🔴 LIVE: Finnish President Stubb & EU Tech Chief Virkkunen at Helsinki Security Forum | AC1E

That was a really interesting illustration of the way the two theaters come together as that sea route opens. But Jana, Germany, I know you're not here to represent Germany per se, but Germany is a massive global trading nation, enormous manufacturing base that relies on exports, and China is a hugely important market. How will the chancellery feel if a future US administration starts to demand that the return for US investment in European security is a more US-facing trade policy?

You know, I think I would turn this around. Why should we be more critical with China or more hawkish towards China? Not because the United States asks us to be on their side, but because we have an own interest to see China more clear-eyed, to reduce our dependencies. If we are serious that this war against Ukraine is not only about Ukraine but about the European security order more broadly, then China is the central enabler of this war. China makes this war possible on a daily basis. So China is working actively against us and against our interest, and we have not made China feel this way. I think in that sense Donald Trump had a point when he was asking the Europeans for more sanctions. I think we have not started an honest discussion about how should we deal with a China that is actually enabling Russia's war and what can we do against it, and we should think about sanctions because it's in our interest, because it's our security, and not to do a favor to the United States.

Well put. Issa, I want to come to you with the last question of all. We've only got about a minute left. There are people in the UK MOD and in fact in NATO headquarters who are using 2028, 2029 as a date by which Europe NATO must be as ready as possible. Now I think that that is just a forcing function. I don't think there's intelligence to say that war is coming that soon. But as a chief executive of a major defense company, can you honestly tell us that Europe has the industrial capacity to be ready on that timeline?

Not totally. Not in every area. There are pockets where that level is already there, but in total, I would say that the clock is ticking. It's only two, three years. And knowing, I'll just give one example: if somebody decides to build an ammunition production facility, decision made today, if there are no hiccups and everything goes like in a movie, it's a minimum of three to four years before it's operational, maybe even five. So that's a reality. Then there are areas where decisions are by far much faster, and then we are just talking about ramping up the capabilities of something the companies are already producing today. So developing new capabilities and so forth, that's another story. So yes, as a total, we are in a hurry.

And the danger is that on this stage last year and the year before and the year before, there was the same creed around what the lead times are, and yet here we are. And if the producers wanted to be really mean, they would mash all of those clips together and show them next year after the same argument has been made again. Essa, Sylvie, Ashley, Matty, Jana, thank you so much for joining us this afternoon. I hope you enjoyed the discussion as much as I did. Thank you all very much.

Thanks. Thank you.

Okay. Hello everyone. Welcome to the last session of the day. It is my true pleasure to introduce my wonderful panelists. We have Miko Haala, the chief political, geopolitical and government relations officer and country manager for Nokia Finland, and Sianda, the vice president of defense at the Technical Research Center of Finland VTT. So please join me in welcoming Si and Mika. So this session is all about how we're actually using technologies in defense. And Sally, maybe you could give us an overview: what is happening today in the defense sector and what kind of technologies are we actually seeing?

Okay, thank you Melissa. So I think what's coloring our view on tech today is really the happenings in Ukraine and what we've seen in the past couple of years or three years of the very rapid acceleration of adoption of new technologies, also kind of improvised use of technologies, harnessing a lot of commercial off-the-shelf things. When it comes to, let's say, communication, you know, using Elon Musk's satellites for enhanced communications when the other means have not been usable, developing drones and so on. So I think it really differs from what the long-term view for defense technologies would be. We heard from Commissioner Birkunan about AI and quantum and microchips and space and what have you. So they are not maybe as relevant yet in today's acute situation what we see, but I would say that technology certainly has taken the driver's seat when it comes to reacting to all kinds of new threats at the moment. So the established means of kinetic impact and all will remain, but at the side of that, when there has not been enough means for volume-based battle, then you have to try to move into a new direction which is more smart and more improvised and more fast.

And you've spoken a lot about how basically everything changed for you after the Ukraine war, and we've now entered this new paradigm of defense. Can you talk a bit about what that is?

