EP 184: Ivy Nguyen | 10,000 AI Pitches and the One Test That Matters
59 minAudio
In 2017 another investor told Ivy Nguyen that artificial intelligence was too small a niche to dedicate a fund to. More than ten thousand AI pitches later, every pitch deck says AI. Single funding rounds have grown larger than entire venture funds, and from the outside the companies that will matter look exactly like the ones a fast follower overtakes within three years. Ivy Nguyen is a Partner at GFT Ventures in Palo Alto, the early-stage firm founded by former NVIDIA and Samsung executives to invest in AI and frontier technology. Before GFT she held investment roles at Zetta Venture Partners, Point72 Ventures and Autotech Ventures. She studied chemical engineering at Stanford and was named to Forbes 30 Under 30. In part one she explains why Silicon Valley money backs the vision before the plan, how a first company that ran out of cash led her into venture capital, why she passed on Coinbase twice, the test every AI company has to pass at her firm, and why the early AI businesses that passed it still ran on 30 to 40 percent margins. Part two continues in EP 185. WHAT YOU WILL LEARN Why Silicon Valley investors fund the size of the vision before the plan to reach it Why a strong business from another market can lose a Silicon Valley pitch in the first thirty minutes How passing on Coinbase twice led her to specialise in AI The billion-dollar test, and why proprietary data is the one advantage money cannot buy Why AI products that work in the lab lose accuracy and margin in the field BEST QUOTES (00:08:35) “The number one most common cause of death for every startup is cash flow.” (00:23:03) “In a two to three ton piece of metal on the highway, I don’t want it figured out along the way.” (00:33:56) “I actually passed on Coinbase two times, because I didn’t know what I was looking at.” (00:39:53) “Niche implies that it’s small. I’d like to use the word obscure.” (00:44:54) “Even $100 billion would not allow a copycat to brute force that level of performance.” TIMESTAMPS HOW AN AI INVESTOR IS MADE (00:00:00) Opening, the investor who called AI too small a niche (00:03:35) Chemical engineering, a first company and no plan B (00:08:35) The number one cause of death for every startup SILICON VALLEY AND EVERYWHERE ELSE (00:09:48) Why Stanford money backs the vision before the plan (00:15:54) Gold, defence research and the roots of Silicon Valley (00:22:35) Why a car at highway speed needs more proof than a grocery app (00:25:35) Seven years per car model, one weekend per software release (00:28:35) Growing up in Vietnam, where failure is shared BUILDING CONVICTION IN AI (00:33:41) Why Ivy Nguyen passed on Coinbase twice (00:35:43) A crypto fund or an AI fund, the 2017 coin flip (00:38:33) Ten thousand AI pitches and the empress of obscure THE INVESTMENT TEST (00:43:20) The billion dollar test and the proprietary data flywheel (00:46:32) Why ten years is the real deadline in venture capital (00:51:25) Why AI companies once ran on 30 percent margins Ivy Nguyen at GFT Ventures: https://www.gft.vc/ivy-nguyen Ivy Nguyen on LinkedIn: https://www.linkedin.com/in/ivyknguyen/ Newsletter: https://lsg2g.substack.com/ Follow Beginner’s Mind for long-form conversations on leadership, capital, technology, and the people building what comes next. Ranked in the global top 10% of all podcasts, named #1 Deep Tech Podcast in 2025, and #8 Deep Tech Podcast in 2026 Send us Fan Mail Support the show Join the Podcast Newsletter: Link Your Next Move: Link
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
Christian SoschnerIn 2017, on her first day at the fund built entirely around artificial intelligence, Evie Nguyen went to a networking event and introduced herself to another investor. His answer is the reason this conversation starts where it does.
Ivy NguyenOver eight years ago now, I went to a networking event and I spoke to another VC and introduced myself.
Christian SoschnerAnd he said, Artificial intelligence, isn't that and eight years later, the Uber drivers I meet talk about artificial intelligence, or as we say today, super intelligence. Every pitch tech in drug development, in med tech, and in logistics now says artificial intelligence, and I'm pretty sure soon superintelligence. Some single funding rounds are now larger than her entire fund. From the outside, the companies that will matter look exactly like the ones a competitor copies within three years.
Christian SoschnerHer test from telling them about comes down to one asset that money cannot buy.
Ivy NguyenNo matter how much money Softbank might be able to throw at you.
Christian SoschnerUm, like even 100 billion dollars will not allow you to the part I did not expect in the conversation was how hard the early years were. In her example, a model that's cost 99% in the lab drops to 60% on a factory floor, and the company price like software ends up on 30 to 40 percent margins. The cost is one piece of hardware. In the the factory, you have to work with the on-site camera, which is basically is a partner at GFT Ventures in Palo Alto, investing in artificial intelligence and frontier technology.
Christian SoschnerBefore that, she had investment roles at SETA Venture Partners, 0.72 Ventures and AutoTech Ventures. She has looked at more than 10,000 artificial intelligence pitches. And I ask her why Silicon Valley Money bags division before the plan. I ask what a company needs to return a venture fund inside 10 years. I ask why a model that works in the lab can still break the business that sells it. If you build fund or govern a company with artificial intelligence in its pitch, this conversation gives you the questions that Silicon Valley investor asks before she says yes.
Christian SoschnerI recorded this conversation as a live stream in October 2025, and this is the replay. I looked at the information that you sent me, and it's really impressive. You have been in artificial intelligence long before anybody talked about it.
