#315 Venture Capital in the Age of AI: Expert Perspectives from Manish Patel

[00:00:00] 

Mehmet: Hello and welcome back to a new episode of the CTO show with Mehmet today I'm, very pleased joining me from the heart of silicon valley. I would say silicon valley Uh veteran Manish Patel Manish Thank you very much for being with me on the show [00:01:00] today The way as I was explaining to you before the way I love to do it I keep it to my guests to introduce themselves.

Mehmet: I have a theory. No one can introduce someone else better than themselves. So The floor is yours. Tell us a little bit about you what you do You And, you know, whatever you want to share with us at an introduction. 

Manish: Sounds great. Well, thank you again for having me on the show. It's an honor. I'm a big fan. Uh, so Manish Patel, I'm a venture capitalist here in Silicon Valley.

Manish: Um, uh, I'm lucky enough to have co founded a firm with one of my best buddies, uh, Freddie Martinetti. Our firm is called Nava Ventures. We do stage focused Thematic investing. So our stage is series a, and we love to have strong points of view. Uh, when we invest, um, we've been around for about three years, uh, and been fortunate to invest with some of the best firms in the world, like Sequoia, Excel, NEA, among a number of up and coming firms as well, mostly based here in San Francisco.

Manish: And we also have a teammate out in New York. We're focused on us, uh, investing in North America. [00:02:00] My background is before starting Nava. I've been in venture for about 13 years now. Uh, I was a general partner at a firm called Highland Capital, which was a firm that started in 1980s. And really, that's where I got my training and venture capital.

Manish: I was out on the West Coast and helped that firm build a bigger presence out here. Their headquarters was out in Boston, and before that, I was fortunate enough to spend most of my twenties at Google. I was an early employee there through the IPO work with a number of amazingly creative, smart people on was part of the journeys that company grew to this kind of global presence.

Manish: Uh, as I always like to say, I was Not a very good engineer, but I was an okay product manager and worked on the early ad system at Google AdWords, uh, the Google televisions team, uh, Google maps. Uh, so a number of really great products, which touch many, many different aspects of people around the world. I also worked on some pretty bad ideas too, because why not?

Manish: It's always good to experiment. Uh, I was fortunate to work with one of the founders, Larry Page, for A number of [00:03:00] years as well, running our corporate strategy and OKR processes, which is a process many startups use today in terms of goal setting and being able to have conflict resolution as they think about what is the right direction for their company.

Manish: So that's where I found my 20s. And then I came out to the Valley 20 plus years ago. Uh, I went to Stanford, uh, I was lucky enough to study engineering there. Uh, and then I also teach at Stanford and that's why I have this little banner up here. Uh, so for when my students ask you me, where, where am I on campus?

Manish: It feels a little bit more like homes when I do my office hours. So that's a little bit about me and I love technology and I've been fortunate to be part of the Silicon Valley ecosystem for so many years. 

Mehmet: Again, thank you very much, Maneesh, for being with me today on the show. We have pretty much a lot to cover.

Mehmet: Let's try to see how much we can do. Um, yeah, great. Uh, by the way, I'm a little bit biased. You mentioned you like you, you were like In product management, I'm a little biased. I work in [00:04:00] tech companies, but of course on the field. So I was the sales consultant and you know, like I work in sales as well, but One of the favorite teams that I used to talk to are the product managers.

Mehmet: So I'm a little bit I'm a little bit biased here now You know, from, let's start, you know, from technology perspective and being in the VC space, give you maybe some more visibility on what's happening. And you would agree with me, Manisha, of course, that AI is the big thing. Now, you know, I've been hearing about it.

Mehmet: I've been hearing about it and everyone is talking about it, but of course, you know, there's a lot of talks about, uh, Um, okay. So, so we are doing this investments in AI and you know, people are saying, you know, 2022 at the end of 2022, we started to see few things, 2023 we went full fledge and now here we go.

Mehmet: We are almost in the end of the third [00:05:00] quarter. first quarter of 2024. So given your extensive background in tech innovation, um, how do you evaluate the readiness of AI technologies in 2024 to deliver significant ROI for businesses? 

Manish: Yeah, yeah, absolutely. So, uh, I think you outlined it well, where, you know, 2023 in my mind in AI was the year of awareness.