Well, we've been using the term 'new defense' and maybe it's good to share with all of you also here. I think that most of you are familiar with 'new space' or 'new space economy' where companies like Elon Musk's companies and ICEYE in Finland entered the very administration-heavy space market maybe 15 years ago or so. But now what's happening in the defense market? This 'new defense' means that there will be a lot of new entrants to the market and a lot of things are changing. It's a paradigm shift and systemic change. So instead of having very long-term development cycles in technology, let's say 10 to 20 year cycles, very administration-heavy, steered and predetermined results, we have then moved into something where we have a lot of new companies. These startups that we also heard earlier today, they want to enter the defense market, and these drones in Ukraine are just one small sign of that. But when we enter more into this discussion, I would love to discuss also how this would be an opportunity for Europe. I mean, this disruption in the approach could be really an opportunity for all of us.

I promise you we'll get to that. But before that, Miko, I want to hear what you guys are doing in defense.

Yeah, thanks. Before going into defense directly, I would perhaps bring back up the concept of national security, which is broader. Obviously, if you don't have national security in the broader sense secured, then your defense rests on a very shaky ground. So what I mean by this is that if you look at Europe these days, we do have massive problems in our digital infrastructure. One of the problems is that part of the infrastructure is already in the hands of, as we call it in our lingo, non-trusted vendors. So we actually don't have the full control of the technology stack, which always means that it's more open and vulnerable to bad intentions and actions. So basically, I think we have to have a broader view on European defense and tech, and one of the cornerstones is actually the critical infrastructure and digital infrastructure, because without this, you basically don't really have a reliable, sustainable defense. On the defense side, I think the fundamental driver behind what the organization that I represent is the AI super cycle. And AI super cycle means that AI will obviously occupy more and more space in all areas of life, including defense. And AI super cycle means that data traffic globally is going to double at least, probably triple within the next 10 years, and it's mainly driven by AI use. But it also means for us that AI doesn't work, it doesn't really mean much without the corresponding connectivity that actually makes sure that the data can be moved from a device to a data center and then back. I'll just give you an example: for a European customer right now, one gigabit per second is considered internet that is really working nice. If you have AI functions, you have to be able to move 800 gigabits per second. So it means that the level of connectivity that you need to run your economy or operate in the military domain, you actually have to be able to move massively more, faster, low latency. So that's the basic case: how do you bring AI to defense? And if you want to do that, and you actually have to do it, you have to also have the connectivity which is able to uphold the system.

And so can you, that's really interesting. I want to definitely talk about critical infrastructure, but I'm really interested in Nokia's role currently. What kind of technologies are you providing for the technology sector?

We are currently active in, I would say, three operations or domains. We are of course bringing connectivity to the strategic level. In defense context, that can mean a military base or a defense data center. Then we bring connectivity to the operational side, like moving bigger troops to forward deployed areas. Then we also have the tactical bubble. So we can bring the smallest version we have, which is actually a 5G independent network in your backpack, weighing about 7-8 kilos. So that's the smallest critical 5G bubble you can bring. Why is this all important? Because if we assume, and I think it's correct to assume, that AI systems are going to be decisive on the battlefield, and like I said earlier, they need massively faster data transfer. The old radios, as we call them, old legacy radios, they simply cannot handle the data volumes we are now talking about. So actually you need new concepts, you need new technologies, and what we are bringing is 5G. That is actually the only standard currently that can really give you 10 times the speed of the old systems at least. So this is the basic argument we have. Then I would also add here that we are now moving towards 6G in mobile networks. That's going to happen commercially in the next five years. We are now setting standards, and that will probably mean that the network itself will become a national defense tool because the network will have capabilities of sensing. It essentially becomes a radar. It can see, analyze, detect things that it cannot today. So this will also mean that the network doesn't only enable operations within the network bubble, but it actually becomes a defensive tool in itself. So these are the long-term opportunities that we have.

That's fascinating. I mean, just this June NATO adopted a new 6G standard. What is the significance of that?