Ivy NguyenYeah, Christian, thank you so much for being here. It's uh it's an honor and it's my pleasure to join you today. Um, yes, I it's been very interesting to see the artificial intelligence uh community go mainstream in the last couple of years because I distinctly recall that when on my very first day um at the venture capital firm that I joined that was entirely focused on artificial intelligence over almost eight years ago now, over eight years ago now, I went to a networking event and I spoke to uh another VC and introduced myself.
Ivy NguyenAnd he said, Artificial intelligence, isn't that isn't that too small of a niche to dedicate a fund to? And that stuck with me all this time. And I would love to be able to grab copy with this person if I can remember exactly who it was to tell if you remember to ask him if he remembered that interaction, because it's incredible to see how much things have changed in in the almost decade since.
Christian SoschnerI always wondered um you started investing in artificial intelligence before anybody talked about it. So basically, you were a contrarian investor and uh base uh placed a big bet in an area where many people back in the day said, um, well, if this ever works, I don't know. I don't know. It's stupid. I always wondered how does it feel? How does it make you feel as an investor when you see take the entire area off when you see that your pet works? How does that make you feel?
Ivy NguyenI I um well, one, I feel incredibly excited for for the entrepreneurs, excuse me, who are actually at the the ones who are actually at the front lines, actually building the companies, right? And putting much more on the line than than I am. I'm just a more of a cheerleader in the backseat along for the ride. But um, a lot of these teams that are now seeing incredible uh success, um, I think that while from the outside it may seem like it's an overnight success, they've been working at building these platforms and just building their expertise in artificial intelligence for so long.
Ivy NguyenSo I'm I'm just very, very excited for them to see uh to finally see that the technologies and the the product um philosophies and so on that they've been evangelizing for so long start to get greater acceptance um and start seeing greater pull from the market instead of um them having to constantly push a boulder uphill for so long.
Christian SoschnerYeah, that's understandable. It's uh 10 years, 15 years, and then suddenly it works. But you didn't start in in artificial intelligence, you had a career before that. Um you worked as an operator, so not directly into VCs. You worked at uh box, medallia, gusto. What experiences uh helped you the most as an investor in working as an operator?
Ivy NguyenSo I would say that I had a personal fairly limited experience as um as an operator. So I studied chemical engineering in undergrad. Um and although I went to Stanford, which today seems like it's crawling with uh wannabe or aspiring startup entrepreneurs and the VC is looking to find the next Mark Zuckerberg or the next Instagram founders. Um at the time when I was an undergrad, but it was still possible, although still not easy, to escape uh tech and venture capitalists.
Ivy NguyenUm and chemical engineering was one of those small areas where that was possible. Um, and the reason why I chose chemical engineering was because I came from a family of immigrants. Um, I was the very first in my family to go to college. And so I was looking for um a path of study that seemed to be like at lowest risk and fairly moderate reward. Um, you know, I wasn't a massive risk taker. Um, but one of the things that that um I was not very reflective on was I was not honest with myself at all about how much I hated chemical engineering.
Ivy NguyenRight. I I think um if now that I'm a bit older and a bit wiser, um upon reflection, I am somebody who likes to take big swings and who likes massive disruption. And if anything, chemical engineering is the exact orthogonal opposite of that, right? A chemical engineer's job is to make sure that um plants and production and so on can get to scale, uh, but very slowly, very deliberately, and in a way that doesn't blow anything up. Whereas, you know, I like the big booms. I like you know tearing things down and like and uh and building something up uh like a new in a completely different, perhaps hopefully better way.
Ivy NguyenUm but um to to um long story short was that um I I didn't admit to myself that I absolutely did not uh want to be a chemical engineer or anything even close to it at all until I was quite close to graduating. And um because I was at Stanford, I was afforded many opportunities to try to innovate my way out of the corner that I felt I felt like I painted myself into. And so I managed to raise a small amount of angel financing and grants to try to build my own company.
Ivy NguyenUm and uh it did not work very well. But I think one of the magical things about Stanford is that there are still people who would be willing to give you some capital and make a bet on you, even if it became quite apparent that I didn't know the first thing about managing cash flows uh and um and scaling up a business uh in a sequential way, um, and it's so that the business itself didn't blow up. And so I blew my business up and didn't have a great plan B for where to go or what to do.
Ivy NguyenThe the typical more um predictable tracked paths that undergrads might take out of Stanford, such as going into investment banking or management consulting, or perhaps becoming an associate product manager at Google. I never even spent any time like trying to do, thinking that this startup would be it, right? That I had no plan B. And so when that blew up, um, I took some time to apply to business school, but opportunistically took a job running a startup accelerator.
Ivy NguyenAnd in the meantime, um like uh finally started to understand a little bit more about the world of venture capital, beyond at the time what it was primarily focused on, which are uh which was mobile, social, local apps, right? Everyone was looking for the next Instagram, the next Snapchat, and the whole category of services that turned out to be, you know, the next Airbnb, Foursquare, uh Uber, Lyft, et cetera, Instacart. Um, and I thought at the time that that was all venture capitalists looked at.
Ivy NguyenUm, that uh the what in the quote unquote real technologies, right? Things that I thought would have a real substantive impact on the world, um robotics, medical devices, space launch companies, um, so I think at the time um self-reinforcing software, right? But I went out what people are now calling AI. Um, it didn't seem like anyone cared about that until I got very lucky and found uh a venture capital firm that cared about all of that. And so I joined a fund um that that is now uh called Fusion Fund at the time uh from basically day zero, when uh the the two partners that founded it were in the process of building out that fund.
