Venture Capitalist on the AI Frenzy: What Comes Next?

The pace of innovation, the pace of change with the AI bubble is so extreme that adjusting our workflows, thinking about this on a daily basis can be exhausting. And I think for a lot of companies, executives, and employees, it can be somewhat exhausting. I think if we actually saw a bubble pop and, you know, things cool back down, I think a lot of people would be happy. Our Bay Area innovator today is Dan Butler, Managing Director at Tamves Ventures, a longtime investor at the intersection of software, fintech, and emerging technologies. In this conversation, Butler shares how he evaluates companies in a rapidly shifting market, the concerns around the potential AI bubble, the race for AI, and robotics leadership, and much more. This is Bay Area Innovators, and I'm your host, Steve Esparza. Don, welcome to our program. It's great to be here. Great to have you here. So, you've been an investor for over 25 years, which is a long time in the Bay Area, and you've kind of seen it all, from the dot-com crash, to pandemics, to the 2008 financial crisis, and now we're into the AI era. Now, the first thing I want to ask you is, like, what's the most satisfying to you as an investor? Is it finding the company you've been looking for, the company going public, working with them? Yeah, certainly. I would say, you're right, that having done this for a very long time, it's been quite a fun journey. And, you know, usually the things that are—the companies that are the best investments, the ones that you enjoy the most, are ones that you took a bet on, where most people thought you were wrong on the investment, and you actually turned out to be right. And if you were really right on the investment, you wind up seeing that company go public. And pretty soon, something that felt like an idea that a lot of people questioned at the beginning is validated. And then you see that validation extend all the way to, you know, ringing the bell on the New York Stock Exchange, or the opening bell at the NASDAQ, and suddenly that now is sort of like an idea that has become mainstream, that everyone agrees that was a good idea. And I would say that, plus, seeing the people that we backed at the very early days become successful. You know, I didn't sort of appreciate that second-order sort of delight in the business, but I think, you know—and I've seen that go from people who've gone public, taken their companies public, and continue to operate them, to, you know, one of the—an old friend whose business we led the series A for, his business was acquired. And he had been working in the technology industry for maybe 25, 30 years. And when I asked about the closing dinner after this acquisition, like, well, you know, his name was Ted. Ted, what are you going to do now? And he said, well, I'm going to buy a house in Tahoe and a house in Hawaii, and I'm going to go back and forth during the year. And, you know, you sort of see that happen to somebody who struggled through all the ups and downs of a startup. And to see that kind of outcome at the end is incredibly rewarding. That's a very interesting point about not going the same direction when everyone is. And, of course, you got to relate it to AI, because everyone now goes into the AI direction. And how do you differentiate that? How do you choose the ones that are different? So I would say at our firm, there's a kind of a couple of ways to look at it. We actually internally, we developed a framework to actually evaluate whether or not we want to jump into an industry. I would say the framework is we look for established industries that tend to have very good profit margins that were based, where the incumbents were based on older technologies. And then where the incumbents also have a very low, what's called a net promoter score. It's sort of like a, you know, it's like your Yelp rating. And so if you have somebody who has a good market share, is very profitable, but a lot of consumers are not that crazy with the service, like that's usually a good place to go with the tech innovation. I would say when it comes to AI more specifically, you know, I'd say we have seen the investment in AI kind of go through a transition. So I remember there was a time probably in 2018, 2019, where I saw people starting smaller seed funds that were AI focused funds, maybe like a $50 to $100 million fund. And what happened there was, you know, people were doing that and were kind of doing that in the face of looking at history. And history had shown that there was, gosh, basically three periods of AI investment and then AI nuclear winter. And people said, you know, oh gosh, how do you know this time is the right time? And what we've seen with AI is, I think from those early days in sort of the, you know, what would have been more called like machine learning or data and infrastructure investing to now the AI investment boom, what's happened is we've seen the investment environment move from one where it was sort of like, gosh, I don't know if this is a good idea, you know, we've seen these past booms and busts in AI, how do we know this is the right time? To now it's the established wisdom. And so what that has led to though, is that now that everyone knows that it's a good idea or that it's the next big thing, everyone's piling in and that tends to diminish the returns because now prices for the companies we're investing in, the valuation goes up because now everyone knows it's a good idea. And so, you know, a good framework for us has been, we look for