Well, we are having a process of basically competing about who gets to set the standard for 6G. Obviously, the aim is that all the different players in the field could agree on a common standard so that all the equipment and devices would be compatible. We are not there yet. The process is ongoing, but it's obviously important that NATO actually in June this year adopted 5G as a military standard. So it means that it basically acknowledged that this is usable and viable for military use.

I want to pick up on something you said earlier about AI, and I'm really interested in hearing it. How do you see AI changing the nature of warfare, AI not just at GPT but in the broadest sense? So what do you think?

Yeah, I think it's unavoidable. So if we think of our pockets, we have mobile phones in our pockets and they have all kinds of computer vision and shape recognition. They will pinpoint your friends by name and they know from every one of your tens of thousands of photos, they know better than yourself who was who at what given time. So it's something that you cannot really stop the trend. And then I think that there is also quite a lot of talk about autonomous systems and drones. Drones are still quite a lot remotely controlled in a way that there is an operator, and there are some mishaps with that so that they are very vulnerable to jamming and what have you. So more autonomous systems can maybe operate when you are losing the connection. So that might be one kind of thing in there. And normally, if you have autonomy, you need to do that through artificial intelligence. Then soldiers typically want to have a human in the loop or on the loop, and not off the loop. So we would like to have a soldier deciding what to do. But that's also relative. So now in Ukraine, they want to design drones that maybe the last 50 to 100 meters are autonomous, sliding into the target when the jamming is at its strongest. So therefore, I would not like to consider that a black and white question, whether you go for AI or you don't go, but it will start. And for my sake, I think that this 'soup cannon' in Finnish, which is bringing pea soup to soldiers, might also be one of the first ones that should be autonomized, and not maybe the most lethal things. But if you're very categorized, then you don't want even to have this soup cannon to drive by itself to the front line, but it's all relative. So you have to start from the easiest areas of applications.

Solid. Thanks for the business idea. No autonomous.

Oh, sorry. Yeah. Okay. We need to IP protect that right now.

That sounds like an excellent idea. Help me put the recent news around drones into context. How worried should we be about this new technology?

The thing what we should be worried about, I think, is what the commander of Finnish defense forces was also highlighting in the news earlier this week: it's a cost imbalance that we have at the moment. The threat is caused by relatively inexpensive means, and on the other hand, the countermeasures are done with extremely expensive means. So therefore, if you think of war economy, this is entirely unsustainable. So we need to learn from the use of drones in Ukraine about how you can manufacture at industrial scale something that could counter those threats what we see at the moment. And I would highlight that probably we can never put all those counter things on a shelf like we have wanted to do with 155 mm artillery grenades. It might be such a fast-moving target that we just have to start to build capability to meet those things. So therefore, this technical capability would be starting from gifted individuals, suitable companies with suitable backgrounds and training, being prepared and increasing preparedness by creating projects that are doing that to some scale. But what we've seen for example in Ukraine is that first you have objects that are coming very slowly, then they are putting some jet engines into those things and they are coming very fast. And first they come so that you can jam them, and then they come having this fiber optic cable behind them and you cannot jam them. And it changes in a matter of weeks and months. So therefore, Europe needs to harness the civilian sector technology sector better to do that. And if you think of Nokia, Nokia is hardly a defense company, but it's an excellent example of a dual-use company that all the technologies can be also used for security and defense. So therefore, the companies like Nokia and even the small startup companies would be the ones that I would opt for making this change, rather than the present missile companies. If I say that your missile price should go to 1% of the original price, I think that they would not go for it as a good business model. But it has to be a hundredth of the original price at the moment what we've seen.

Excellent. Miko, can you tell us how are you thinking about this and shifting from a company that operates in peacetime to potentially wartime?

