Solution Spotlight - AI Enterprise Solutions & Technology Independence | unDavos 2026

Laura Herman (Potentiary), Richard Hong (TNE.AI, ex-Microsoft) and Mark Terrell (unDavos) debate whether trillion-dollar data-centre buildouts are the battleships of our era. Richard reveals how his company shrunk a one-trillion-parameter cloud model to 1.7 billion parameters — running on a $1,000 laptop at 25 watts instead of 20 kilowatts — while delivering the same results for enterprise KYC and anti-money-laundering. Mark argues the real unlock is secondary markets and pre-exit liquidity that could free up hundreds of billions in early-stage capital, enabling four-person startups in Nairobi to build what used to require Silicon Valley budgets. The conversation spans AI sovereignty, critical infrastructure fragility (Berlin lost power for two days from a single substation attack), quantum computing's looming "Q-day" that could expose every encrypted record, and why every country now needs to think about education, energy independence and technology sovereignty as a single stack.

SPEAKERS

• Laura Herman — Founder, Potentiary | 20 years in nuclear science & tech transfer
• Richard Hong — CEO, TNE.AI | Ex-Microsoft (12 years), former VC (15 years)
• Mark Terrell — Founder, unDavos | PhD in Collective Intelligence (Intel)

unDavos — unconventional conversations at Davos. 200+ sessions across 12 tracks, bringing together founders, investors, policymakers and scientists for the discussions that matter.🔗 https://undavos.com📺 https://www.youtube.com/@undavos

🌐 undavos.com

---

TRANSCRIPT

And the stage is yours. All right. Thank you very much. Richard, if you want to take the microphone and you want to sit, I'm going to moderate. So I'll stand. Sure. We've got two chairs. That makes sense. My name is Laura Herman. I'm with Potentiary. This company was founded as an investment advisor after I had 20 years working in nuclear science and technology across the entire fuel cycle, doing tech transfer from our national laboratories in the United States out into the startup world. So in the last decade or so, I've had the opportunity to work with a lot of startup companies who are trying to break into large industrial systems-oriented types of industries. And so that brings me here today and to Davos this week. And I am excited to be having a conversation with Richard Hong, who is former Microsoft. He's going to tell us a little bit about his company and Mark Terrell, who has been organizing these spotlights here on Davos. The one and only. Yeah. So Richard, please, do you want to take a few minutes and tell our audience about what you're working on and how you got started in the AI infrastructure? Well, thank you, everybody. I think that, you know, we have an AI house here, we've got all these crazy things happening, and I guess I'm here to tell you that actually making these things work is way harder than you think. And I often ask for a show of hands, because if you look at the press, it seems like it's just happening everywhere. And I guess I spend a lot of time talking with the very biggest companies, and the answer is it really is very hard. So by way of background, I worked at Microsoft for 12 years. You can blame me for all the issues with Windows Office and our server products, I guess that's one thing. And then I spent 15 years as a venture capitalist, and in the last five years I've been doing AI companies. And we are a company called Total Neural Enterprises, TNE.AI, and we're really trying to help the biggest companies adopt this technology in an efficient way. I love to hear you say that transferring artificial intelligence into these enterprise systems is not easy, because too often I hear it kind of presented as a panacea. So we'll talk a little bit more about how it can be difficult. It's called the San Francisco AI House Party, by the way, if you're wondering what the actual techni

Transcript

SPEAKER_00 0:09

And the stage is yours.

SPEAKER_02 0:12

All right. Thank you very much. Richard, if you want to take the microphone and uh you want to sit? I'm gonna moderate. So I'll stand. Sure. We've got two chairs, that makes sense. My name is Laura Herman. I'm with Potentiary. Um, this company was founded as an investment advisor after I had 20 years working in nuclear science and technology across the entire fuel cycle, doing tech transfer from our national laboratories in the United States out into the startup world. So in the last uh decade or so, I've had the opportunity to work with a lot of startup companies who are trying to break into large industrial systems-oriented uh types of industries. And so that brings me here today and to Davos this week. And I am excited to be having a conversation with Richard Hong, who is former Microsoft, who's gonna tell us a little bit about his company, and Mark Terrell, who has been organizing these spotlights here for on Davos.

SPEAKER_00 1:12

The one and only.

unknown 1:13

Yeah.