Ivy NguyenThey were they were trying to raise capital from mainland China um and the rest of Asia. This was a very different time, right, than the geopolitical environment that we live in today. And the idea was to take that capital and that ability to scale manufacturing to um to work with US entrepreneurs and to be able to try to translate that technology across border. And um, that was where I got my first entry into venture capital um and where I learned that uh the number one most common cause of death for every every startup is cash flow.
Ivy NguyenUh they they run out, they die because they run out of money. It is as simple as that. Um, and and that's a lesson that uh I've I've brought um with me throughout my my very uh my very varied and uh very twisty career in um as a as an AI investor.
Christian SoschnerYou mentioned uh the Stanford culture, and so this is something that I find very interesting uh before we dive into other parts um that we tackle for sure in the conversation that's upcoming. Uh you mentioned that uh it's pretty normal in Stanford that uh angel investors invest 100,000 into a company in people who actually don't have a plan what they do. When you look at the European culture, it's pretty much the opposite. I mean, uh the first thing people get asked here is uh where is your business plan? And if it's not 200 pages and with a data room uh that's at least one terabyte, it's not good enough.
Christian SoschnerIs this really in Stanford uh such a unique culture?
Ivy NguyenWell, I would say it's not quite as blasé as I'm making it sound, right? But I'd say that um that that um that like um diametrically opposite culture that you observe is very true, uh Stanford and I'd say Silicon Valley Sand Hill Road as a whole has a very long-term big picture vision. And if the um if the big picture vision is there, if you can convince me that um no matter how we got there, and we'll we can figure it out, even if there's not a clear zero-to-one plan today, if the the long-term view is that we are going to tear down and completely remake a 10 billion, 50 billion, 100 billion dollar industry, or um that you know this this new product, this new widget can create demand uh for an industry that may not exist, but is so universal in its applicability that it could be a multi-billion dollar outcome, um, then I think that you can find very many people who would be willing to throw some support, to make a bet, to buy that lottery ticket, even if that plan from getting from zero to one is completely unknown.
Ivy NguyenUm whereas in other innovation markets, um, and um and I'd say that it's not just Europe that that is guilty of this. I'd say this is true even for many um many innovation started communities in the US outside of uh out of outside of core Silicon Valley. Um local investors tend to uh very reasonably, in my opinion, right, um ask for a lot of rigor in the zero to one.
Ivy NguyenWe want to know how you can get to cash flow break even. We want to um, you know, to to uh return to that point I said about well, what is the number one cause of death for for startups, right? It is running out of cash. How do you throw out as much free cash flow as possible as quickly as possible so that you can maintain um that so that you can you can gain control of the business, you can gain control of the fate. That like that is um what gets uh I think very reasonable first and like primary importance, but that also comes at the expense of the big picture vision.
Ivy NguyenI found that many of uh investors in other markets would tend to, um, because of their different risk appetite, uh, invest in fundamentally robust businesses in the near term and trade off on the scale of a vision. And so I see founders uh who have been quite successful launching an MVP. Um, and they probably are actually tackling problems that are multi-billion dollar problems, if not hundred billion dollar um potential outcomes.
Ivy NguyenBut because they are so used to uh having to overly prepare, like to uh and anticipate all these questions to an incredible level of robustness for their local investor audience who gave them that initial round of financing. When they go to Silicon Valley, all their answers are focused on the near term. This is what I've done so far. This is how I'm gonna de-risk things in the near term, this is how I'm going to increase margins. And that um those answers take up the vast majority of, let's say, that initial 30-minute slot um that you have with a Silicon Valley investor.
Ivy NguyenAnd it comes at the expense of really telling the full hundred billion dollar massive opportunity story. Um and so that investor uh leaves that conversation thinking, I don't think this entrepreneur is thinking that big. I don't think that he or she is taking that big of a swing, even if that may not actually be true. Um, and and so unfortunately, you know, in that in that interaction, and I've seen it so many times, that founder gets unfairly dismissed for not pursuing a sufficiently large opportunity to be able to deliver to that angel investor or that venture capitalist that hundred, uh, you know, that 100x return on their investment.
Ivy NguyenI hope that makes sense. But it's uh it's it's something that it's a theme that I see recurrent for whenever an entrepreneur who's done very well for himself or herself in a different market show up in Silicon Valley and they start pitching. And they they come back to me with a shocked face look. I don't understand. I don't like I, you know, I'm like I'm doing everything right, and they are, right? Um, in terms of running the business, why am I not getting enough credit for that?
Christian SoschnerNo, it totally makes sense. And I think um you meant you said it's unfair uh to the founder, but I don't think it's unfair. I think it's the founder's job to also research the culture in an area when he or she uh wants to pitch in Silicon Valley. I mean, they should also read some books and uh just find out how is the culture there, what RVCs looking for, and what you described is pretty clear. In in Silicon Valley, the culture is big swings. You want cat you want people that define new categories that have ideas to transform entire economies at the end of the day, like Nvidia did um in the 90s, or like Apple did in the 70s, or Microsoft did in the 70s.
Christian SoschnerBut you don't want this incremental progress, something make a little bit better, doing well, quick cash flows. You want the big swings, and if don't work, it doesn't matter. Uh go go go go find your next idea and uh come back then again. Yes. Um my question to you is um I find it really interesting because this is a unique culture, it's I think globally unique. Um it doesn't look like much, it looks like a subtle shift. But when you look at the result today, uh how Europe developed since the 90s and how Silicon Valley developed, I think almost all trillion dollar companies uh have their roots in Silicon Valley or at one point got help from Silicon Valley uh VCs.
