Episode Transcript
[00:00:03] Speaker A: Welcome to AINA Insights where prominent leaders and influencers shaping the industrial and industrial technology sector discuss topics that are critical for executives, boards and investors. INA Insights is brought to you by INA AI, a firm focused on working with industrial companies to make them unrivaled segment of ONE leaders. To learn more about INA AI, please visit our website at www.aina.AI.
[00:00:40] Speaker B: Hello and welcome to another episode of Titanium Economy podcast series hosted by aina. Today we are excited to be joined by Dong Soo Kim. He's a seasoned venture investor with a rich track record across corporate VC. Dong Soo is the CEO of LG Technology Ventures, the corporate VC arm of LG Group and manages nearly 900 million in assets. The firm's portfolio spans a diverse range of verticals including AI, biotech, clean tech, consumer enterprise technologies, and they're all focused on building the next wave of industrial and technological innovation. Dong Soo's career began at Samsung's corporate VC arm, followed by a stint at Asia Evolution and then he later joined Samsung Ventures America where he invested in over 30 plus companies. He holds a Bachelor's degree, light Physics from Caltech and a PhD in electrical engineering from Princeton. Dong Soo, we are really excited to have you on the show and thanks so much for being here.
[00:01:36] Speaker C: Well, thank you. I'm really honored to be here.
[00:01:38] Speaker B: Perfect. So what I want to first start with is you have a very academic background with you know, degrees in physics and engineering. What prompted you to go towards investing and that early in your career?
[00:01:50] Speaker C: Sure. So somewhere during my Ph.D. i was, you know, became quite interested in where all this technology will go eventually and, and I was really wanted to see how the underlying technology becomes a product. Actually my PhD was in compound semiconductors. I studied indium phosphide related devices.
So yeah, so, so a lot of my friends ended up in academia, but I decided to join at that time Samsung Electronics as an R and D engineer and, and I was developing fiber optic components. But somewhere along the line there was for me to do a career transition because at 2001 actually Samsung decided to exit the fib optic component business and I was offered to actually join the technology strategy team. So that gave me a chance to work with a lot of different startups and other companies along with the corporate venture capital team within Samsung. So eventually I make made the transition into the corporate venture capital arm of 2005. And so I've been doing this for almost 20 years now. Over 20 years now. And here I am.
[00:03:01] Speaker B: No, that's, you know, that's actually a very interesting career that you have taken. Right. Transitioning from academia to, you know, more research and then to a, you know, corporate VC side. And you have spent like 20 plus years in corporate VC, right. What excites you about this space compared to a traditional VC or a private equity.
[00:03:19] Speaker C: Yeah, so I think there are several advantages of being in corporate VC versus being at a purely financial vc. I mean I've known actually a lot number of ex colleagues who from the CVC who decide to migrate to vc. And again, I think the biggest advantage is that if you're a vc, especially at a partner level, you're constantly fundraising and it takes a lot of effort. Whereas for corporate vc, I mean yes, it does require some effort to justify, you know, the investment and all that. But. But at the end though the effort is a lot less. So you can focus a lot more on just the investing itself and then managing the portfolio. So that's one advantage. The other advantage is that you can really provide a differentiated value in that you can connect your portfolio companies with the right people within the R and D center or the business unit and then promote the partnership that accelerates the growth and you know, which can lead to even better financial return.
And also that strategic benefit that you can provide sometimes allows access into a very competitive deals because this company sometimes wants to do business with LG for example and so they will give us some allocation to invest.
[00:04:32] Speaker B: No? That's very interesting to hear. Given that we are talking about the LG technology venture, can you give us a high level view of the LG technology venture and what are its core strategy behind the fund?
[00:04:42] Speaker C: Sure.
So we started in 2018, so, so we are just over seven years now. Before then LG, different business units was just investing out of the balance sheet which wasn't working too well for various reasons. I don't want to get into this a whole new sec, you know, talk but, but so, so actually I was recruited out at the time I was with Samsung Ventures. I was actually managing U.S. europe and Israel offices first for Samsung. And they approached me to come in and set up and run it. So I had a very easy unique experience of setting up a corporate venture capital team from scratch.
And again from day one I really wanted the corporate venture capital team to work very similar to a financial VC in that we raise funds and we look for financial gain on every investment and so forth and we can lead rounds, sit on boards and all that. So again, I think at least in my opinion that's a model that not just provides the best financial return but also provide the most strategic value to lg. Mainly because when you're investing in these very companies that you think is going to make a financial return, these are the companies who are actually growing fast. They know the market, they have a good technology.