SPEAKER_02 1:14

So, Richard, please, do you want to take a few minutes and tell our audience about what you're working on and how you got started in the AI infrastructure?

SPEAKER_00 1:22

Well, um, thank you, everybody. I I think the you know, we're we we have an AI house here, we've got all these crazy things happening, and I guess I'm here to tell you that actually making these things work is way harder than you think. And uh uh I often ask for a show of hands because if you look at the press, it seems like it's just happening everywhere. And I guess uh I spent a lot of time talking with the very biggest companies, and the answer is it really is very hard. So, by way of background, I worked at Microsoft for 12 years. I ran I you can blame me for all the issues with Windows, Office, and our server products. I guess that's one thing. And then I spent 15 years as a venture capitalist, and in the last five years I've been doing AI companies, and we are our company's called Total Neural Enterprises, TNE.ai, and we're really trying to help the biggest companies adopt this technology in an efficient way.

SPEAKER_02 2:12

I love to hear you say that transferring artificial intelligence into these enterprise systems is not easy because too often I hear it kind of presented as a uh as a panacea. So we'll we'll talk a little bit more about how how it can be difficult.

SPEAKER_00 2:26

It's called the uh San Francisco AI house party, by the way, if you're wondering what the actual technical term is.

SPEAKER_02 2:33

Great. Mark.

SPEAKER_01 2:35

Hello everyone. Uh so I'm Mark Tyrrell. Um let's say a few things. So most of you will know me from, particularly in Davos, will know me through Un Davos. Uh very good. These days, by the way, un Davos means unconventional Davos. Uh at the beginning, it was underground Davos because there were only about three or four people that were doing side events. I mean, very hard to imagine. Um, and then it became uh unconferencing Davos, making use of the space. And now Un Davos itself is quite big. Yeah uh big. Um, but for me it's really about uh creating spaces where we can all do stuff. Uh my background. So uh I did my PhD work in collective intelligence and crowdsourcing uh when I used to work at Intel. Oh, we've got a chair. Thank you for making the chair. Lovely.

SPEAKER_00 3:18

She made a chair.

SPEAKER_02 3:20

Thank you.

SPEAKER_01 3:25

So right now, um, so I did my PhD work in collective intelligence and crowdsourcing while I worked at Intel. Um I had a database uh PhD, so I was looking at server log data. So originally it looked as though I'd lucked out because the person that the sort of uh master's student who was working for them as well, she got to do interviews with people in San Francisco, in uh in Milan, in Stockholm about how people using this was uh Lotus Notes, groupware, very early days of technology.

SPEAKER_00 3:55

I remember deeply. Yeah, yeah.

SPEAKER_01 3:57

So and I was one of the first people to use serverlog data to see how people actually use computers. And what I found then is that everybody was lying. All of the human now, they didn't intentionally lie. Some of them would say, Oh, I think this program is fantastic, and I'd say, No, I think you use CC mail, you didn't even use the Lotus Notes. People are saying, Oh, it's terrible, it's slow, I get no value, and they'd be using it massively more than anybody else. So just I realize that humans are weird, but technology when it could be used well, could do fantastic things. Um, but right now then, so I do a variety of stuff, but I'll I'll share stuff as we go.

SPEAKER_02 4:32

Great. Well, so good, because Richard and I started this debate in the hallway. Okay. And I've got to continue this saga. Because uh, you know, we think about how AI applies to these large enterprise systems, and and when we're talking about energy use in particular, which is an area I'm interested in, um, we were talking about the whether or not this is a um energy consumption and really minimizing consumption at the use point is something that AI can help a lot with. We've seen that with demand response in the utilities. Utilities in the electricity sector are not known as early adopters. Right. And so, Richard, can you talk a little bit about your perspective on the use of AI in the industrial sector and where where technology such as AI can be helpful and what still remains to be done from an engineering and a science perspective?