conventional wisdom about an industry or about a company or about a product that is no longer true. And it's conventional wisdom because everyone knows this to be true. The reality is the technology is going to disrupt this and make this very different. Well, let's take a step back and go to where you started your career. So you actually went to UCLA, did the major in Chinese and then Stanford political science. That's right. How did you end up in tech? You meet interesting people who show you an area that you didn't know anything about. And so when I was at Stanford, my roommate at Stanford was a PhD in economics who had just come off of working for the Federal Reserve as an intern. And, you know, having dinner with him at night, talking about what he was focused on while I was studying political science was fascinating. And it made me think, OK, I should go to go to learn about Wall Street and try this. Like, this is too interesting to pass up. So I wound up working at Lehman Brothers. When I was at Lehman Brothers. From Stanford, you went to? Yeah, so I went to Stanford twice. And the first time I went, I went to Lehman Brothers in New York after that. And then the second time I came back to work on political science. And while I was doing that, Internet bubble was starting to happen. And, you know, I said, you know what, this is this has been too many interesting things going on to ignore this revolution. And that's when I was introduced to Peter Thompson, the founder of Tomvest. I never thought I would work anywhere more than five or six years. Peter is a phenomenal gentleman. He's been awesome to work for. And so, you know, I wake up and it's been 25 years. Are you into car racing as well? You know, I know Peter is. I've had the pleasure of being an observer or sitting in the navigator's seat as he sort of did a test on the track for a rally car that he has. Monterey here? No, this was up in Canada. And, you know, it's funny because he he has done over time rally racing. You know, it's been phenomenal. But also, I don't participate the way he does. Right. So so he's involved, of course, with Reuters. And is there any connection in that sense between journalism and news and tech? You know, not a formal connection, I would say an informal connection. So over the years, we have gotten to know different people through him at Reuters. And what they've been looking at is how technology is impacting reporting and journalism. And so from time to time, we've we've compared notes on tech trends we're seeing or sometimes because of because of the connection to through Peter to Reuters, we've sometimes had people come to us with new technologies that might impact journalism. And so oftentimes we sort of said, well, this is very interesting. But, you know, this is not where we're investing, per se. But, you know, to the extent that it might be interesting, happy to connect them to folks over at Reuters or vice versa. If they have questions on vendors, sometimes we've been able to sort of give them our views on what we're seeing in the valley. Now, let's talk more about how you choose companies to invest in. And what I'm thinking now is there's the markets are changing so fast and usually investment could be years, could be a decade before you get your money back now in a decade from now. Who can imagine what what would the world's going to change? Yeah. How do you make the determination when you when you look at companies? What we found, I would say there's an internal process and then sort of looking at the company, the internal process. We decided maybe about 15 years ago that we should become experts on certain industries. And so we decided to go very deep and pick a subsector of areas that we would go after. And then we do a lot of research in those areas. The research for in those areas typically means we're not just doing sort of looking at the research that's been published or data sources we subscribe to, but also going out and actually meeting with customers, meeting with vendors in the ecosystem and then trying to figure out, OK, typically in an industry, there's like a frontline of innovation happening where is that frontline happening? And therefore, when we then flip over to looking at companies like in cybersecurity or fintech, where do those companies fit on sort of the frontline of innovation? And that's helped us quite a bit as we've selected companies and then also as we have helped them grow their business. I would I also say for us, the criteria that we look at, we actually developed an internal scoring system for how we evaluate startups. That scoring system has evolved so that we have an early stage framework and a late stage framework. And I would say the early stage framework, if we're looking at a company that's just getting going to like a series C or a series A company heavily weighted towards the team, because oftentimes the market might be brand new. There might not be really that subsector of the market might be brand new. There might not be any proof points on the technology. Revenues might be zero or very low. And then you're really betting on the entrepreneur and the team that that entrepreneur has has started the company with. And so in the early stages, we're heavily looking at the team. And then as the company gets later, we start to look more at sort of the the company as sort of an institution. How good is product development? How good is engineering? How good are the teams that the founding team has have actually assembled around them and how much can they make an impact on the