Christian SoschnerDid you ever spend some time thinking about how this culture evolved over time? Uh, did do you have uh an interpretation? It's just interpretation and just uh uh uh hearing your opinion on that topic.
Ivy NguyenWell, um I I am a big uh Silicon Valley history nerd, right? And I um I I'm not sure if um many of the folks who are joining us in the audience today who might be listening to this episode later are aware, but the roots of Silicon Valley actually came from um from US Department of Defense research out of Stanford uh and a lot of the local uh university labs um that initially got commercialized um and um uh combined with the unstoppable kind of uh force of of California hippie culture, right?
Ivy NguyenCounterculture. Um it's an odd mix, but I think that um it um like the it tended to bring in people who are incredibly smart from all sorts of disciplines, not just the engineers, the physicists, the scientists, but also I think people who think about um massive like change from a societal and cultural level. Um and you know that that um that crash collision, right?
Ivy NguyenThat like massive collision from um both sides uh in the 60s and 70s really gave birth to the the internet culture, I think the um the venture capitalist culture kind of that that exists at its fringes um that we've seen evolve into what we have today. Um it's um and then I I would also uh combine that with the fact that because California is one of the uh newer states in the United States and is on the Western frontier of the US, um there was always even if you go back to the founding of California, um the uh the gold rush, right?
Ivy NguyenThe discovery of gold in the um the late uh 1840s uh was really what kicked off um a massive dash and a massive population growth from across the world, not just the US, to California. And so I think that that something about the waters, I think in in the Bay Area has always attracted people who might um be genetically predisposed to uh you know throwing throwing everything into um into trying to make a you know big life-changing bet or the swinging big.
Ivy NguyenAnd I I I wonder if over time that that mentality has just continued to persist and build on itself over the last you know almost 200 years um of California history.
Christian SoschnerYeah, it makes total sense. And I think uh people or entrepreneurs, especially who want to pitch their RVCs, who want to syndicate with investors, should just uh read a little bit about this culture and understand that you need a big vision, you need to swing big if you want their money. And uh if you don't have it, um maybe allocate your time otherwise.
Ivy NguyenWell, well, Christian, I I wanted to pick on something that you said a couple of times, right, which is reading about the culture. And I'd say that, you know, I uh for anyone who's interested in learning more about Silicon Valley culture, it's it's so hard to capture what what this actually means in practice and how you know how it even feels and how it can even change your mindset, how you communicate, um, and and how you just engage by by trying to read about this and learn about this uh remotely via a book or via webinars or so on.
Ivy NguyenYou you really have to be immersed here in San Francisco, here in Silicon Valley, to just really. appreciate how how uh different of a culture it is right um even you know like the i i started uh us down this thread by talking about some of the communication norms right um of a uh a founder whose company was born in Silicon Valley um and was funded from the cradle by Silicon Valley investors versus investors in other markets um being implicitly trained to focus on the big picture versus um versus instinctively or being explicitly being asked to focus on the near term fundamentals um at the at the expense of the big picture story.
Ivy NguyenBut um like that's uh that's something that that that that's that that kind of behavior I think is very much um something that you learn via osmosis right now I like I I work with a number of um folks in innovation and strategy um in um in the auto sector um like prior to coming to GFT and many of them were from Detroit many of them from are from from Stuttgart from a lot of the the the German auto manufacturing hubs and the number of friends that I've had who have uh had temporary postings in Silicon Valley where you know they're mandated to spend a quarter six months a year here just the first couple of weeks right um of of having led it in Silicon Valley the number of serendipitous conversations they might be sitting at a Starbucks or a blue bile coffee um and uh taking a meeting only to opportunistically look up and overhear in the table next to them right um somebody pitching like exactly the the kind of solution that they've been looking for you know for whatever whatever mandate it is or getting invited to a hackathon or getting invited to um a uh humanoid robotics fight club right and uh like just the the the kinds of things that you actually encounter uh having boots on the ground here and actually uh getting to to really immerse yourself into um into this yeah you know uh culture and and just the the heartbeat of Silicon Valley uh cannot be replicated in my opinion uh by by trying to capture it in a book or or in a podcast or so on.
Christian SoschnerYeah no absolutely so the best way to connect with the community is having an office there and allocate people there and work work with them and not only uh try to fly there for one pitch event and then go back to your home.
Ivy NguyenYou mentioned immersion yeah that's right yeah absolutely you mentioned um working with German car makers and you're based in Silicon Valley how did you experience the cultural difference between Germany and Silicon Valley um well I would say that uh I I there there is a lot of hand waving in Silicon Valley right uh as I mentioned earlier we we do focus we focus on the big picture how we get there I don't know like we'll figure it out and and um I think that's fine um when you're working with a a consumer game like far or you know can or a um grocer delivery app if it fails if there's a massive AWS outage like no like we'll survive right um but when I am in a kind of two to three ton piece of metal uh going down the highway at uh like 250 kilometers an hour um it I don't want it to be figured out along the way right I want to make sure that it works 99.99% of the time um and that uh when it fails I have all the different redundancies to make sure that uh it it fails in the safest way um and I think that that that culture um like the the culture required to produce to produce that kind of outcome thank god right is is very very careful very deliberate um and requires uh a very high bar from the the startup founders that wish to engage with them or the VCs that wish to engage with them before they lean in right you you have to go in and really prove efficacy and credibility in order to to be like even let in right and to get more information in order to start collaborating and to start going through that iterative loop iterative loop together.