Right. So it's precisely these kind of companies that can provide most value to lg. Right. So contrary to some, a lot of people's belief, I do believe that financial gain and, and strategic benefits, they go hand in hand. So I, I try to achieve both.
[00:06:18] Speaker B: For me actually that leads to the next question, right? Because a lot of the perception outside is, you know, traditional VC is a lot more about thinking about the, the financial returns versus a corporate VC is thinking a lot more about alignment with the broad priorities of the parent organization.
And seems like your experiences, you know, it' or the other, but you can actually drive both. Can you elaborate a little bit on that?
[00:06:44] Speaker C: Yeah. So in the especially like, even like 20 years ago when I was first getting in, the CVCs weren't as sophisticated and everyone was saying oh yeah, we're just investing for the strategic benefit, financial return is not important. And what happened was that as a result they were putting a lot of strings attached with the investment for the strategic benefit. Either exclusivity or free IP licensing, you name it. Right.
And as a result that caused two things. One was that because of the strings attached a lot of good companies didn't want to receive the investment from CVCs because again if you're growing and if you have a lot of investors lined up, why should you be taking money with a lot of strings, right?
So, so a lot of CVC ended up investing in companies that couldn't get funded otherwise. And that eventually led to not a lot of strategic alignment because these companies eventually couldn't get outside funding and then the survival rate was very low. So that's one. And two was another negative outcome of that was that the CVC started getting very bad reputation as difficult to work with, slow decision making process and, and just not helping the portfolio but just looking after the interest of the, of the, of the corporate. Right? So, so again that also made it difficult to invest in the best deals because of that. So all in all I think people who have done it for a long time or at least the CVC needs to survive through the ups and downs in the industry learned over the years that yeah, I mean really the right way to do it is just, look, just go after the financial gain first and then try to create strategic benefits afterwards.
[00:08:29] Speaker B: Perfect. And I think given that you had a chance to set this up from get go, especially at the lg. You probably sort of took a lot of that learning and experiences of running it previously and bring it to bear in LG Ventures, right?
[00:08:43] Speaker C: Yes. Again. So all the experience within SAMHSA when I came to LG tried to take the best practice and then also made some changes to make things even better, at least in my opinion when I joined lg. And one of the few changes that I made, for example is made to headquarters here in Silicon Valley.
So the investment committee its also within LG Tech Ventures. And because we are here local and we can make a quick decision and also understand the Silicon Valley culture and all that. So that's I think the biggest improvement that I made. Another improvement I made is we have a business development team in house here also. And so we brought there the business development team's role is to connect the startups with the right people within LG and help with the collaboration.
And again I brought these people, actually hired them out of LG and brought them here to Silicon Valley so they can work side by side with the investment team and also if needed, interact with the startups directly. And I think that also made the business development effort that much better.
[00:09:47] Speaker B: Perfect. No, that's interesting.
I'm going to travel a little bit back, you know, to your Samsung days where you used to do a lot of investing within the semiconductor space. And I know you have done like, you know, a lot of investments in that, that, that space. Right. And starting with fabulous companies.
How do you assess an early stage opportunity? Right. When the, you know, when the sort of IP is limited, the access to fab is limited. How did you used to do that?
[00:10:11] Speaker C: Yeah, I mean even 20 years ago when I was starting out, fabless semiconductor company was just very difficult to invest. Even back then the cost for doing a tape out and turning a chip, it just has becoming very expensive.
So at the end, I mean you really have to think how big of a market are they going after? Right. If it's just a very niche component that has some customers, that's probably is not a good kind of as a venture capital investment it really has to be you're going after a fundamental processor or, or, or you know, so that's you know, just a big market. And then also you have to look for teams that knows how to execute. I mean making a chip is just very difficult. It's. And you cannot make mistakes. I mean after tape out maybe you know, if, if you, if the chip is unsuccessful then you have to do a wrist spin and that again it's a lot of money. So, so we look for teams who, who have experience building chips. And so those two I, I think would be the key criteria in addition to all the other things that, that you look for when you make an investment.
[00:11:22] Speaker B: Yeah, now that's actually very helpful. And I mean these days semiconductor definitely is very hot. Right. Because of all the AI, cloud compute, data center related things.
Do you think the, the valuation within these sectors are justified or are they like, you know, overhyped right now? What, where do you, where do you, what do you think about that?