world? One other observation I'd give from the companies that we've seen that have been the most successful, that have gone public, that have done things is that usually the the most successful teams are also driven in an incredible way and driven to, you know, to the point where they're not willing to accept the status quo. Like they're so passionate about what they're building that they're willing to sort of take the risks of raising a bunch of capital to go change the way the industry works or change sort of what, you know, consumer behavior sometimes even. And I would assume that you're looking at the team as a strength. And then if markets change or technology or or anything changes, the team can change as well. They're not going to. The best teams can change as well. Yeah, that's that's, you know, I think I heard a phrase from Peter one time that, you know, the best teams, if they're confronted with a wall, like they're locked in a room, they're going to break through the wall. Like, you know, they won't take no for an answer. I would say one real life lived experience. We had a phenomenal team in a cybersecurity startup. This was a company in Boston, and they had raised we were part of a series A. They had raised a series B financing. So probably at the time they may have raised like twenty five, thirty million. And I remember going to a board meeting and they had warned us ahead of the board meeting. We get to the board meeting and they say, you know, we ran into a technical wall. Like literally there is a there is a this technology for what we've been building, what we're trying to do is not going to scale. And the team had the maturity to come to us and say, everything we've built up for the last, you know, two and a half years isn't going to work like it's not going to cut it. And so they sat with the investors and said, so we have two options. Option A, we give you the remaining money back. Like we can say, we're sorry, this didn't work out. Here's, you know, half of what you invest in coming back to you. Option B, you can keep that money on us. We might need some additional money on top of that. And we're going to start over and go in a slightly different direction, but slightly different direction was we're going to rewrite our technical stack. And so in some ways it was almost like a restart. And I'd say out of like four or five investors, I think one of the investors said, I want option A. And so we we paid them their their money back, basically. The other investor said, you know, this team is so phenomenal that we even within the missteps, we think they're the right team here. And the crazy thing is, about 18 months later, we sold the business. It was a great return for us. Great return for the team. And what had happened was that they hit on a sort of an innovation of how to do what they were trying to do. And that innovation was scalable. And suddenly they were like in the right place with the right customers or, you know, people they could upsell and with the technology that actually could scale. And and so they got acquired by Cisco. Yeah, it was a great outcome for us. And, you know, but for me, that's ingrained in me the lesson that the right team, even in the toughest of times, will find a way. Can you give us some examples of some of the more successful investments that that you had that you invested in and they they did well? I think we've had a mix of companies that have gone public and then companies that have been acquisitions and like very large acquisitions. I would say, you know, recently we had a we were part of the series A for a company called Figure Technologies that went public. We were investors in the same entrepreneur. Mike Cagney started a company called SoFi. We were part of SoFi in the early days. You know, if I look at another company, we were part of it. There's a company called Vungle. Vungle was acquired by Blackstone, and they have now filed to go public as Liftoff. Well, when it comes to AI, we got to talk about the potential bubble. Yeah. You know, there's everyone potentially looks at it. You know, there's reports one point four trillion dollars investments versus 13 billion dollars revenue. What are your thoughts about this potential bubble? Is it is it more of a question of if or when? I think it's more of a question of when it's interesting because we've been looking at the comparison to this bubble, to the dot com bubble. And because we live through the dot com bubble. When I when I joined Peter Thompson and Tom Best, it was February of 2000. And so right before the bubble sort of popped. And at that time, we had a portfolio of about 20 companies. And so we sort of lived through that. And I would say the, you know, it's interesting the amount of perhaps the problem with the bubble right now is that so many people recognize that this is a once in a lifetime moment. This is a once in a lifetime opportunity to invest. That then everyone is rushing into this area. You know, what you find, though, is that, you know, I've found historically that whenever that happens, pricing goes up and valuations go to extreme heights. And they tend, you know, out of the mix of companies, there will be one percent of the companies, two percent of the companies that will be the next Google's or the next, you know, Microsoft's of the next generation, like absolutely. Those companies are sort of up and running and getting invested in. But then you're going to have a lot of other companies that were sort of, you know, four, five, six after that. And, you know, a good good way to