Ivy NguyenWhereas in Silicon Valley, you know like um because some of the uh earlier successes of Silicon Valley are lower stakes right the the willingness to try um is is just that much higher because the ability to fail and the safety of failing um is is so strong right that uh you can like um I I I've seen business opportunities right or collaboration opportunities form impulsively at the drop of a hat you know truly because the person A met person B at a coffee shop and thought hey what the heck let's just make give this a go whereas that's just not possible like in in in Detroit in in Germany right in in Asia where a high level of functional safety um like like needs to be met um both uh for reasons for regulatory requirements um for uh you know uh the reason like for the threat of commercial of um commercial fines and civil lawsuits and just for a matter of ethics right um like do you would you put your own family in a death trap or in something that um that you quickly cobbled together over the weekend and decided to push out to see how it might behave?
Ivy NguyenAbsolutely not right and and so but that just makes the ingestion of innovation that much harder. Maybe what one last um one last stop that I could I could leave you with on this point um before I drag us all further into this tangent because I could talk about it all day is just think about product with lease cycles, right? And the difference between the two. In in um the traditional automotive sector, it takes seven years to design a car from scratch to get that car launched on the market.
Ivy NguyenAnd maybe like a vehicle model your update it takes two or three years and that's kind of like that's going in lightning speed right to push a new technology into an existing vehicle model and to get that into cars on the road. Best case. In Silicon Valley like a a new product innovation um can happen in a matter of a weekend right you can release you can completely rebuild a new version of Uber which is now a 15 year old company and launch that um and uh I'm sure there will be problems right but uh like the um with that level of speed but uh you can get that product out to market just that much faster.
Ivy NguyenUm and like we're talking like many, like that's effectively like a like many, many generation like dozens of generations of difference right in um in product evolution between one weekend versus seven years. And um you know I I I think that there are a lot of people prescribing things that the auto sector the traditional auto sector should do or shouldn't do in order to be able to innovate at the speed of Silicon Valley.
Ivy NguyenI'm not that smart. And so I I don't presume I won't presume to to give any advice myself but something has to change right I'd say about about the traditional seven year loop um because like it's um it's it's very very different than the speed at which Silicon Valley um is is ingesting new technology and engaging with the state of the artly agree I think the difference I mean especially for Europe it's crucial to change that and I think for me the difference is big corporation thinking and um small company thinking just puts the innovation in a small company lets them try and if it works um move forward with it and if it doesn't work just throw it away and try the next one which is not so easy in big corporation politics.
Ivy NguyenI find it's really interesting these cultural differences I mean at the end of the day it's just people working together then the people form a unique culture and then one or two decades in the future you can observe how different the outcomes really are talk talking about culture and you mentioned in your material that um you grew up between Vietnam and Silicon Valley how did these cultural differences influence your decision making I don't think anyone's ever asked me that question before um I'd say that Vietnamese culture uh at least the the uh the southern part of Vietnam where I grew up um is very more similar to how how you and I have been characterizing the European startup culture right which is it is um it is fairly insular it's very community driven everybody knows everybody and so there's a bit of you know anti-tall poppy syndrome you don't want to stand up you don't want to stand out too much and um because uh everybody is in everyone's business you you also don't um like you are discouraged from taking too much risk and rocking the boats uh because when you fail it's not just your failure um in isolation you know your failure has ripple effects onto your family your uh immediate community and so on and so that sense of shared responsibility um and and that sense of shared risk um is is kind of I think a different orientation than in Silicon Valley whereas in Silicon Valley it's you know it's more a matter of um uh finding time I think just that much more precious right so the number of board meetings that I've been in where and the the the consensus in a um product discussion discussion and your time is so precious I would much rather you bet the company right and on um this move and uh and like uh take like put all the chips on the table and run 100% at a brick wall and see if you can bust through or if you fail then then then that's fine because your time is precious.
Ivy NguyenIf you fail take a break clean up pick up the pieces and start again right as if there like there isn't like there aren't real consequences to that right as if these people may not have like don't have mouths to feed or as if the employees that you're taking along this you know speed run into a brick wall don't have you know don't have bills to pay don't have don't also have mouths to feed and so on. It's um it can be very uncomfortable right just hearing this and thinking well yes on the one hand from an investor's perspective where my job is to make the big bets and to cut our losses on the ones that aren't working like this is the right um the right advice to give as a board member and as a friend to the entrepreneur.
Ivy NguyenBut on the other hand there could be like this this might be a team of 20, 30, 300 people, right? Whose whose jobs uh you know like are are on the line here. It is very uncomfortable.
Christian SoschnerLike it is a very like that I think that cultural difference um makes uh the uh I think the the impact of the kinds of decisions that uh I am involved in supporting and in in driving um just that more acute yeah failure culture is uh is pretty different outside Silicon Valley also here in Europe yeah when we look when we look at your career um ai is the guiding star in your career and let's uh look at this space now with the next couple of questions um I'm curious to hear was there as a moment in your life um where you decided that ai is it this is the the the one thing that you want to pursue was there a single um decision making moment um you know I I don't know that there there actually was uh I think it was something that um that I built very strong conviction on incrementally but I'll tell you how I got my start in AI so I I'd mentioned that my first job in in true vc not just with the startup accelerator was at a firm called fusion fund and at fusion fund um if you if you have a chance to visit their their portfolio page fusion fund looks at everything in deep tech literally nuclear fusion but also um you know medical device robotics ai uh rocket launch companies um brain implants you name it and when I was starting at Adventure I was also simultaneously learning how to uh how to practice as a venture capitalist like what is a term sheet what does a cap table mean what is pay to play etc right not just the technical terms but how to apply that in in different deals um and I was simultaneously trying to learn about effectively the entire universe of companies right and of technologies um I didn't know up from down and I couldn't figure out well why why was I like why didn't it feel like I was making any incremental progress?