Well, basically Nokia has been in defense for decades. When the mobile networks and phones came, it kind of was forgotten for some time, but it never went really away. So it's been always there. Like Si said, we've been always providing this technology for defense purposes. What is different now is that we see the need for this massive data transfer capacity also in tactical settings on the battlefield. It has to reflect the need for data that you have there. So this is perhaps something new, and this is something that we are currently busy with. I would simply say here in this context that what Si said, I totally agree with the aspect that instead of trying to figure out what we do need exactly now and produce a lot of that, I think it's far more better and far more important actually to really build up long-term technological capability so that you will be able to address whatever the shape or form of the threat in the next 10 years, 20 years is. So there's no fixed solution to this. It has to be a moving target by definition. Then I think in the context of Ukraine, I've been trying to study it hard and being there trying to understand what is the critical lesson for us. I think it's not only technological capability. Actually, part of the capability is how do the companies and Ukrainian armed forces work together. Because the old practice used to be that you have armed forces, you have companies, then you have a massive layer of bureaucracy and process in between those two things. And by European standards, the quick process was seven years. And now you have a completely different situation. So instead of having this kind of insulation between companies and defense forces, they actually have to work together from day one, and they have to make new iterations. They have to find a new solution. They have to update things in a matter of weeks instead of having the formal process of years. I think this is going to be quite difficult for Europeans since we don't have the urgency, the absolute urgency of a wartime, when you can actually easily throw away all the layers of purchasing rules. So I think we have to learn to be faster, but the critical thing is that armed forces and companies have to not only collaborate but actually co-create, because that's the only way of really moving fast enough in this field.

May I comment? Even though we are now talking and having a European viewpoint at the moment, what we've seen lately in NATO and most of the European countries are NATO countries. So there are some fabulous new processes in NATO that have been launched very recently, like in June this year. So for example, the Rapid Adaptation Action Plan, RAP, which means that there are exactly those elements what Mika was highlighting: there is a faster route for the companies to get insight from the defense forces or NATO forces about what the needs would be. There will also be some support for the companies to recognize companies, startups and tech companies that have something good to give, and faster access to, for example, field testing or other things. And then Finland also declared at the same time in the summer a NATO innovation range which would be established in Finland. And that is actually highlighting next generation communications. I mean, we're Nokia country, in the Nordics we are Nokia and Ericsson countries. So next generation communication is a very natural technological choice for us, no question about it. And also talking about NATO, Finland is now starting its first NATO DIANA accelerator next year. DIANA is a system where there are more than a thousand or 2,000 applicant companies that have technical ideas and they are approaching NATO, and NATO is filtering out 97% of them as waste and 3% are coming into this program. Now we have had four Finnish companies so far in that NATO system, and next year let's see how many we are getting. But it's less than 100 companies annually that are getting training where military soldiers are telling what they need, investors are telling how you build up your company, and tech people are saying how you should increase your technology readiness level. So my message for Europe would be: let's utilize NATO to Europe's advantage. Because we mentioned US and also something about the Far East today. US interest is maybe not in land battle in the same way as in Europe. So if you think of all these communication things, drones and other things what we have been discussing today, they are very European issues. So therefore, if we operate within NATO and concentrate our forces on these topics that are very relevant to Europe, then we are combining our needs to some well-functioning mechanisms. And I don't downplay any European mechanisms, but the European Defense Fund, European EDIS, others, they are not maybe as mature or as impactful as some of these NATO things in some cases they are. Miko, did you have something to add?

Perhaps I would simply take a jump to a more strategic concern that I have. Because I'm actually less concerned of the European capability of handling the defense tech. I think we have the industry, we are capable of scaling it up. So we have all those ingredients in place to tackle the immediate short-term, middle-term threats. I think what is more fundamental is that the real reality is that Europe is massively lacking in its digital capabilities. And I'll give an example from my world. The share of the 5G standalone networks in Europe is 2%. And what does 5G standalone mean? It means that the whole network in all its elements is designed for 5G standard, which is the fastest that you can have right now. In China, the role of fully-fledged 5G systems is 80%. So in our case 2%, China 80%. So basically everybody is surpassing Europe in that respect. And this does mean that we simply don't have the digital infrastructure to build an AI-based economy and also defense on it. And how do you ramp this fundamental thing up fast enough? I don't think I have an answer. I think Europeans did wake up on this at least a year ago with the Draghi report. But the fundamental threat here is that unless Europe catches up, it will mean that we will progressively lag behind more and more. And why this is important in defense context as well? It's because I think it was one of the pre-events where it was said really clearly, and I do agree with that, that technological dominance in the AI world will equal to military dominance. So if we don't have the ingredients, the basic infrastructure to succeed in the AI world, whatever the practical solutions at some point will be, I think we have a massive problem in our hands. And this needs to be addressed. Then everything else can be built upon that.