look at this. I remember in the when I started with Peter at Tom Best, we were looking at the antivirus space. So if you have antivirus software, McAfee and Symantec and all these. So at the time, in early 2000, there were twenty seven vendors, you know. And we all remember, you know, McAfee, you know, Symantec, maybe Trend Micro. Yeah. You know. But then after that, I think that's part of Symantec now. So but then it's like, OK, what about four or five, you know, number four through twenty seven. Right. And so I if history repeats itself, as it often does, I would expect that we'll see that phenomena play out in the bubble that we have right now. The one thing that I didn't think was interesting when we were comparing today's environment to the dot com environment is that while today's environment is fever pace, like a fever pitch in terms of pace of investment, dollars of investment. Actually, some of the multiples are not as bad as it got to. And I'll give you an example. We were looking at this as of like a month or so ago. We were looking at the price to earnings multiples of public companies in the tech sector. I think it was somewhere for like the top 20 companies or something like twenty five to twenty seven times earnings in the dot com era, though, that was up to like 40. And so what you find is when you look at some of the metrics, it definitely feels like there's overinvestment in the sector. It feels like there's going to be too many companies funded. But the sort of the fever pitch that we hit in the dot com bubble, we're not there yet is basically the gist of what we're saying. And there's still some room. I think there's still some room. And then on top of that, too, this time around, we have the revolution being led by the biggest of incumbents. So a lot of the capital infrastructure that's being capex that's being incurred is being incurred by the meadows of the world or the ghouls of the world. So these are companies with extremely healthy balance sheets. And so I think just as it does not appear as kind of overvalued as it was in the dot com time, that means that if something were to correct, I don't think it would be as bad as it was. I also think there is a funny sort of social observation that I would give to, which was the the pace of innovation, the pace of change with the bubble is so extreme that I find adjusting our workflows, thinking about this on a daily basis is can be exhausting. And I think for a lot of companies, executives and employees, it can be somewhat exhausting. I think if if we actually saw a bubble pop and things cool back down, I think a lot of people would be happy. You know, I think a bit of the angst that you might feel because of the rate of change might be tempered if that were to happen. And so I don't think it's quite as like I feel like it wouldn't be the worst thing to happen if we saw sort of a pull off on the level of investment and sort of the bubble now. Very interesting perspective. Now, let's talk about job displacement. There's a lot of talk. I think I remember this conversation between Jensen Huang and Elon Musk. Yeah. Saying that, you know, as I develops, they're going to be busier. So, you know, and and I can see that because there's so many ideas that can be put into practice. Yeah. So they'll be busier. Maybe hundreds, maybe millions of people will be busier. But what about the hundred millions of people who are maybe the factory workers, the cashiers or, you know, the logistics? Yeah. You know, and we've seen in past sort of AI or past technology platform shifts is that usually, you know, there's a new set of rules that are sort of that that are created. And I don't think we've yet seen that. You know, there might be prompt engineers. There might be things like that. But, you know, if you look at after sort of the the mobile revolution, you know, we nobody anticipated that influencers were going to like, you know, individual influencers were going to be like a profession that, you know, you would generate a lot of a lot of income off of. Like, you know, we didn't foresee that where today I think today's youth will look at that and say, yes, of course, you know, I aspire to become an influencer. I think we'll see the rise of jobs like that. But we just haven't really seen it yet. And so I'm optimistic that we will find a way. At the same time, I am a little bit nervous that the the rate of change is so dramatic that it will probably lead to more job displacement or automation than we've seen, you know, in quite a long time. I would give another comparison to the dotcom era. It felt sort of like it was an equalizing force that I remember seeing, you know, driving down the 101, seeing a plumbing truck go by like a plumber with his web address sort of stenciled on the side. And I thought, oh, my gosh, you know, like everyone can then put their address and like have their own form of advertising. And like and anyone on this, you know, on the freeway, they would see that. We know I can go there and see who he is. And, you know, maybe if I need something, I. And so it felt kind of like it created a level playing field in terms of awareness and information this time around. It feels like there's more of a split where people who have either capital or who are working in the right companies are definitely going to be better off. But, you know, I don't feel the same level of kind of kind of like equal, equal sort of distribution of opportunities that we saw in the dotcom boom in this one. It feels like it could be more impactful on a lot of a lot of the