[00:11:43] Speaker C: Depends on what kind of product. Definitely. I mean AI every now is hot and some, you know, AI chip companies are, you know, becoming, you know, a bit overvalued actually we invest in lg, I mean test Torrent, which I think is also very rich. But we did it because it's just very, very big potential market and is a team that really knows how to make chips. I mean Jim Keller is almost a legend in the semiconductor business.
But otherwise I think there are other sectors other than AI that is more reasonable. Right.
That begin to make sense. But again it's just still a very tough field to invest. You will have to have a really compelling product for us to consider an investment.
[00:12:28] Speaker B: Actually then thinking a little bit about that given that there's a lot of so much growth driven by AI beyond just the model and the compute layers where the leaders are already emerging. Are there other parts of the AI value chain where you think it is? You know, they offer more upside.
[00:12:44] Speaker C: Yeah, so, so you mean like semiconductors?
[00:12:47] Speaker B: Yeah, semiconductors.
[00:12:48] Speaker C: Yeah, yeah, yeah. So other than, I mean people, they're, I don't know, dozens of companies. Right. Who wants to be the next Nvidia and going after the gpu either TPU or you name it, I mean there's all these different varieties. I think those are, has been getting a lot of attention but it's very competitive. I think so many companies fail trying to overthrow Nvidia.
One of these days somebody will come up, but I don't think it's going to be five, 10 companies. It'll be only maybe one or two. Who, who takes market share away from Nvidia at least at this point. But there are other sectors like you know, like network interface related with the AI, you know, servers or memory related. And there seems like all this wherever there's a data bottleneck and there are people are trying to, you know, address the problem. It almost reminds me of like, you know, 15 years ago when I was investing quite a bit on the, on the memory interface. So I invested in Infight for example. Right. That for the dram and then also looked at a few companies on the solid state drive sector because again, I mean just it wasn't just the memory and the processor, but the link between that really required a lot of both semiconductor and software. So I think a lot of people are thinking about those problems now. I mean obviously Nvidia has that the Mellanox solution and, and again that's another reason why a lot of servers can move away from Nvidia. But there are people who are trying to address their problems too. So I think there are pockets there of interest.
This is not a semiconductor Fabless or anything like that. But again there's chip cooling and heat management has becoming an issue. So we've been looking at quite a bit. Packaging is also another big problem. So. So, so we are also looking for interesting comparison on that sector too.
[00:14:37] Speaker B: No, and I heard, yeah, those are actually some of the very hot sectors. Right. Especially given the amount of energy consumption that is needed in some of the data centers.
More efficient cooling and a better packaging. Right. It becomes so much more important now.
[00:14:51] Speaker C: Than it was ever before.
[00:14:53] Speaker B: Yeah, perfect. Moving a little bit to AI and because I know like at least in LG ventures, that's one of the core things to focus on within the AI value chain. You know, of course, other than the model, what are the other pockets that you think have a lot of upside?
[00:15:11] Speaker C: Yeah, so I think actually even a few years ago models were getting a lot of attention and there's also a lot of tools and all that. I think that has matured quite a bit, at least from a venture capital investor side. Now there seems to be a lot of new emerging applications of the AI, whether it be speech or enterprise or you name it. I think that's becoming very interesting partially because the technology itself is mature, but also the cost of large language model has come down quite a bit. I mean they are now open source models and even if companies who wants to build a very small, very dedicated model that are relatively small, then those cost us also come down quite a bit. So, so I think it's now a matter of just using the technology and then solve real people's problems. And for example, we invested in a company like Cresta Labs, they make AI solution for call centers. Right. You know, if you all these big companies have these call centers which are, which are very costly. I mean all these phone operators, you know, it's just, you need a lot of them and also it's a very stressful job. There's a lot of turnover by using AI to facilitate their work. For example, for Crestlight Lab solution would listen into the conversation and suggest for example some model answers to the telephone inquiry. You would also record the calls, catalog it and also fill out the forms that the operators need to fill out after each call and all that. So it just makes things a lot easier. And because these people know the contact center business very well, they have quite a bit of experience that they, they, they, they can address their, their customers problems. They, they know which, how to sell to these enterprise and all that. So it's just going back to the original point. I mean if you know certain business vertical well and then you're using AI to solve those problems, I think those are the best companies.
[00:17:11] Speaker B: I think what you're saying is like, you know, basically really using AI to solve a specific and building a solution around it is essentially what are the kind of things that you're looking for and that's probably where the next set of value is.