But what do you mean? Does that mean Europeans building their own large language models, or concretely what?

I'll give you an example. It starts from even data center capacity that we have, which is clearly different from, for example, the US. It also, specifically for me, means the lack of connectivity. And like I said, Europe 2% pure 5G systems, China 80%. It may sound like it doesn't matter much, so Chinese will simply get their cat videos faster. No, it will matter massively. So it's about really the infrastructure. Our problem also in this respect is that we don't have incentives in Europe to really invest heavily. I'll give another example. A US operator typically who invests in these networks, they have 100 million subscribers. Average European operator has 4 million subscribers. So talk about scale here.

Sally, I'd love to hear your thoughts on this.

It's a grim picture. I'm struggling to try to find a silver lining to this thing. And I'm personally very concerned about the...

I have a solution.

Okay. Well, that's positive, but you save it to the later stage. I think one example what VTT has been doing is that in Finland, the investments in strategic technologies like quantum have been centralized or focused somewhere. VTT is not really holding a quantum computer in its basement and holding it for itself. Now we have had hundreds of millions of euros of investor investments into these startups and new companies that are rising. That is a follow-up of the strategic investment. So rather than having every town have one qubit computer, we are having 50 to 300 qubit computers that are central investments, and we have this LUMI computer, supercomputer, that we discussed and heard from Commissioner Virkkunen. So that might be something that we have to do more, and it's also painful for smaller nations like Finland to say that okay, Europe opts for more centralized solutions. Because Finland as a marginal thing in the outskirts normally is not getting everything by definition. But that's unfortunately, or fortunately, that's what we have to do. And going back to the physical world, if you think of the number of platforms that we have in European defense systems, they're immense. There are tens of different platforms for tanks or aircraft or what have you, compared to the US where there's consolidation and focus. Therefore, you can make better choices. So I guess Europe needs to pick its battles. I think the consumer business has gone. I don't know. I'm sorry. But it could be that the remaining option could be that if we could manage the data for more tailored or business-driven or sector-specific things, that could be one chance. Because there's nobody really collecting wise data about the battlefield. So in Finland, the winters are every year, and we could collect all kinds of things and learn, teach the AI about anything what you can see in the snow and be prepared. And I think we do that much. So all the data is from consumers, and all the consumer data goes to the US and China, and it's not really going to go to Europe in any greater sense. But I'm trying to find the silver lining in a way that there are some specific areas where we could be much stronger as a power.

I'd love to hear from both of you. How do we fix this critical infrastructure dependency or problem that we have? How do we make that more sovereign? Miko?

I think we have to take a look at the European regulation in order to allow for more scale. That means cross-border mergers. That means allowing companies to grow bigger. Then obviously we need to support R&D pretty massively. And I think we have to of course focus on the areas where we still do have a chance. And my sobering analysis, you can contest that, but we do have only a handful of globally critical companies in Europe left. You have ASML which is making those chips. Then you have Nokia and Ericsson, one of the few examples where also the US is kind of dependent on Europe and not the other way around. Then you have perhaps Airbus, SAP, things like that. So I think connectivity is obviously one of those fields where we still do matter more than anywhere else. So I think we have to pick our battles like Sally said and think meaningfully where we can make a difference. But currently the problem is that we need to invest heavily in R&D. We have to allow the market to grow bigger to invest, to really invest in scale, because otherwise it will lead to a situation where we simply continue to lag behind, we can't scale things up, and it gets more dangerous now when the AI super cycle is going on.