Ivy NguyenI didn't know how to tell a good company from a bad company beyond how well a founder was able to articulate his or her story. And I I have um I um I I was only like finally brave enough to admit this to um to my my first boss adventure just uh probably three months ago um when I had lunch with him but I actually passed on Coinbase two times really um because I because I didn't know what uh what I was looking at and I remember having uh conversations first with an one of the earliest investors in Coinbase and they kept throwing out all these terms that didn't make sense.
Ivy NguyenI'm not I'm still not convinced that blockchain as technology has any near term business ROI yet right but I was just trying to get to well but why like why would you want to have a like a crypto wallet like what what is this crypto thing? And so I declined that first meeting and then I I had I was at um some um you know like a young person in tech like get together I think it was a part like one like a party one evening um at someone's someone's home and I was sitting next to Brian Armstrong the the founder and CEO of Coinbase I just thought I don't think this guy like this guy must be on something because like I only understand maybe three words of what he's saying.
Ivy NguyenI have no idea what this is I I'm not gonna pursue it. I think he's crazy. And you know frankly Crystal I was right for for probably a good three or four years right because not not many people um I really invested in Coinbase. The ones that did early on they're probably very happy with that decision right but um I to I to to bring us back on this tangent uh I I bring this up just to illustrate how much I didn't know what I was doing and what I was looking at and frankly how impossible of a task that was right to be a generalist deep tech investor um who um didn't have much functional experience in any one sector and who also came into venture after crashing our own company into into that proverbial brick wall.
Ivy NguyenAnd so I got I got advice from a couple of trusted mentors and friends that I should focus I should more narrowly focus and I should join a fun that is entirely focused on a technology theme or a sector, whether whether it was video games or blockchain or AI. And because I I didn't really know anybody in Silicon Valley having tried to um avoid as much as possible the whole startup ecosystem while I was a tech startup ecosystem while I was at Stanford I I just took whatever conversations I could get and at um at the end of the day I had two opportunities.
Ivy NguyenOne was to join a um a crypto fund uh and this was in 2017 and the other was to join one of two um at the time like uh 100% AI focused funds um and I didn't um I made the decision on just on the basis of of like I think what people are calling vibes now right I I like the people at the AI fund more.
Ivy NguyenI uh I thought I could learn from them more um and um there were more words that I could understand frankly about AI and the the business value um while it was still very much prospective just seemed a lot more tangible. So it it it felt like a corn flip at the end of the day but that was how I got into AI in the first place um and uh started learning a bit more about both the tech the differences in technology as well as um kind of product architecture and so on compared to enterprise SaaS um and and over time just kept following my instincts and building on that in a confounding way.
Ivy NguyenAnd so you know eight years later here I am but I I um there there frankly wasn't one clear aha moment that I thought um yeah I I should pursue this 100% of the time this is real. Like it was very lucky right at the end of the day.
Christian SoschnerYeah so there was no uh mountain that you climbed and then you had your revelation and enlightenment I like the stories that you tell about uh Coinbase it's very interesting it's I think very I'm horrified thinking back on it especially given how the company yeah but but when you look at Bitcoin it's even worse I mean uh just imagine investing in Bitcoin 2010 I think the device was 1.0 one one point uh 0.1 or something dollars so a big opportunity and in 2017 when you decided for AI uh there was this crypto bubble so everybody was talking I mean here in Europe everybody was talking about crypto but you went for the contrarian bed you looked uh at artificial intelligence and when I think back to the material that you sent me did did I read the number right that you looked at 10 000 AI startup pitches yes I think the number might actually be higher because I so I I try to meet with at least three entrepreneurs a day um but uh sometimes that that are some days that that might be as many as excuse me uh 2025 right um especially if if it happens to be a demo day and so if I just think back and and I include weekends on this and so if I if I think back on the last eight uh to 10 years in in venture um the number is at least uh 10 000 yeah really 10 000 in eight years that's an impressive number this means um you look at pitches every single day material it is there's not a single day that you can't uh work on pitches oh i it's there's not a single day that i i um i want to do it right it's like it it almost feels like an addiction because just it's it it is incredible the the number of uh massive like I I hate the word niche right um like my um like a a a a former colleague calls me the the Empress of niche because niche implies that it's small right I think it I'd like to use the word obscure right because small is a bad word in Silicon Valley right we we like to think big we like to go for the billion dollar swings and just the number of sectors that I had never thought about I had never even considered considered um to to exist um that are driving multi-billions of dollars in spend that exists out there right it it is just so incredibly fun to uncover that um and and so just so fun to learn about the different workflows learn about the different value chains just learn about the just meet the people right who um who are trying to bring to fruition uh innovation and and just transformation in all corners of of the world and all sectors of our economy I am I think about uh I was just telling a mentee of mine that you know as as much as uh there might be things to be frustrated by in in the sector uh and there's a lot um at the end of the day I I um I realized that um every every conversation I have um is with the most like an optimistic person right in in that person's orbit like whether it's a startup entrepreneur or whether it's someone like like uh driving innovation or strategy at a large corporate this is like among like the smartest like people and the most optimistic people just trying to to like um bring something new to life and I think that that energy is just so special um and and just like so so different than probably any almost any other job in the world.
Ivy NguyenRight. And so to me like um see going through a dozen pitch sex a day um every day right the first time I wake up is is not like at all work. It's like it's fun. Right. It's like it's the Addicting part of my uh my work.