SPEAKER_00 5:27

Well, I I just want to say that I have deep understanding of these issues because one of my investments was in an energy trading company in Australia, and I spend more hours than you want to know analyzing spot pricing of electricity. So, with that perspective, I think that there's two ways that AI can help. The first is that uh just load prediction helps a lot, right? Because the fact is uh these systems have to be sized, you know, size for peak load, uh, small modular reactors are another nice solution. But one of the things I think that AI can help a lot with is doing that demand prediction and then figuring out where the load is really going to be and when it's gonna be. But the second is actually way more fundamental, which is that uh today, almost I would say 99% of all the AI compute is done in the cloud. And that is some of the most expensive in terms of electricity performance per watt. And the reason is uh like a car, right? Since we have Formula One here, a Formula One car goes about three times faster and costs about$250 million to make, or you can go buy a small car for$8,000 or$10,000. So I think what's happening now with AI is this gradual migration of that compute from just data centers to PCs uh to phones. And this phone, for instance, is more powerful than the most the fastest Cray computer that was doing defense nuclear warhead things in 1995. So I think you're gonna see two things happening. One is more load prediction and so forth, but the second is just moving the workload out to the devices people already have. Uh for instance, uh Apple just announced a feature I can't believe. Actual live translation. You put your AirPods on, and by the way, when it's doing that live translation, it's not going to the cloud. It's just running on the local machine. And these machines are drawing five watts, 15 watts. So that third that efficiency I think is going to help us a lot too.

SPEAKER_02 7:26

Mark, you do a lot of work with On Davos with venture capitalists like Richard, and really helping bring investment to startup companies that are bringing these innovations to market. And I'm curious from your perspective, when we're talking about things like what Richard's describing, this sort of systems change, where you've got to not only have your suppliers of your phones developing phones that can do these things, but there is that larger system approach. Like that type of investment doesn't happen on a project finance basis, one company at a time. So, where do you see the role of conversations like on Davos fitting into that building those big ecosystems for those transformative technologies?

SPEAKER_01 8:08

Well, and I think that so there are many, many different pieces. There's also many different countries involved. Now, and that's why the Davos conversation is very interesting, because to some extent it's fairly easy to take a US-centric view, even when we're in Europe. And at the same time, then uh almost like the thing that's absent from our sort of most of the world, at least sort of almost by choice, is the Chinese world and then the emerging Indian world as well. Because if you like, there is a let's look at whatever hyperscaling. So I was having conversations yesterday at the Belvedere uh with people doing hyperscaling. And the thing that gets me is that okay, uh I lease cars because I'm not an idiot, but if I were to buy a brand new car, it loses 50% of its value the day that I buy it. If you look at these chips, these chips that we have right now, will they really have a 10-year lifespan? Well, they might be used in 10 years' time, just like I use my iPad to pretty much stick it in the kitchen to open up uh recipes, but it's not going to be used for anything really, really cool. So one of the things that gets me is first of all, some of these technologies will be obsolete in two years. We're throwing massive amounts of money in. The next thing, and this was a Davos observation I had last year, is that last year during Davos, when it was AI, AI, AI everywhere, what then came out is a little company called DeepSeek had a completely different way of doing it with a completely different cost model. And I think what the Davos conversation does is that it brings in perspectives from different uh countries, mindsets, technology base, which could mean that even though there is a consensus, we're marching off in this direction, it's all good, goes whoosh. And I think we're very likely to have messy, many sort of whoosh moments through technology in the next months and years that will be dramatic. We'll have a quantum uh computing uh session that's going to come up sort of after this thing. Eventually, when Q Day happens, all of this data that we have that is um encrypted will basically be in the clear. There's no conversation talking about like all of your private pictures, all of your visa records, everything will be in the clear in is it five years, is it three years, is it ten? There is no, I mean, but Davos is the type of place we can at least start to have these conversations. Now, sure we should have had these conversations ten years ago, but quite often there's a consensus that no, we should but so uh right now, yes, I I do see this. Davos gives us the place to have this ecosystem conversation within the different domains of technology and the different countries as well.

SPEAKER_02 10:34

So, Richard, as a as a VC, how do you approach your investments in this kind of world where things can change so fast? Do you try to stay diversified? Are you looking for specific sectors where you think you can have the biggest impact?