work that we do today. And I know, you know, even I'll tell you a story that I heard from a friend of mine who works at an insurance company. So his company had employees working in a call center. I think he said it was about 300 employees. They wanted to start using AI to do call center support and using voice bots to handle this. And so they said, OK, we're going to take these the staff and we're going to retrain them. And let's think about what where we can redeploy the staff as opposed to letting them go. They started retraining them into outbound sales. And then, you know, six months later, along comes outbound sales sort of voice chat bots. And they're like, wait a minute, you know, maybe that's not the right place to retrain them to. And I feel like we're going to be a little bit of this, you know, sort of, you know, hopping from one one island to another as we sort of figure out what the next set of roles will be for people. Like, I feel like that's coming, especially at the at the kind of middle class level. Tell us about some of the companies. What's the percentage of companies you invest in that are fully AI focused? I'd say of the newer investments, probably like 70, 80 percent. We also I find even the ones that are in the remaining 20 to 30 percent, they might not be fully AI focused, meaning they're not necessarily building AI infrastructure, but they're leveraging AI to deliver something new into their own areas. So we invested in a company that is a company called Baseline in New York. It's a neobank basically for people who rent out homes, rent out properties and leveraging sort of AI capabilities to handle transactions on behalf of the people that are their customers. And so but of the 70 to 80 percent, those those tend to be in the AI infrastructure build out. You had an example about investing in you try to diversify. Yes, that's right. From the eye. What happened? Yeah, that sort of goes to the goes to the scale of the impact of the investment wave. So our our we have a we're part of this broader family office in Toronto. The family office there, I think, was trying to diversify away into other areas both to balance sort of the the mix of investments and mix of risks. And, you know, one of the things that was sort of remarkable was that they had invested in a fund that basically invest in farmland. It's like a farmland investment fund. And, you know, the other day it was funny because they got a distribution coming back from that manager and they sort of say, hey, we didn't expect to see this much capital coming back. What happened? And it turns out one of the one of the farms that they own was actually bought to put in place a metadata center. And so, you know, it's like one of the new data centers. And so even when you try to diversify, I kind of feel like it comes back. And, you know, I think for asset managers, finding something that is not correlated to the AI wave probably is becoming a little more challenging just simply because it's you know, this is I would say in my lifetime, this feels like the biggest platform shift. I would even put it above the sort of the dot com era. So what happens when things don't go well? Yeah. You know, as far as the investment went the wrong way. Yeah. But, you know, I'd say just to sort of if you step back and look at the the the industry and the way the industry works, probably especially if you're investing in an early stage company like the seed level, 65 percent of the time, it's not going to go the right way. You're going to end up losing money. So, you know, yeah, you're expecting that. I think the you know. And so it's it's very much a part of the business. One time we did a study. And so this is dated, but I thought this was fascinating. We were looking at the investors who had the most IPOs and investors who also had some of the most write off companies that had failed. And I was very surprised to see that some of the best investors into Lincoln Valley topped both lists. And it's like, OK, they truly are taking quite a bit of risk with quite a bit of capital. And, you know, on the early stage investment, maybe the majority of these don't work. And but it's being right that one percent, two percent of the time where you get a company that has a true power law return. You know, you invested in the next Google or the next Uber. So that's what this industry returns really driven on. And so I would say on the one hand, it's the nature of the industry we operate in. Now, where we're as an investor, it's easy to say if you're an employee of the company, then it's much more kind of concerning because, hey, what happens if this company doesn't work out? What do I do next? And so what we've tried to do, and I think we have a very good track record of this is when we have something that is not working out, oftentimes sort of working with a team and saying, OK, how do we find a good home for this company? Like instead of simply writing it off, walking away like if they've built something interesting, is there someone who would be interested in acquiring the company? You know, and so and I found, you know, when I've looked at the investors, we've had not everyone does that. A lot of folks will say that's the business move on. I think we because we're investing the house's money, as it were, we on the margin tend to maybe care a little bit more like a smaller volume of investments. But the ones we invest in are very high conviction and we want to see those companies through. And so I think oftentimes, you know, we have stuck with companies, stayed on the