Christian SoschnerYeah, that's good. That's good too. Yeah, I think uh they say you need 10,000 hours to become an expert in any area. And in the startup world, it's 10,000 pitches at the end of the day. So it's a visit. I guess so, yeah. Yeah.
Ivy NguyenI I wouldn't call myself an expert yet by any means, right? But uh, but it's um like I'm I'm I'm certainly trying to get my to put my 10,000 hours in.
Christian SoschnerYeah, yeah, absolutely. Um what I'm curious to learn from you. I mean, 10,000 pitches, you have a lot of expertise in that area when you look at them and talk with the people and process it and uh can compare between them. Um when we have a bubble time, like we have now in artificial intelligence. I mean, everybody's talking about it. And when I uh take a cab here, um taxi driver is talking about artificial intelligence. When I go to ABUB, they talk about artificial intelligence, so it's really mainstream now. And the odd thing with mainstream is uh since everybody talks about it, it's in every pitch.
Christian SoschnerI mean, every company in drug development now has artificial intelligence in their pitch tech, every company in MadTech artificial intelligence, every company in any area, even the hairdresser announces artificial intelligence. Did you did you did you develop a method uh to separate the weed from the chaff to get a pile of really true innovation in that area and can weed out fast and with precision those who just copycats and not real innovators and uh are likely to not work out?
Ivy NguyenUh so I I would say I I wanted to nitpick on something you said, Christian, which is uh you know sorting the weed from the chaff. So I'd say that um there are many very like valuable um and just or just fun applications of artificial intelligence um that may not yield great venture investments. I'm looking for a very specific thing, which is um I'm looking for the formula that I'm looking for is can this company get to uh a billion dollar exit value at minimum uh in minimum or maximum 10 years, but ideally five years with the minimum number of capital, minimum inputs, maximum output.
Ivy NguyenUm and um my hypothesis, my my firm's hypothesis for how he would do that is to um for to find companies that have access to proprietary data and can generate um and use that data to generate a reinforcement loop or reinforcement layer that leads to or is highly correlated with um a improved, like a compounding perform product performance improvement over time. So as you accumulate more customers, you accumulate more data, you use that data to further and further fine-tune your models that um gets you credibility, that gets you um access to be able to deploy even more models to set off more of those flywheels.
Ivy NguyenAnd over time, your out-of-the-box performance just gets that much better over time compared to um some like a follower, right, who's behind you or a copycat. And um that because that data is only generated with continued engagement with the models, uh, there's no way that no matter how much money SoftBank, right, might be able to throw at you, um, like even a hundred billion dollars would not allow you to a copycat to your brute force and replicate that level of performance, no matter how many, you know, uh $10 million signing bonuses you might give to an AI, the best AI researcher, um, et cetera, et cetera.
Ivy NguyenUm, and and so not not every application of artificial intelligence can build that. And and that's okay. Like there, there, I think there are other um like there are other investors that uh are um that still would be interested in in backing those companies and like you could still get um like very attractive um investment return, probably right, with um with that um with those types of companies. But uh at GFT we believe that you know the companies that are like be able to return the 10x, the 100x, right, the 1000x, which is what um our brand of uh venture capital is all about, really, really benefits from that data flywheel um and that that reinforcement loop that I mentioned.
Christian SoschnerSo to understand you right, when you look at the pitch that it qualifies for your fund and you say, I take my time for a second and third meeting and look deeper into that and talk with my partners, you want to see a big swing and a clear pathway to a 1 billion valuation. So this is uh one of your hard criteria.
Ivy NguyenAnd and and then importantly, in under 10 years, right? Because in in venture capital, the the fund lifecycle is designed to return capital to its investors uh in 10 years. Like I'm an investor, but I also have investors to answer to, right? And I only can can can keep having a job and keep having capital to invest if I make money for those investors, but they don't have all the patience in the world, right? Like um, time is precious, you know, like uh like you know, um net present value of your dollars is is like is important too.
Ivy NguyenAnd so I can't say, hey, give me 50 years, um, and I promise you I'll I'll make you 100x on your investment, right? Like IRR matters too. Uh and so generally, um like the the implicit contract and sometimes explicit agreement we have with our our customers, our investors, is um we will return, uh we will get we will do our best to get you that 10x return in 10 years' time. And so I'm like I I because of that have to filter for companies that have the ability to grow to that level of return in that in that time horizon.
Christian SoschnerThis is still not mainstream know-how outside Silicon Valley. So in Europe, I'm always startled that many founders don't understand that a VC has to return capital to their investors and has to beat. So it's an important information uh to put out there. And thank you for for uh touching on that point. I mean, you have to beat basically the S ⁇ P 500 or Berkshire halfway at the end of the day, because alternatively it's uh not very healthy for an LP to put money in a in a venture fund, lock up the capital for 10 years if you then get nothing back at the end of the day.
Ivy NguyenYes, that's right. That's right. And I think the the um it's it's not a it's not a very well-kept secret, right? But if you actually look at the average venture fund, they lose money, right? And so venture is very much a high risk, um, high reward, um, but I think on average, like a fairly volatile asset class or fair, unreliable asset class. Um, but uh to your your earlier point, I I'd say that what no matter where you are in the innovation ecosystem, whether you're operating as a VC or a startup entrepreneur or someone in corporate strategy and innovation, right?