SPEAKER_00 10:48

Well, this is where I'm glad I only spent 15 years as a venture capitalist and now spent the last five years as a startup person. You're you're asking the hardest question. Because it is very, very hard to predict the future. And the problem is uh it's called selection, choosing the right company. It's sort of like uh you go on a date and you get married for 15 years, and that's venture capital, right? Because you have a it's a great meeting, it's wonderful, you have the first day, and then you realize it's the first day of the rest of your life, and you're gonna live with this investment uh typically through two or three CEO changes, typically through four or five market changes, six pivots. So I think it's a very, very hard thing. And I think particularly now, because um just to follow up on your on Davos comment, I think this is the first time that we are we have a technology where we were just talking about Zambia and Kenya, for instance, where every other technology thing I've worked on, you can buy a phone, but you can't make a phone. You can buy cloud services, but you can't make AWS. This is the case where you mentioned DeepSeek, you can actually download that model in Kenya, build your own system, and you have zero dependence on any trading partner. This is a truly revolutionary moment where everywhere around the world you can develop your own state-of-the-art company. So, in fact, um we we started in Seattle and we're in the middle of moving to Singapore because we see the difference in worldview being so great between the United States, Southeast Asia. Well, we're gonna start an office in the Middle East so we can access Africa, and of course, India is right in between. So I anyone who's doing these things, I just encourage you to realize that it is unlikely that the way AI is deployed is gonna be the same around the world anymore. In the old days, you just take Microsoft's model and just import it. Today it's not that clear because maybe you don't have to have gigantic power grids. Maybe you can be more focused on PC infrastructure and micro generation and not to find a not find a 50 gigawatt plant.

SPEAKER_01 12:52

Yeah, I can see I've got an extra thing though. So um being an angel investor is a stupid idea. I know I am one. Yeah, I know. Um being an early stage VC is a stupid idea.

SPEAKER_00 13:02

Even any VCs here. Could be even dumber. Yeah, yeah, absolutely.

SPEAKER_01 13:06

Even late stage has got a phenomenal amount of risk as well.

SPEAKER_00 13:09

I've done that too, that's not so funny there.

SPEAKER_01 13:12

One of the things that so I did an IPO for my first software company. And I did an IPO because I wanted 50,000 euros to renovate my kitchen. And not a single investor, even though I had a company that was doing something, but not a single investor would basically buy a little piece of my stock so I could so I had spent$340,000 and 18 months to do an IPO so I could have divisibility, divisibility of shares. As a founder, it sucks.

SPEAKER_00 13:36

Liquidity is hard. Yeah.

SPEAKER_01 13:38

Well so one of the things that I'm obsessed about, and I started a couple of years here in Davos, last year really got into, but this year is big, is secondary markets and pre-exit liquidity. Yeah, it's a big market. Because, I mean, if we look at this right now, what you're doing is stupid. It's just who on earth would say, like, I'm gonna bury the money in the ground and basically keep it there and I'm gonna hope it's okay. It makes particularly when there are instruments. So as I've been working, so my what I would love to do is free up hundreds and hundreds of billions of dollars of early stage capital. I had a few friends that put money into perplexity at an early stage. They're looking at what, 10 or 20x already. They are very active early stage investors. If they could exit half of their stake, they would have an extra 10 companies that they could invest in. And if they could invest globally, which means you don't have to pay San Francisco prices, you can get the Kenyan engineers doing the Kenyan model, this could free up hundreds and hundreds of billions. But you just have to have a few people think slightly differently on how they manage a cap table and what document they sign.

SPEAKER_02 14:40

So but but talk about that again from the perspective of these uh capital intensive improvements that we're talking about when OpenIAI talks about the the billions of dollars that they want to spend on data center build-out, which somebody else has to do for them. When you think about uh the energy system and like scaling these startups is is a whole nother animal, right? And so how does that fit into this model?

SPEAKER_00 15:07

Well, I think there are um I think AI right now is such a classic tale of two cities, right? There is the highly capital-intensive, we're gonna put, you know, for instance, the US, uh I think I read statistics like 80% of their capital expenditure is now data centers. So kind of amazing. Huge percentage of GDP growth is just data centers. It's not clear that is the model that's gonna work worldwide. So let me just construct a story.

SPEAKER_01 15:34

It's as though America decided, like in the in the age of it feels like it's a difference between canals and railroads. But we're just at the cusp where the first railroads kind of there, but they're kind of crappy. So it's much, much better to raise capital to make canals everywhere.