board, seen them through to some sort of exit. And so but that that's I find that's a little more of a firm by firm sort of decision of what to do when things are not going the right way. And by the house's money, I mean, it's Peter's family. That's the Peter himself. Yeah, Peter himself. Yeah. And so, you know, and so take take everything, every dollar very seriously. So put another way, we're not investing other folks money. And so I would say with that, we tend to be. And that's one reason why we have compared to most venture firms. I'd say we have we might be on par with them in terms of overall size, but we tend to have a not a high volume of investments. We're not trying to play the odds, as it were. It's more sort of falling out. Yeah. Selective and following kind of our industry focus process. How involved is Peter with all the investments? Is he aware of one? Yeah, no, he keeps on top of the investments. I would say we handle the day to day interactions. But certainly because of that, I'd say we're an important part of his overall sort of sphere of interest and influence. Tesla announced recently that they will stop building the Model S and Model X and instead they'll focus on humanoid. Yes. Yeah. The one robot. Now, you posted up a panel at Stanford to talk about robotics. Tell us. Tell us about it. So I was fortunate to moderate a panel as part of the conference was held by the US. I think it's the US Asia Technology Management Center at Stanford. This was maybe three or four months ago. And, you know, one of the things that was remarkable about that was the fellow who leads the center, a fellow by the name of Richard Dasher, had given a presentation about robotics more generally and sort of how we see this industry evolving. And I would say in tandem with that, we see a lot of interesting investments that are related to now AI sort of at the interface of physical and digital. So, you know, instead of just working on text or images online, now being able to interpret the room and take action accordingly. And so one of the things that came out of that, that I thought was sort of fascinating was given. And we had a lot of Japanese corporate venture capital firms there as part of this discussion. And one of the things we found was it does feel like there's an opportunity to pair up with the Japanese industrial infrastructure between sort of the US LLM vendors or other AI vendors to leverage sort of their industrial base to create robots that will be on par with the companies we see coming out of China. And so it does. I think Richard had put it in and gone so far as to say, you know, this feels like Japan's opportunity to lose. They're so good at industrial machinery. And, you know, I think that that's an industrial base that we sort of, I think, had lost to some extent. I do think, you know, folks like the Elon with what he's done in Fremont has been able to bring some of that back. But it does seem to me that they could be like a very good partner for us in this next era of, you know, inference and deployment of LLMs at the edge through robots. Now, what I'd like to figure out is how much do you or your team or, you know, investors in general, think about what's going to happen five, ten years from now as far as what will people do? You know, we have all these robots. We have this technology, this AI. And it's getting to the point, I think Elon Musk said that work would be optional in the near future. Right. How do you look at that? Like, will we have to find things for people to do? I mean, there's so much you can go hiking at Tahoe, right? Yeah, that's right. It's great. I love hiking in Tahoe. But, you know, can you do that every day? And if everyone goes to Tahoe every day, it's going to be pretty crowded. Yeah, that's right. Yeah. But then, you know, it's funny. So if you think about in a way, it may be a question of like, do we wind up with universal basic income as a form of income for people? Well, that too. Yeah. But there's more than that. It's more of how the people can enjoy their life. That's a, you know, I suspect the way we enjoy our lives will be 25 percent, 30 percent different than what it is today. And, you know, I think right now I feel like we're actually in the process of searching for what that's going to be, you know, partly because if you think about all the things you do during the day that might be sort of routine tasks to the extent that those can then start to be automated in some way. You know, I tend to have a bad habit. I wake up in the morning, I check my email on my phone. You know, if I have an LLM interpreting everything for me and say, this is all you need to worry about right now, you know, focus on this. Like, I definitely see that coming. But it's like oftentimes with technology, the shifts are unanticipated. And I would tell you, there's a very interesting study by published by Morgan Stanley. It was one of the research team at Morgan Stanley. And what they looked at was they looked at sort of how long it took for technology shifts to generate sort of the next generation of companies or employment. And what was really interesting was they found on average is about three years. So from the introduction of the iPhone to the point where you got like Uber started, because you can now think about ordering a car on your phone, Airbnb, because while you now you can book things on your phone, you know, that was about a three year lag. And it seems like the entrepreneurial ecosystem where the societies are broadly takes a bit of time to digest what the implications are