Maybe one pass to Miko again. I know that in Finnish media there was one quite recent article by our former CEO Antti Vasara from VTT, Pekka Lundmark, the previous CEO for Nokia, and Peter Sarlin from Silo AI, and they were bringing up an idea that for example in next generation communications, it's not only about communications. I'm sure that Nokia would not like to only transfer contaminated messages, contaminated data which is compromised in security and cyber and what have you. And those three leading tech guys in Finland were saying that when you have this new combination of next generation communications with cyber and AI and quantum and space, you are redefining the place. It's a bit like blue ocean strategy, so that we are maybe behind in many of those things, but we have all the freedom to really find what would be the future concept of trusted communications. And for me, I'm trying to find a name, whatever diamond or fireball or rock or whatever that would be, but something that would say that whatever is going there, it's resilient and it's trusted and it flows and it's sharable. But that's something I think that my main takeout from this article was that this could be an option, and companies like Nokia would have the industrial muscle also to take it that far. So in that manner, I think that there's always a way forward for us, but we also might need to change the rules of the game sometimes when we want to play to our advantage.

When you think about AI, until now Europe's approach has been very much to focus on the ethical side and make sure it's responsible. Do you think by doing that Europe has sort of shot itself in the foot?

Yeah, the ethics of AI is something that typically people are very concerned about, what the autonomous machine is doing, what the computer is doing when it thinks by itself. Five years back it was always about a car trying to opt for hitting one person or two persons or what have you. And I think that one thing is that we need to accept that Europe should be driving regulation for ethical and trustworthy AI as it is doing now. So therefore, the public acceptance of more autonomous systems and the infiltration of AI in all our lives will be somehow manageable. But then there's another ethical question. I don't know if it's a good time to bring that up, but it's also about when we talk about the ethics of this whole defense thing, what we're doing here, why we are sitting here with Miko today. And for me, I have been trying to tell myself that unfortunately it's an ethical imperative also that we are sitting here today. We would not like to be talking about defense issues as much as we do today. But our democracy-based societies are now under existential threat. So that's one ethical dimension: why we have to do this. And the other ethical dimension is that sometimes it's better to have a machine helping an elderly person rather than no person at all. And I also learned from one of the Finnish politicians, she might be here today, I don't recall who she was, but she was from a more remote part of Finland, and we were talking about ferry traffic that could be autonomized. And she said that that's a perfect ethical thing: people can stay longer, they can live out in the islands, they can keep the island inhabited because the cost level of this unmanned system is better than with a manned system. But you have to still do the ethical part, so people can trust that it's safe to ride on those things. I don't know if that answered your question.

Brilliant. I'd like to take some audience questions here. We have a mic, please. Could we get one here? Quite readily audible without a mic as well, but maybe this is being streamed. Wonderful.

My name is V. Lever, I'm professor of technology policy at Aalto University. Thank you for the very interesting remarks. If Europe cannot develop domestic AI frontier AI capabilities, should we still adopt AI across society and let it infiltrate into every aspect of society as you've suggested? Thank you.

Yes.

That's a very Finnish answer. Could you elaborate please?

Oh, okay. Yeah, you cannot neglect AI because like what Miko was saying, everything is about data and how data is being a part of our lives, and all the systems will be more connected, more cognitive, and what have you. So therefore we have to be good in adapting AI, no question about it. But probably the big owners of AI data models and the big servers and what have you, they might be a different ball game. But adapting that, for example in defense, we can be leaders in that, no question.

So even if that means we are even more dependent on foreign...

Just to repeat that for people in the live stream. He asked: even if that means that we're more dependent on foreign tech companies? Why not thinking of a risk-based approach so that you consider making dumb systems in order to avoid your dependency on US tech?

I would opt for going to wise systems or smart systems. But it's something that you have to consider. If you want for political reasons to reduce your dependency and at the same time your product range would be dumber, then probably it would not be wise. But it's not a clear balance.

I would say that if we somehow decide not to adopt AI, I propose that we all start to retrain ourselves as museum guides.

Yeah. Okay. Love the questions. We have a lady here in the middle, please.