SPEAKER_00 15:47

Uh I actually have a different analogy. I have a different analogy because I just went to the defense thing. So for those of you who are history buffs, uh when World War II happened, everyone was building these gigantic things called battleships. Right? And they were amazing, they cost hundreds of millions of dollars. And then this uh, and by the way, a typical battleship then cost a million US dollars. An airplane costs about 8,000 US dollars. Turns out that an$8,000 airplane can sink a battleship in about 10 minutes. So I think it is much more analogous to say that the current way that America is building things has to do with the fact there's a huge capital base and you've got to use it somehow. It's not clear to me that that capital intensity is really appropriate for most AI applications. Let me just talk about one of ours. So we build advanced know-your-customer applications, anti-money laundering applications. Uh, we use very small models, we can run them on laptops, we don't have to, we don't have to buy anything from OpenAI. We typically can deploy them even on a phone. And uh it turns out you can have one or two software developers, AI architects, write the whole thing for you.

SPEAKER_02 16:55

And you can do that at scale.

SPEAKER_00 16:57

Uh that's what we do. So we actually have, uh because the university education system isn't there, we have our own internship program. And anyone here who's got a smart, who's once smart can apply. But basically we have an internship program that puts people through a six-month boot camp that teaches you all the business skills and technology skills. So you can literally, because this is the way our system works, you can literally write a business plan and then write the application. So you don't, you're not writing code, you're just saying, I need an application that can do this, this, and this, because it turns out most software projects fail because actually the business never made sense. So I think this is the kind of radical thinking that can happen in an undavo's world, because I can build a company in Kenya with four people and a laptop for 100,000 quid instead of 400 billion dollars. So I think that the unpart, you know, the uncola part of it is yeah, you can really think differently about building companies. I'll tell you, our company, you know, we'll be cash flow positive this year. We we've just done a small amount of angel funding, it just works. Because the thing is you can take the efficiency of this AI development and apply it to a company. So I think they're really two different models. So everyone time every time you hear someone saying we have to put a trillion dollars in, I just think about those big old battleships that people were building, and it could be the wrong move.

SPEAKER_02 18:16

So you see a future where these things are decentralized, they're pushed out across enterprises? Absolutely.

SPEAKER_00 18:23

I think that that is going to happen in big enterprises, but to talk about the and Davos part, uh, you know, I'm going to the African house later. It can happen anywhere, it can happen in Leeds, it can happen in Manchester. You know, this old world where you had to lift up your boots and go somewhere, that sort of ended now. Because you can actually, and for instance, uh in Wales, we're doing a whole training program for a big university there. We'll probably do it in Greece. And so this is much more approachable than ever.

SPEAKER_02 18:50

So, what would you either of you say to the critics who look at OpenAI, uh a chat GPT inquiry, ver the energy intensity of a chat GPT inquiry versus a Google search?

SPEAKER_00 19:03

Oh, I think that um I just give you an example. So we've built a sales, uh it's a sales engine, right? So it figures out which customers you want and so forth. And we first deployed this in the cloud model, and then we started shrinking. And we have a technology that you can take something that works for big and just keep making it smaller and smaller and smaller. And we thought we could hopefully get just technically speaking, have it run in a PC. So just to give you the numbers, uh OpenAI is probably one trillion parameters, it's vast, right? A PC model is about a hundred times less, about thirty billion parameters. We got to run in 1.7 billion parameters, so almost a thousand times smaller and a thousand times cheaper, and it will run on a thousand dollar computer. That's great. So it has no and and by the way, that computer burns twenty-five watts. So you've we've taken something that probably cost, you know, I'd estimate 10 to 20 kilowatts, and now you're 25 watts. So I think that's the engineering that's important.

SPEAKER_02 20:00

And you're pushing the the processing out of the data centers and down to the It turns out that it's in the computer science world, uh the expense goes like this, nonlinear.

SPEAKER_00 20:11

So if you only need a little bit of compute, it's very cheap. If you need a lot, it's very much, much more expensive than it should be. So you always want to move down as far as you can on the curve, and that's one of the things our our company TNE.ai does is try to figure out where the right part on the curve is. And when you're in deployment, that's really important.

SPEAKER_02 20:27

And it's a huge, it's a huge human behavior change for big enterprise systems that are used to being really centralized and understanding how those systems work.

SPEAKER_00 20:36

Well, here's the funny thing. If you look at uh if anyone who used their web browser today on their phone, what percentage of the computation for that web browser page do you think was in the cloud versus on the phone? Turns out when you're scrolling back and forth up and down, that's all on the phone. On the phone. All it's doing is squirting a bunch of data down. So if you ever wonder why any of these apps are so fast, it's because they're all running on the phone. So actually, our uh one of the ways to think about it is there's about 20 million servers in the world, and there's about six to seven billion dollar end devices. So all the computing is out there.