of a new capability. And then founders say, oh, you know what? I think we could go build this type of company. And if you think about it, we're right at that three year mark since like the introduction of Chachi PT right around that time. And so I kind of feel like right now, probably some of the most interesting future companies are being built with that. Probably some of the future employment will be built. I would tell you another example that I've seen a company launch this week that I would not have foreseen, but it goes to employment. There's a and the name is escaping me, but I think there's now a couple of companies doing this. And basically what they are is that they are employment platforms like an indeed or something like that, but their employment platforms where an agent hires humans to go do things for them that they need done. And so if I'm an agent and I'm in charge of filing your taxes, but I think these tax returns need to be, you know, signed by a notary public or they need to be mailed at the post office and get the stamp of mailing in, then that agent can hire somebody on the platform to go do that work for them. And I saw, I think as of yesterday, there were already like a 150 people that had signed up on this one that just launched this week. And, you know, it was interesting because the reviews and I think they had already been doing some of the work in like a primate phase. The reviews were good. It's like, well, I get paid on time. I know exactly what I need to do, you know. And so I do wonder, like there might be forms of employment, you know, that we can't yet anticipate that that. And it begs the interesting question, too. You know, would I be happier working for an AI boss or a human boss? You know, you know, I suppose it depends on how the two are trained. Right. So, you know. And you gave an example earlier about the chat between the text message. Yes. Tell us about it. Well, and so, you know, one of the things that's been most fascinating to watch this week in particular is sort of the advent of a platform called Maltbook. And malts are basically small sort of bots you can create where you download oftentimes a small, large language model onto your computer. And then you have you can create like a little autonomous agent to do things for you. And so one of the examples that I saw from a friend was it's a physical device or it's just it's a software that you load on your computer. And the trick is that you can give it permission to do the things that you would do on your computer. So if you spend money through your credit card, if you make calendar appointments, if you send emails, you can, depending on what you give it access to, you can then tell it, hey, I want you to do these things for me. You know, if I if I tell it I want to go to New York for a meeting, I can then say, hey, you know that I like to fly on American Airlines. Please look at flights that are under this price point and go book my flight and it'll go off. And if it has access to your login credentials and your payment details, it can go do that. And so what we're finding is that Malt says they're sort of being referred to recently, you know, have started taking on behavior. That's kind of intriguing. And one of the examples I saw was there was somebody who I think his wife wanted him to text her during the day, say how his day is going. And then, you know, look at your calendar when you might be back home. And in this particular example, he set up a mold to text his wife and say, look at my calendar. Tell him and tell me how busy I am when I get it. And what we'd heard was several days went by. He then asked her, like, you know, my text messages are OK because he didn't tell her he was going to do this. Right. And her response was, you know, I'm really enjoying our conversations. And, you know, you sort of looked and I think at the moment he was like, wait a minute. You know what? I should I should be monitoring. Like, what is the little Malt saying? Because if she asks the Malt about, you know, should we take a vacation tomorrow? It might respond back. Well, you know, here's typical vacation options or something like, you know. And so I do think there's that. And then another thing that we've seen with recently is this platform emerged literally in like the last week called Maltbook, which is a place where those agents can communicate with each other or talk to each other. And so now, you know, another example of something I didn't think I would see this quickly. Yes. Any any final thoughts where we're going with investments and with AI? I'd say, you know, part of part of what's both exhilarating and slightly concerning sometimes is I feel like the future. Certainly, we sort of feel like we're hurtling down the tracks, but where we're going is not entirely clear. I think that's generated some some angst, obviously. But at the same time, you know, I tend to be more of an optimist. I think like we'll wind up having new services or new capabilities, kind of like staying at an Airbnb or using an Uber that, you know, the next generation of kids will take for granted that for all of us who are here today, we'll sort of look at and say, well, that feels awkward or that feels weird, like like being employed by an agent. You know, I suspect we'll wind up having things that feel new and different, but actually will be sort of played with, tested and then widely accepted by society. And so I kind of I do think we're in for a period of continued change. But on balance, I also tend to be quite optimistic about, you