Manish: Uh, everybody started talking about it, you know, chat GPT launched and really the product took off last year, 2022, the product took off in 2023. And that's done an incredible job of bringing awareness to the power of these large language models, LLMs. I mean, every day, my mom sends me a poem. She writes on chat GPT.

Manish: T, uh, on WhatsApp, you know, that that's pretty incredible, right? And when I think about your question of like ROI, I do think that 2024 is going to start to become the year of ROI. As we think about AI, um, I think [00:06:00] about it from a few perspectives. There's the consumer lens, there's the foundational model and the company building that some of these big models, there's the infrastructure lens, the companies that are sometimes powering those models, the underlying piece, and then the enterprise lens, and I'll walk through each of those and give you my perspective.

Manish: Consumer lens is easy. What's the ROI on that? Our end users are being delighted. There's value created. It's fun. It's interesting. It's different. It's somewhat magical. So that's an easy one. I, so I do think there's going to be continued to be more of that. And I think consumers that generally will benefit, which I think is great for the ecosystem.

Manish: I think with these large foundational model companies, uh, whether it's the opening eyes of the world and tropics of the world, Pretty amazing companies that I think have big ambitions. Uh, the question I think about for those on ROI is that, you know, is the business case going to be valid from a direct to consumer experience, you know, as an early employee at Google, and obviously I'm biased because of that, Google is highly efficient at answering these queries, these [00:07:00] search and answer queries, right?

Manish: It's very fast. It's very cheap. They have fantastic infrastructure that's been refined over many years. If you look out there, and I don't have the specific data, but you can see on X or Twitter, as you call it, Sam Altman and others talking about the cost to serve these queries, and they say, you know, each query or each chat is, you know, Single digit sense.

Manish: That's that's pretty expensive. You know, if I think about my own querying behavior with many of these systems, um, I'm sure I'm costing them more than I'm spending on my monthly premium subscription of 20 or so, right? And so, and the interaction is very different opposed to Google, where it's very quick interaction.

Manish: I get my response. There's this back and forth happening, which is also just more expensive. So I think the big idea there, this company's trying to build a habit is awesome, but I think they're going to have to. Also think about new and different ways to monetize beyond consumer subscription. Um, because Google and others, uh, mainly Google though, it has this amazing advertising machine that's built out over many years, which is [00:08:00] incredibly effective and incredibly efficient.

Manish: So I think that's something to think about the ROI on those companies. And if you, uh, from an investor lens in particular, how big do they have to be and how many problems they have to solve before there's really good ROI on those investments. Um, but I do love the awareness they brought to the world.

Manish: The third category, which I think is interesting to talk about in this idea of, you know, 2024, the year of ROI, um, for AI, is some of the infrastructure companies, namely Google, Amazon, NVIDIA, and Microsoft, that many of these foundational model companies are building on, they're clouds, right? Uh, if you look at the data, actually, the biggest funders of some of these foundational model AI companies.

Manish: are Google, Amazon, NVIDIA, and Microsoft. Um, and that's pretty interesting. Um, you have to think about why are they doing that? Um, and I think this is a nuanced thing, but I think it's critical to understand. Like, you know, why would Google invest in, uh, uh, a foundational model company when they're building their own?

Manish: Well, you know, these companies [00:09:00] have spent hundreds of billions of dollars in CapEx on huge infrastructure. They all have cloud services and want to sell more of that. Uh, they've been building them up over the last decade. And to make those cloud services, if you want to advertise them, the best way to do it is to show some of the best high end companies in the world, the Ferraris, so to speak, of technology are using that.

Manish: And that funding that these companies are getting is oftentimes in the form of credits. It's not just cash investment the way a venture capitalist would invest. And so that's an interesting thing happening where it's this kind of, Recycled full circle effect where a company like Google might give cloud credits to Anthropic.

Manish: Anthropic then spends those credits back in Google Cloud or OpenAI does that back in Microsoft Azure. And then it's a revenue gain for those companies and that, you know, helps them in terms of looking at and reporting those cloud businesses. So there's this kind of interesting cycle happening. So I think Many rents will accrue, as I like to say, to these large [00:10:00] tech incumbents because of this unique dynamic.