[00:17:24] Speaker C: Yeah, yeah. At the end of is providing a return on investment for your customers. Right.
And some of these enterprise solutions, I mean you can show that even now. Right? Right away.
[00:17:37] Speaker B: That's perfect. Actually. On the flip side then if I ask you a question, are there parts of the AI ecosystem on the value chain where you believe, believe are is overhyped and sort of overvalued?
[00:17:46] Speaker C: Yeah. Where do I start? There's so many. Right. I think a lot of these consumer related applications like video generation or you know, those are, I mean of course it's a big market but it's. The barrier of entry is not that high and it's just so competitive and also they're so overpriced that I think most companies would fail. Right. It's just very difficult. And plus consumers are. The willingness to pay for these kind of applications is very low. I mean people are used to getting stuff for free.
So I think yeah, those are really hard large language models. Yeah. I mean these companies are valued billions. Right. But how many large language models will survive? Right. I don't know.
I don't see more than five surviving because it's just very general solution and they're all competing against each other and just costs a lot of, I mean it takes a lot of investment to even maintain and run it. We'll see.
[00:18:51] Speaker B: No, that's perfect. And I think it's sort of what you're pointing to is it's like more like in a B2C way we really solving a tangible problem for a company. Right. Like that's where the value is. But for B2 sort of B2C it's you know, not that much because the willing to, willingness to pay is still a bigger.
[00:19:09] Speaker C: Sure, exactly.
[00:19:11] Speaker B: No, that's perfect. And, and I know you're making a lot of investments especially in the AI space. Right. Like what are the kind of specific traits or fundamentals do you look for other than some of the things that we talked about. Right. And making in evaluating a high quality investment.
[00:19:27] Speaker C: So I mean again depends on the sector. I mean if it's application I already mentioned about that particular vertical and all that. I think if you are more of a platform then I think it's just again I mean it's just where the customer pain point is. That's always comes back to it. If you're solving a problem that is good to have then it's very difficult. You really have to have a solution that is a must have that without it you really, the customers can't, you know, can't stay in business or just it's something that makes them so much more efficient or something like that. So again that's a solution that we're looking for. And then, and there are, you know, not that many companies who are, you know, who is a must, has a must have solution because at the end you know, you have a lot of okay, this, this is good all that. But it's at the end if you think about it, it's just nice to have solution. And so you just have to filter out a lot of these kind of companies.
[00:20:28] Speaker B: No, that's very helpful and insightful. One last question that I'm going to ask you is there's a lot going on in terms of the macro trends and shifts and what are the key things that you pay a lot of attention to especially when you're trying to make an investment or try to figure out where the next opportunity is.
[00:20:47] Speaker C: Yeah. So again I mean we are still a strategic investor after all. So we look for companies that can create synergies with LG in one form or the other. Whether it be a joint development or it can be a strategic supplier or it can also be a potential customer and all that. So, so, so we look at those kind of companies and, and then, but we, we have the flexibility of also looking a little bit into the future. Right. If you think, if it's a company portfolio that we think is, can, can be relevant to LG a few years down the road, then we can also still invest and then wait for the company to mature a bit before trying to create strategic value for lg. So we can do that. If you are asking about what kind of sector other than AI that we think is interesting, that's going to be big in the future, I think is electrification of energy. So everybody is aware of EVs and all that, but it's just that I think EVs are just the start. I think even at at homes, I mean there'll be more electrification including, you know, heating will become more, you know, use using more heat pumps and people cooking with induction cooktops and all that, not just because of the environment, but because it's just these kind of electrical solutions are a lot more safer, cleaner and more efficient too. So and same thing will happen in manufacturing, a lot of things that are done by with using fossil fuels. It will be gradually transforming to use electricity as energy. And this means that there's quite a bit of infrastructure that needs to be upgraded. Not just generation of electricity, but also transmission, storage, management and all that. And so I think there's a lot of interesting play there. Some companies who do solve some of these infrastructure problems will really end up a very big market.
[00:22:36] Speaker B: Oh, perfect. No, it was really nice chatting and connecting with you and thanks a ton for sharing your thoughts. Really appreciate it. And thanks again for being here with us.
[00:22:45] Speaker C: Yeah, it was a pleasure. Thank you very much.
[00:22:53] Speaker A: Thanks for listening to AINA Insights. Please visit AINA AI for more podcasts, publications and events on development achievements shaping the industrial and industrial technology sector.