I'm Palen from University of Helsinki, associate professor. I'm interested about these ethical rules which you were saying. I give you an example: we can create a lot of new drug candidates for medicines via AI, we get the properties which we want to get the new drugs. And then on the other hand, we can also create a lot of toxic chemicals which can be used for example as chemical weapons or perhaps even biological weapons. So how could you rule this? Can you guide the computer, supercomputer, that if there is something wrong started to do with somebody, it says no, we don't do this? Or have you any ideas about this? Because we know that mainly for the military purposes, the aim is to kill people and to protect your own country. But where are these ethical rules? Is it only in the person's minds and the campaigns and the governments? How do you think about it? Is it possible even?

Maybe a dull answer would be that pretty much all the technologies could have malicious uses as well. Communication could be disinformation or miscommunication. One thing is that if we as mankind are winning by harnessing AI into medicine's development and all of that, that has value, that is saving lives. So maybe you have to again balance like we're discussing earlier on the ethical issues. So it would be quite dangerous to start from a fact that you are slowing down or creating so much regulation that you actually slow down also the positives. I learned some years back that for example in the German automotive alliance, there was an ethical guideline that it is an ethical imperative to take new technology into use if it has been proven to be safer than existing solutions. That is quite an unorthodox thinking: you have a responsibility and accountability to actually harness AI if you can prove that it actually has a positive impact. I can't claim being an expert on this field, but I would simply assess that obviously when you harness AI for doing good in society, you will have the other side of the coin as well. And obviously you can try to regulate that with ethical rules, even legislation. The difficulty here is that because AI is at the front and center of the global power competition, how do you make those powers who want to gain more power and challenge the others? How do you regulate them to stop this kind of weaponizing of AI? So this is the dilemma where even a democratic rule of law country will have to face. So I think the discussion is going to be: shall we simply go further? Shall we even weaponize AI? Shall we plan for how to use it for those kind of purposes with the hope of finding a balance so that nobody can threaten us with this kind of super capabilities because we can then counter? It's pretty much the same dynamics as you had with nuclear weapons back in the day, when actually nobody could really afford for those great powers to stop that. They were rather incentivized to move further. So I'm afraid this might be the logic of the future. I certainly don't hope that, but I think we have to be prepared that something like this will also happen.

Do you think we should have red lines, something that we just won't accept, like autonomous killer robots for example?

We can have red lines as many as we please, but the question is how do we enforce those red lines in the middle of the massive deep great power competition? It's not about... it will be about the sort of capabilities. So how do we enforce those? This is the question we simply can't answer fully at least.

There's a question over there, the man with the paper please.

It's a unique identifier. Thank you. Vivo, University of Trier. Thank you for the discussion. Miko, I think you raised a brilliant point on highlighting the importance and lack of capability in the European infrastructure from a connectivity point of view. Super important. I would like to also highlight energy into the energy infrastructure. A recent report says that if you want to develop an AI factory, a leading AI factory in the world in 2030, it will cost 200 billion, which is a lot of money but can be found, but will require the equivalent amount of energy from nine nuclear reactors. Nine. So all this talk about we want to be a European AI superpower or Finland wants to be an AI superpower, where do we get the energy? Is this futile talk and are we just trying to deceive ourselves that we are in the game, or like Billy said that we should recognize the dependencies on the technology of the great powers? Thank you.

Thank you. Sally, do you want to take this?

No real answer to that. So this whole energy debate is something that looks like with the amount of renewables, we are actually being able to meet the normal requirements when it comes to European targets, if the technology development goes the way how we see it today with nuclear and wind and solar. But if that's an exponential increase in energy consumption, that would require some other things. So normally I guess that throughout history, the more you have had energy, the more you have spent it. So there's never any excess energy in a way that it's all going to be consumed. So the more we are having it, it will find its way to our systems and processes. So one thing is that at least we have to try to consider this. I would like to combine this energy question that you might now touch from a sustainability point of view into a resilience point of view. So that also if you are building your energy systems, resilient energy systems would be less dependent on fossil fuels and would be more based on renewables, and therefore that would be one option for that. But if you are running your AI server farms with fossil fuels, that doesn't seem very clever either going forward. I think if you ask Finnish energy companies, they will give you a slide where they can prove that they can scale up the energy production that will cover those needs. I think on your question, probably we have to also keep in mind that you can calculate things assuming that the energy efficiency stays the same in the future. But what if it improves massively? Just to give an example on 6G, we expect that the energy efficiency will drastically improve in the next 10 years. So you may have things on the horizon that would enable you to cover that as well. So we just don't know that.

