and XAI and others are going to have to deliver on. It's just completely new territory. We've not been here before. So I think from an investment point of view, there's a question mark about how long people are going to carry on throwing$7 billion just for training and inference, which is what it's cost to OpenAI.

14:38But at least they're making revenue. A lot of the bubbles, there was not as much revenue, at least some of these big foundational companies. Yeah, but the multiples are incredibly high. You're looking at a public sector multiple environment that's as peak as it ever has been, especially in tech. Does anyone know what the unit economics are on actually serving a query from OpenAI? It's a million pounds a day to run. A million dollars a day to run. Microsoft is saying they're pretty good in excess of 50%. Go ahead, Marge. Okay. Delomax, dethrone? Dethrone, not dethrone? I mean, I don't have a lot to add to what's William said.

15:14Other than I would say it's like, for the moment, the market is growing at such a rate that actually fighting over your market share is probably less relevant you know it's great to have the you know ego trip or the halo effect of being number one but again i don't think this is necessarily when it takes all for the moment at least you know we're talking about the next 12 months there's enough for everyone to fight over that yeah i mean it may lose its number one spot but does that really matter for open ai at the moment i think andrew's got a good point it's like is open ai is an organization set up set up for success in the same way that its competitors are and i think it sounds like probably not which is more of a long-term kind of infrastructural problem that they have i think but isn't there isn't there always conflict isn't there especially with changing from this this whole uh public serving to full profit i mean there's gonna be conflict well the aid is always conflict for sure right in all of this but b i'm actually in a way the open AI model of moving from what it was to what it is, not many people have actually successfully done that.

16:18They're not closing the not-for-profit side. They're not closing that. That's the point they're making is if you look at all of the big companies that have been created the last 20, 30 years, by and large, most of them are founder-led, right? Properly founder-led. And I'd argue that might not be the case for open AI. Yeah, no, I think it's fair. I think it's fair. I want to stay with China. And Andrew, one of your points, you put into the docket that this will be the year that decides what happens with China and what happens with going to war over Taiwan or what happens with trade conflicts and other challenges between the East and West.

16:51What were you thinking here? Yeah, I really just think a failure to defend Ukraine in the next 12 months is going to give a green light or not to countries with ideological goals, right? And the assumption is that Trump won't back Ukraine. Is that the assumption? I'm not going to assume anything about Trump. I think he's proven to be hard to predict, which sometimes plays to benefit in foreign policy making. But I just think fundamentally, if we show weakness towards Russia, then that does indicate to other countries that we don't have the stomach for it. And that's how we've got to where we are today.

17:27By saying nothing about Crimea gave him the confidence that we wouldn't care enough to do anything about Ukraine either. So that's absolutely critical because China's watching and other countries are watching as well. North Korea is watching. So that has a knock on effect to not just China, but I think also general geopolitical security. I mean, you've seen attacks on, there's a lot of asymmetric warfare going on. And a lot of it's been going on for a very long time, right? I mean, China attacks the UK infrastructure hundreds of thousands of times a day with hackers and AI bots and tries to probe, tries to steal secrets, tries to shut things down.

18:08Can you see any world where Taiwan has not been absorbed by China within a decade? I think it comes down to whether America wants to maintain its authority of the world or not. Because they've made it very clear they will defend Taiwan. doesn't mean they will. I thought it was always a point of strategic ambiguity. The Chinese government has just put in an order for one million attack drones. You know, China's manufacturing capacity is just totally unrivaled. And these drones completely change the nature of warfare and in some respects renders the American naval supremacy irrelevant because you can sink the carriers with all these attack drones.

18:53But let's look at what that does longer term. So let's say that China in some way attacks Taiwan. America does not respond. What does that say to the rest of the world, not just China, but every other country, but also China itself? If China decides actually, well, that was quite easy. And they're expansionists. They're ideologically expansionists, the greater China. They don't want the power rival status quo that exists in the world today. So that message is that, well, do what you want, basically. I'm not saying it's a good thing. I'm just asking whether you can see a scenario where it doesn't happen.

19:23I'm struggling to see the U.S. facing – I really hope not. Listen, I don't disagree with you, but I'm just struggling to see a scenario where the U.S. faced with China being as strong as it is today on the manufacturing base. So the West won the Second World War on the back of the American manufacturing base. That's been eroded. China is supremely number one today. I just don't see how the U.S. – maybe the U.S. could win, but you'd have to go to such an extent. The cost would be so high. But it always is in any great power conflict. And that's the biggest danger. That's why deterrent is so important.

19:59That's why spending the gold now rather than later, which would be gold and blood, is absolutely vital. I wouldn't underestimate the ability of America to step up and put all that industrial might into a war footing. And I still wouldn't bet against the United States. And also, warfare is not just about the ability to produce drones. you've then got the ability to run a military and run military operations. You know, China is not anywhere near as experienced as actually, ironically, the West in actually fighting wars. One of the reasons North Korea is putting men on the ground in Ukraine is so it gets some war fighting experience.