SPEAKER_02 21:10

I know we're almost out of time. Mark, did you want to add one more thing?

SPEAKER_00 21:13

I can always add extra time.

SPEAKER_02 21:15

Okay, all right. I I know he's in charge, man. I know we're going. He's in charge. I try I'm trying to be the good moderator, timekeeper. You're doing very well.

SPEAKER_01 21:24

So the um Last year we had a logo which is called It All Changes. Now I would have done the same again, but it would have been a little bit lazy. So right now, I think because compute is relatively cheap, because the knowledge of making this stuff is basically everywhere, because there are so many module pieces, because money is not is is quite available, because there are lots of different countries, it means all of the enabling conditions to allow us to invent all kinds of stuff are all here. There are also some technologies like whatever nano atom by atom building of materials that might have been 20 years in research. It's also coming out. You've then got the quantum stuff that is also coming out. We've then got robotics, humanoid robots, everything else. I actually think the tech stack or the tech platform that we have is going to radically change almost every single business and organization. And what I'm looking forward to is there will be whatever, some VC brave enough or sovereign wealth fund, whatever, will say, you know what? I'm going to take somebody that used to run a Denon or used to run a Bosch or a Ford, and I'll give you 50 million to destroy that industry and rebuild it with all of the new tools. Not just like a little bit, like we're going to get rid of our SAP system and use something. No, no, no. But imagine that right now you have uh well you'll build the factory with humanoid robots at night. You'll still use the humans during the day, but at night you can get the human because they sit, they they can see in the dark. So right now, but that'll just be one piece. I think right now, I think me sort of my hope or expectation from a place like Davos is there'll be very forward-thinking people that'll realize that all pieces of the stack can be changed at the same time. And that means you don't need as much capital, or you can get much more expansion. So I think this is an incredible renaissance time. Now, I do think it's going to be a little bit tragic for the people that aren't going to be following along. However, from my view, on our phones that we have, we have the gift of knowledge. People could learn Korean, they could read Cervantes, they could do anything. The fact that they choose to look at Instagram videos is their fault. Just my view. They have the ability to do anything and they choose not to use it. So I do see though that those people who sort of can sort of rethink and rewire will just this is a phenomenal time to be alive. As long as you're on the top of the wave and not underneath it.

SPEAKER_02 23:44

Yeah, I was gonna say I'm gonna have to play devil's advocate because while it's exciting to see all of this technology get pushed out to something as small as a phone, the the robots that you're talking about, all of these things require electricity. They require energy in some form. And when it comes down to it, we're still boiling water to turn turbines. So to keep these machines running, we have some incumbent industries that require really um massive change. Absolutely. And and so that is, I think, one of the bottlenecks that we have to contend with to make this reality, this vision that you've described a reality.

SPEAKER_01 24:24

So at the um just sort of from the defense panel beforehand and connecting these things together. So I was having a conversation with actually with Bronte, and we're having playing a little game about um will you ever have city defense systems? So um Ukraine sent-cause we did in the Middle Ages. Well, and we're gonna we're we're moving to the dark ages back again, cities states, and everything else. But the logic is that so uh last year uh Ukraine Special Ops sent a truck full of drones, they drove the truck 2,000 kilometers, and then the drones just had to open up its roof, travel five kilometers, and it destroyed whatever, a quarter of a trillion a billion dollars worth of military hardware. If you think about it, all of our infrastructure is incredibly critical. There is no defense for any piece of infrastructure, basically. Particularly when you could just have a truck and drive it through. So right now I've been sort of playing the game, and this I mean, this might be when energy could change. Now let's imagine an actor decides, probably for a budget of whatever, one or two million bucks, to go out and take out pieces of the electric electricity infrastructure of the United States. How much would it cost to take out a dam? Two million bucks, maybe four? Go after the power lines. And power lines are cheaper to get out. Well, it's the okay, so so right now then, at the moment sorry, I really love military history. Well, and we are living, I mean, right now, the fact that maybe having tariffs will allow a country to take over Greenland. Maybe it won't. But we're living in a war, like I think warfare is going to be the next topic for this next Davos. But then it means then that our in energy infrastructure, if we look at let's say, how we then begin to have changes, war is a way that you have a lot of changes. Ukraine is already seeing that now. At the moment, there's massives of complacency that everything will just be the same and maybe 2% different. Right now, when we see pieces of these infrastructures dramatically changing, it will force through changes, kicking and screaming. And if not, quantum computing will take the board of directors, it will find out all of their emails, visa, card transactions, and everything else, and that will happen. This it'll be crazy shit. Like great function.