And maybe quantum computers someday, depending a bit on what kind of time frame we have. So there might be some disruptive ways of computing also in the future, but they are not yet today.

Thank you. Any other questions in the audience? Oh, some pointing over here.

Peter Watkins from Chatham House. Sally mentioned new defense and drew the parallel with new space. I just wondered if you could say how quickly you think this transition might occur. I mean, in space it's happening pretty quickly, but I would suggest that in defense it might be a lot more difficult. There are more and bigger and more influential incumbents, all sorts of issues, vested interests and so on. So how quickly might it occur and how might it be accelerated?

Thank you. Excellent introduction to the topic, and those are the kind of things that are actually hindering this type of thing. Like I said, NATO and many nations have new methods and tools to try to attract more companies coming into the market. I would say it takes a long time. When it comes to space, it's not that massive. I mean, it's been massive for the past maybe five years or so, mainly from American companies. And the small companies like we have ICEYE in Finland now, they have 800 people and I don't recall how many satellites they have, but it has taken a long time for them also to grow. I think it will be the same in defense as well. So one main thing in that new defense is: do we have to channel all the new entrants to the market through existing incumbents? And I think that the existing companies, established defense companies, might be filtering a bit. It might not be in their interest to disrupt their own business. If you want to get a disruption in cost level, then you might need to create a challenger. That's what happened pretty much in the space industry: you have a different cost level with these new operators. So therefore they might need to go to the market directly. And again, we're here with Nokia. Nokia has products that they can export to the defense market directly, not always using a prime or an integrator in between, but of course in many cases yes. So I'm just thinking: how can we make sure that the large industrial primes are not starting to be blockers for new technology introduction? But many times, the administrations that are acquiring new technology can also set all kinds of rules about what they want to have. They might set a cost level target or availability target or tech target. So maybe they have some ways of directing this. So no clear answer. I would say that it will take a long time before these small companies are coming. We have been analyzing this Finnish dual-use or new defense landscape with roughly 200 companies, and they are growing much faster than existing companies. We had Patria and Linko here today. The growth of these small companies is extensive, but they are starting from near zero. So it will take a long time before they are new Patrias or Airbuses. Creating a new Airbus will take 20 years. Even though we have companies like Palantir, those are interesting, but they are exactly the type of companies what Miko was bringing up that are not from this continent. These data-based companies can come into the 10 billion range not overnight, but within a couple of years, but not in Europe.

Thank you, Si. We have a couple of minutes left, and I want to give you both a final task. I want to know from you: what is the most important thing Europe can do to future-proof its defense sector? And you both have 30 seconds. Miko.

Well, I think the critical thing is to address those foundational problems that we have in our digital infrastructure and build all those models on that. Unless we do so, any subsequent effort we do in defense or any other field is quite likely to fail because it doesn't have the foundation to build upon.

Thank you. Si.

I would say: beef up the existing defense industry with dual-use or civilian industries. Build those nuclear or polar icebreakers for us, export them from Europe to the US, safeguard the whole North Pole with European technology, export Nokia stuff in there, and try to get as many civilian companies to also learn how to address the defense market. Not only the defense industrial companies know how to do it, but that's where the growth potential is, I think.

Excellent. Thank you so much for this fascinating conversation.

Thank you.

Thank you so much.

Thank you. And I'd like to thank all of you for joining us today and tuning in for your brilliant questions, and everyone at the live stream as well. That's the wrap for day one. Thank you.