20:40You know, it hasn't fought a war in, what, 60, 70 years, right? Men haven't seen armed combat. So all the enormous infrastructure that backs up actually just being able to deploy a drone, from the avionics, from the integrated communications, all that stuff has to be exceptional to win. And I don't think actually China's at that point. There are areas where it's strong, but you're right that their advantage is overwhelming numbers. There's also a question of proximity. I think if the argument is whether China will be able to invade and hold California or even Hawaii, I 100 % agree with you. I just don't think the U.S.

21:19will want to go to the lengths you are describing and the country would have to do to hold and defend Taiwan against China. I just I don't see it happening. And I think the determination on Xi's part to absorb Taiwan eclipses everything else. And you see, the Chinese economy is really not doing great. And his population is starting to suffer. I mean, there are many, many warning signs. The bond yields have fallen off a cliff. The domestic economy is struggling in many ways. But I think he's so determined to do this. And I think at the end of the day, power accrues to those who want it the most.

21:56I think Xi wants Taiwan to come, you know, be reunified with or be absorbed by China. I think vastly eclipses the US's desire to defend Taiwan. And therefore, I think it will happen. I think that's my forecast. But I agree with everything you're saying. I don't disagree with any of it. I just don't think the US is going to step up in the way that would be required. Well, two quick things. One is that Trump doesn't want to be world police. I think that's fairly, I don't think he'll turn coat on that. And two. I think Trump says that, by the way. But I actually think he won't necessarily act in that way.

22:29But yeah, okay. Well, I think I'm not convinced that he wants to be, you know, world police. And also, secondly, once the chip manufacturing really comes online in the States, what would be the commercial need to protect Taiwan? We've already seen the plant, the TSMC plant in Arizona, now producing the highest quality chips. So America has got the production capacity. So what would be the strategic need to stop China from? But as we've seen with their models, maybe they don't need them. I mean, I don't know. I mean, we will never. I think I think this will come down to like, how does the US and without, you know, in a way a little bit out of our our our lane to this in a sense.

23:10But you know, this impact so importantly, so much on on what we do as a day job, but you can't ignore it is that if you a bit like a bit like Putin in the Crimea, if you can Taiwan on a plate to Xi, then you're basically handling the whole of Southeast Asia to have one of the You know, they're already saber-rattling at the Philippines with Vietnam, etc. I think it's a message. If you don't, if the U.S. does nothing in respect of Taiwan, I think that says the message that the rest of Southeast Asia is open for it. Are you talking about what you think should happen or what you think will happen?

23:44I think that's what happens if the U.S. decides to not do anything about Taiwan. So should you think the U.S. would put American Marines on the ground in Taiwan to defend Taiwan against Chinese aggression? It's a very good question. Depends how the conflict starts. So to imagine your question around like why would the U.S. or whether the U.S. would defend against Taiwan? I mean, it's not just about the fabrication plant. I think, Dan, that's a really good point that that decreases the most obvious reasons for protection technology point of view. But you've also got to think about trade routes, freedom of navigation with the Strait of Taiwan, one of the busiest maritime routes that would put China in control of that.

24:28Japan, South Korea, the Philippines, all of them would be terrified at the prospect of China having taken over Taiwan and unopposed because as an expansionist power, they're some of the obvious next in line targets, right? So one would really hope that we've learned our lesson by now that you don't let a capable, large, powerful country evade, you know, evade unopposed, you know, as a first step of its expansionist policy. And the problem with both Russia and China is that they are an expansionist country, but they're also insecure countries. So on the one hand, you want to show enough deterrence that you demonstrate the will and capacity to perhaps make them think twice about invading a ukraine or b taiwan but on the flip side because they're both insecure psychologically as well in that china thinks everyone's out to get it and and putin certainly does and we showed some signs of that unintentionally or not with nato if you show too much strength then actually that makes them more paranoid and then they they do something like invade ukraine anyway so they're very different difficult states from a statecraft point of view to actually manage there's an interesting point you raised mad when you asked me is that would would u.s put range on the ground you know you know i think like we need i think we need to start thinking about warfarin in a different way the u.s can deliver comprehensive like very impactful support to taiwan without uh putting boots on the ground like it's particularly with modern weapons and you know it's actually takes me to a point i was thinking about over the over the christmas break which was like there is an element of the world that feels very like 1914 at the moment with all of the different kind of insecurities that we have everywhere.

26:12But that also feels like a little bit like 1914 in a technological standpoint, which is very interesting. It's particularly tied into defense tech and what we're seeing, which is that in 1914, you only just had the first of the industrial wars happen. You had a little bit of the industrialization of warfare in the US war. Then you had the Russia-Japanese war in 1905, some war. 1914 was when really the impact of the industrialization had a massive impact on the way we fought wars you know you still had the french charging you know in 1914 you know still charging horses into battle right and i still but i feel like we are analogously at that kind of point in time now with warfare right you know where everything is becoming more automated you know more drones you know automated ships etc so i think that's a very interesting like infection and then ties into, I think, our world of investing in the banks.