know, the things we can go to accomplish. So generally, I think there's there's always concern about the development of AI technology because of fraud. You know, there's so many things can go wrong. Yeah. But there's if there's one positive thing that I can think of is that people will be much more self-guided, more aware. And there's already a very young people who say like, oh, this video is so AI or yeah, or this call. I got it. It's it's so obvious that it's fraud. You know, when before you would be like, oh, yeah, I watch my credit card. You know, they called me. Yeah, yeah. Yeah, that's right. Now it becomes so obvious. Yeah. Yeah. At least for some people. So that's the one positive thing I can see that people will be more aware. More. And will that continue? I wonder. Yeah, I think, you know, it's funny because we because we do cybersecurity investing. We also see how clever hackers can be in terms of. I do think at the same time we can adjust to that and sort of overcome it. Like I would say our experience over the last 25 years is that, you know, there's an innovation like LLM's hackers learn how to exploit that and move very quickly to do so. Companies and then employees, for instance, or consumers learn how to adjust their behavior in response. And so it's kind of a back and forth. So in that regard, I don't think it'll be, you know, even if we get phishing emails that look, you know, somewhat questionable. I would say that we also, you know, it's funny that we have an investment in a company called Bolster. Bolster, if you see a docusign that is, you're not sure that's an accurate, accurate docusign or not. They actually power the service that docusign runs, where it's I think it's verify a docusign.com. You send that to them. They look at they come back and say, this is an actual docusign. You should sign it. And so that's an innovation that just came about not too long ago. And so I kind of feel like, you know, we will also find ways to overcome that. And I do feel like, you know, to the extent that some of the mundane work of today gets automated or gets taken care of with AI, that we'll begin to find we can spend more time on things that matter more or things that skew a little more towards like the higher level abstractions like art or, you know, start to I think, you know, if you don't have to worry about taking example of like, you know, the time that I might have invested in sort of taking checking notes from a call. Yes. Now I can, you know, use a note taking tool. It'll give me a very nice summary. I can learn both. I find I learn both more from the call afterwards. Like I go back and I study our conversation. I mean, like, OK, I should have learned this lesson in this conversation. But then also that that feels like I'm educating myself more holistically by having the note taking app sort of do that for me on each interaction I have with somebody. So I do feel like, you know, the rate at which we can sort of learn is probably going to be impacted as well. So in that regard, I'm, you know, on balance, optimistic. Now, of course, just just one last point is that that obviously applies much more for the younger generation. And, you know, if it's fraud or things that are, you know, difficult to figure out if it's real or not. Yes. When it comes to the older generation who are not as familiar or not as well, the acute that this is. Yeah, it's just much more difficult. It feels like it. Yeah. And then, you know, because we also do fintech investing, give you an example of a social response to exactly this problem. So if you think of fraud as me sort of impersonating you to a bank, then maybe I've gotten your credentials. I present myself to the bank and I try to do some withdrawal. Big withdrawal, something like that, right? So another thing that I've seen in the cybersecurity world and in the fintech world, people making a distinction is that versus what they would call scams and scams is I try to get my elderly mother to move money to me on out of her own account. So I've somehow managed to get to her. And she actually moves the money. And so technically, you know, she went and moved her own money. But she thought she was moving it to her son. Actually, she's moving it to a fraudster. So, you know, in addition to banks putting in place technology to try and prevent that from happening, one social question that comes up is in some places like in I think it's in the UK, they passed a law that basically said if a consumer, if an elderly lady is convinced to send money to what she thinks is her son and it's not, that's a failure of the bank's technology stack. And therefore, the bank has to reimburse her. And so there's been discussion in the U.S. Do we bring that to the U.S. or not, too? And so I do think as as we sort of evolve, we'll have questions like, you know, if if especially with AI, if if sort of the older generation ourselves are somehow taken advantage of, do we find a way to remedy the situation such as by asking the banks to pitch in or to cover this? Like, I think we'll probably wind up coming with some sort of social accommodation as well. Fair questions. Yeah. Don, thank you so much. Yeah, thank you. It's been awesome to be here. Yeah, it's been a great discussion. Yeah. Thank you so much. I really appreciate it. Thank you. Yeah. Thank you for watching another episode of Bay Area Innovators. If you enjoyed this episode and you know someone who should be on our show, please send us an email at sfbay at EpochTimes.com. We would love to hear from you. Thank you and see you next week.