Ivy NguyenThe common through line here is uh I think that you and I are arriving on is it's so important to know your audience, right? Um and so if you're a startup founder, your your audience is twofold. It's the VCs that back you and you know it's the it's the end customers, right? And and what they actually need. And sometimes, um sometimes you know what I might be asking for as a VC could actually feel like uh very much intention, right, with um with what a corporate customer might be asking for. Or um, or even as you pointed out earlier, um, as a Silicon Valley VC, uh, I might be asking for something very different than um what you know your the um initial group of angel investors and local investors that gave you, you know, the entrepreneur your start, um, might be asking for.
Ivy NguyenSo it's a matter of knowing, like it's a matter of understanding incentives. It's a matter of knowing your audience, right? And figuring out um if if you can solve it with concept switching, with code switching, communicating to them and highlighting to them the things that matter, uh, or um if if um if the incentives are truly very misaligned, right? Perhaps trying to find the the right stakeholders to to um bring around the table to you.
Christian SoschnerYeah, I totally agree. And it's good to hear it from you and as a reminder for the founders, um, it's a hard criteria if a founder cannot promise it. I mean, there can always something happen, but if they can't promise a 10 to 100x return, I don't know if it's a wise idea to pitch to BCs at the end of the day, then.
Ivy NguyenThat's right. And and if anything, uh I I also warn entrepreneurs against taking venture capital too early, right? Because it's not just a matter of, okay, well, I'll I'll give you a check and then you you disappear for nine years, and then you pop up um in year 10 and you tell me, here you go, Ivy, here's the here's the 100x uh return, right? Um I I've increased the value. I I did nothing, or at least I had nothing to in the near term um on the value of the company. And then um year 10, I somehow found an investor or a buyer, right?
Ivy NguyenI will value the company at 10x of what you initially paid for it. In like I have near-term deliverables, right, that I have to I have to uh to bring to my investors too. And so um what um like like you know, and and this was a this was a massive uh tension right for a lot of for old-time um AI entrepreneurs is that uh AI companies took a lot longer, especially when I got my start in AI investing, they took a lot longer to build.
Ivy NguyenLike AI products took a lot longer to build, they took a lot longer to deploy, and then in many cases, they made it take a lot longer and a lot more um human capital and services in order to drive business value for the end customer. And so for the longest time, um AI entrepreneurs and AI companies, uh, if you if you compare them head to head against more traditional enterprise SaaS companies without a massive AI model uh underlying their product, like AI companies look pretty bad, right?
Ivy NguyenUm compared to like a classical SaaS business that's operating between, let's say, 75 to 90 percent margins, your only cost really is your cost of cloud, right? Um, things that can be deployed in a couple clicks or like in a self-serve way. So you push that work onto your customer, et cetera. By comparison, AI companies might take six to nine to 15 months to deploy from an initial contract. Oh my God, right? Especially if the contract is only one year long.
Ivy NguyenUm, you like they um, and then you you you deploy only to find that, oh my God, you know, this model that worked great um in in prototype stage, or at least in in um in the lab, so to speak, at 99 efficacy, for whatever reason is only operating at 60% accuracy or confidence levels in the factory floor, possibly because in in the lab you train on state-of-the-art cameras and in um you know in the factory you have to work with the on-site camera, which is basically uh you know potato quality, uh, you know, 320 pixel uh resolution.
Ivy NguyenUm, and to make up for that difference so that you can still provide some kind of business value for your customer in order to retain them at renewal. You have to throw uh four deployed engineers, right, uh, like or professional services in to make up that difference. That wizard of oz, you know, man behind a curtain approach um to making that model work, right? Um, and filling in that 40% of the time when the model fails. All of a sudden, that took what should have been a 80 to 75 to 90 percent margin business.
Ivy NguyenAnd oh my God, now you're operating at 30 to 40 percent margins, right? Your investors are screaming at you, your customers are mad at you, everyone's angry. What do you do? Right, and and how do you how do you keep going? And I I it was it was not easy being an AI entrepreneur um for the longest time until until like the rest of of um the mainstream uh market caught up.
Christian SoschnerYeah, and part one of my conversation with Evie. This first hour went from a chemical engineering degree and the first company that ran out of money to 10,000 AI pitches. And to the test, her firm applies to every one of them. A path to a billion dollar exit inside 10 years built on data a competitor cannot buy. It ended on what the early years cost. AI companies that worked in the lab ran on 30 to 40 percent margins in the field, with investors and customers angry at the same time.
Christian SoschnerMy next question was what an investor does in that moment, whether she writes the company off or stays at the table. Her answer starts with a captain and a backseat driver. Part two also brings the third quotient. She uses to charge a management team, one I had never heard of before, and her answer to the biggest fear in the AI debate whether artificial intelligence replaces all roles. Part two goes live next week as episode 185.
Christian SoschnerFollow the show and it reaches you the day it is published.
Invest in the wrong startup and you lose one times your money. Miss the next Google and you lose ten thousand times. Most boards are built to avoid the first mistake. Venture capitalists play by different rules. They...
The trait a Silicon Valley AI investor rewards most in a management team has a name almost nobody uses: XQ. IQ brings the idea and EQ reads the people across the table. The execution quotient shows in something as small...
A seed investor at Novo Holdings would rather read your rough idea than your finished pitch deck. Most founders spend a year building the deck, forming the team and waiting until the story holds together. By the time it...
What does it mean when the most respected adviser in Silicon Valley tells you that you are the best chief executive he works with, while your company is heading straight into a wall? Bill Campbell said exactly that to...
What do you do on the morning when there are no good moves left? In March 2001, Ben Horowitz took Loudcloud public in the middle of the dot-com collapse. The bankers cut the price from ten dollars a share to six in the...