27:09But I think that maybe answers your question, Mads, is that there's a lot the US can do without putting boots on the ground to deter China. Where else did you want to go, Lomax, with your defense tech surge? Because that was one of your points. There's going to be a big 2025 push. Yeah, we've seen the first cohort and wave of new wannabe defense primes comes through like Enduro in the US and Helsing in Europe and you now have Sironic that raised a massive series C I think last year in the US on the naval side we've seen that first wave I think that will continue and I think we'll have like more maybe vertical specific problem specific companies like coming up raising money if you look 10-15 years ago none of these countries apart from the US were really in you know western countries were investing in defense you know the percentage of GDP was tiny they'd all whittled away their defense budgets to nearly nothing now they decided to up this and they're now meeting their two percent NATO threshold and you know countries like Poland are going way in above that right those decisions will kind of stay in place that's policy now for the next five ten years I think yeah no one's going to touch it which means that you know our people that we're backing and working with who are starting companies now have a kind of long runway to build companies and actually have a kind of winning and cashed up client base to fund the innovation that will be needed in the private sector so i think i you know i'm bullish on the sector as it were despite the difficulties of public sector procurement etc etc etc just got a question do you think the vc model is appropriate for defense tech companies though because there have been so few trade sales or ipos and a lot of these companies are not you know there's only going to be one or two new primes right and therefore how do you get the returns with the traditional vc model for companies that produce something that might be really important but he's he's either a subsystem of a larger system that you know a prime produces or is you know a small piece of tech that you know okay the troops on the ground want but isn't going to give you a multi-billion dollar outcome it's a great point and not not to patronize you because you know this better than me but like let's not be backwards working in a force industry right so let's be like okay yeah you're right it is not necessarily traditional we're not going to have like tons of vertical like sas companies right in this in this field there's a couple of nice things i'd say is one is there's tons of because because you know a lot of these companies are building hardware like difficult things there's tons of non-dilutive funding big 100 million government contracts right available which mean actually the dilution that you might be worried about that problem might not be as bad as people might think but to answer your question i think you're right like i think that is and we think about this when we look at sort of precedence sage defense companies i suggest there will be consolidation there will only be like this will be a little bit more winner takes all unlike you know what we talked about earlier but you know what like i mean i don't know about you guys i've done this tons of times where i where companies were very vertical specific they've actually consolidated and teamed up with another team and you you know you can buy and you receive shares in in what will end up becoming the winning company i still think you can make your 50 hundred extra turns like that you know so i still think that the consolidation as long as you're not overpaying for the shares that you're receiving in these deals so in a way as long as the company is performing such that they have bargaining power when they do a share-for-share exchange and actually consolidate, I still think you end up with the potential for a decent ownership of that new prime.

30:42There's also the dual-use aspect. I mean, it's going to be such a tech push. As you already alluded to, Lomax, the next warfare that we see will be very, very technically driven. And a lot of the stuff that we do is already dual-use. But Mads, have you got any thoughts on the defense piece? You both have really good points. I think you are looking at a sector which will see massive secular growth just by virtue of the fact that defense spending as a percentage of GDP has been very low between 1 % and 2 % and it's going to increase to between 2 % and 3 % and 3 % and 4%. So you're going to see at least a doubling of the market size here over the next few years.

31:19And that's billions of dollars that will flow into some new tech areas. If it just goes to the traditional existing defense primes and nowhere else, you're not going to see the innovation we need. So I think my concern about it is it will be politically driven. And we know that politicians don't always make the best decisions when it comes to sort of being forward looking and thinking about things in a tech orientated way. But if things are done right, there'll be a great opportunity here. And, you know, I think, Andrew, that's your point. It's just it's the risk of the of the segment of the ecosystem.

31:50It's also important to remember that the defense is not just bullets and bombs and attack drones. It's resilience of critical infrastructure. It's all the cyber security. It's offensive AI. Energy. And that's a far bigger sector than just the kinetic weapons that you'd automatically think about. It feels like the whole defense push will also push up the agenda, manufacturing, supply, that whole protectionism, nationalism, modern language set will create more opportunities for countries and therefore startups to support their countries. I think it's worth saying that the UK and Europe are grossly behind, particularly Europe.

32:37I mean, there's a fundamental ideological change that needs to happen. At the moment, a lot of VCs can't invest because their LPs, the EIF, still have mandates within their legals, which means they can't invest anything that even slightly smells like something military or something even to do with defense. That's a big problem. That needs changing before a lot of VC firms can invest in a lot of these companies in Europe. You've then got an institutionalized dislike of hardware. everybody you know i was in vienna an event a really really great day that speed invest ran a room full of you know 30 gps pretty much all of them bar two won't touch hardware oh hardware takes far too much capital oh hardware is really difficult you know they're just europe's just so behind mentally well that's a that's a question of fun size isn't it isn't that isn't that a factor because actually because actually the idea that today hardware automatically take more money is not necessarily true.