SPEAKER_00 26:31

Can I make one prediction for you? Yeah, sure. Um I think now just to follow up on your point. I think everyone now knows that food security, energy security are vital. I think that every country and every place needs to think a little bit about their AI sovereignty now.

SPEAKER_01 26:49

Because the, you know, the easiest way to And a tech technology sovereignty. I mean, I wouldn't limit it just to AI, but technology.

SPEAKER_00 26:57

Because today there are countries that can just be completely shut down with an executive order from somewhere that just says you can't use data center X or you can't buy this critical piece of hardware. So to me, these are the big issues now. It's not just uh defense or war, uh, but it is also about do you how are you self-sustaining on how you educate people? Right? That's one big thing. Where you get the power, and being in, I do think, just so you know, I think this multilateral uh cooperation is gonna be very interesting. I'm going this afternoon to the you know the the new renaissance of Europe summit, which is all about how Europe, where are its dependencies, should it develop its own uh I'm gonna use the the new US term war technology since we no longer have defense technology, war technology, should you use your own, should you have your own data technology, semiconductor technology, AI technology, how independent are you? Because from independence comes strength. Yeah. So these are I think the issues that I'd love to see next year when I'm here. Totally, absolutely.

SPEAKER_02 27:58

If we need to sum this up, we're talking about if we're s if we're as we're doing that, um, the solution spotlight for AI enterprise solutions to really work, we need to consider education and resilience, is I think what you uh where we're concluding this conversation.

SPEAKER_00 28:17

I mean I would say the message is you need to have an you know, this is where small modular reactors are great, right? You need to have enough energy independence so you don't have to worry. You have to have enough a strong enough education system so you can move people together, and then you have to have enough um uh technology. Technology sovereignty. So are you using a stack that is primarily one where no one can say, I'm not gonna give this to you anymore, and your entire country falls apart.

unknown 28:42

Okay.

SPEAKER_01 28:43

I mean another thing as well is that unfortunately there are no three things that will help. There's 50 things. And I think right now there is the we have a habit of simplifying things so in our poor, stupid human brains we can handle stuff. Now already AIs can do it better than us, because even like your basic chat GPT or even your phone version can handle 50 things at the same time, which we can't. The thing is that if you look at all of these challenges that we're working on, they all require at least 50 things to be worked on. At least. So right now, whilst I would love it to say, oh, here are the three secrets, it's obviously the first thing is that there is no three secrets. Yeah, it's complicated.

SPEAKER_02 29:18

The human behavior aspect of it does require, though, we find some ways to simplify it and keep it focused on a solution set that people can wrap their minds around.

SPEAKER_01 29:28

When the machines take over, we'll find out if we were right.

SPEAKER_02 29:34

Thank you very much for hosting this conversation. Great. Are we taking questions?

SPEAKER_01 29:37

Um I guess so. It's uh up to you.

SPEAKER_00 29:41

You can just get machine guns, you don't have to get submachine guns. Just go for the full assault rifle.

SPEAKER_01 29:48

Well, so I live in Berlin. Berlin had no power for two days because a bunch of left-wing activists decided to uh basically take out uh the sort of power substation. So right now, our infrastructure is completely fragile. It has zero defense, zero. And for some reason, the people that are in charge have decided not to do anything about it. It is incompetent. Generally, it's incompetent. But anyway, but it's the I'll we have a couple of uh by the way, um for those people who are expecting to speak still, if you put up your hand, Max, you want to Okay, yep. And then uh I think we had three more people to uh to fit in.

SPEAKER_02 30:26

Excellent.

unknown 30:27

Very good.

SPEAKER_01 30:27

And and by the way, and I realize that some people have problems traveling in, so that we have to be a little bit flexible for closing in. So it's the I guess very good.

SPEAKER_02 30:35

Well, uh with that, I think we're wrapping up. Thank you, Richard. Thank you, Mark. This was a great conversation. Thank you very much.

SPEAKER_01 30:41

Thank you, thank you. Thanks for wondering. Good job, good job.