Manish: And then the last thing, which I think is really interesting to think about here, that is on the enterprise side, right? I think the story is different, uh, when you think about enterprise, uh, So the businesses of the world. So every CIO in the world today, and we talked to a lot of us based CIOs is getting enormous pressure from their CEOs, from their boards to figure out, well, what is our AI strategy going back to awareness in 2023?

Manish: And now there's pressure to do this stuff. And that is an opening opened a window for AI. We always ask the question and my team, why now, why is something going to happen now? What's going to change the inertia? The existing trajectory or something. And I think because of that awareness that happened in 2023, which is pretty incredible, everybody's asking this question of like, what is our AI strategy?

Manish: And this is similar to what happened, you know, in the web in the 1990s, you know, what was your website strategy for companies that didn't have websites? How are you going to get on the quote unquote information superhighway? In the two [00:11:00] thousands, people would be pressured to be like, what's your big data strategy in 2010s?

Manish: People are like, well, what's your cloud strategy with the emergence of AWS and some other incredible services? So now we're in this moment of a Cambrian explosion with ai. There's all this euphoria happening with ai. It's like, oh my gosh, it's gonna change the world. All these scary things people say as well.

Manish: It could end the world. Um, I think we're in this unique moment where people oftentimes with new technologies. Overestimate the effect of technology in the near term and underestimate it in the long term. So, from the enterprise side, I think all that excitement has opened the door to let enterprises start to experiment with AI technologies.

Manish: And the key word here I want to emphasize is that they've been impressed by demos, but not necessarily by things that are in full production yet. And I think that's going to be critical as we think about the ROI of AI. AI within the enterprise and 24, 24 and beyond. Because right now I think people [00:12:00] are so excited.

Manish: They're overlooking some of the basic requirements of any good system design. When you pull those technologies and enterprise, namely, I think there's a few principles that really matter. One, I think technologies and systems need to be deterministic. Right now, the challenge with some, a lot of these large language models and these systems that are using is they're not deterministic.

Manish: You could ask the same questions at different times of day and run the same things and get different answers. That's not good. Uh, you need to have deterministic behavior, right? You also need to have interpretable behavior. I think transparency is really key as well. You can't just have these black box systems.

Manish: People really need to understand what's happening with them. Um, in low stakes use cases, like a copywriting use case. Maybe not as important, but certainly in regulated use cases, whether that's financial or in health care, you really need to be able to unpack and understand what's happening within those systems.

Manish: And lastly, these systems are hugely expensive, as we talked about earlier, and as you can. [00:13:00] Tell by NVIDIA's rising stock price, you know, so much money is being spent on GPUs today, which are not cheap. And so these things have to just become more cost effective. So those three principles, I think, are really important as being about the enterprise side of ROI.

Mehmet: Manish, you unfolded a lot of, of actually questions that I had. I thank you for doing that. Um, Yesterday, of course, by the time we publish this, it would be like almost two weeks or two weeks and half. So I don't know why this idea came to my mind that I decided to put a short post on LinkedIn when I thought about what you just mentioned.

Mehmet: So I say, Hey, hold on one second. So everyone is rushing to AI and even the cloud companies, which are giving It's kind of a different marketplace slash economy that is happening, giving credits and then investing. And then I started to think about what happened with NVIDIA and the question that came to my mind, and you know, [00:14:00] I want to share that with you.

Mehmet: For years, people went by the famous quote of Mark Anderson, software is eating the world. Now, I would not say AI is eating the world. Do you think the chips, the AI chips, will eat the world? Do you think, like, this is where Because we started to see everyone want to become a chip maker also as well. What is your take on that from both like investment perspective and technology perspective?

Manish: Yeah, no, it's a great question. Um, I don't think the chip companies are going to eat the world. I think Nvidia is an incredible company. I would not underestimate them. But at the end of the day, my view is that the hardware does get commoditized over time. There's software layers on top, which I think, you know, build habits, and there's brands on top as well that I think build habits if you know, something is trained on GPUs versus the TPU or other types of chips just from a psychological, human psychological perspective, which have lasting power, um, you know, the [00:15:00] thing I think that's interesting to think about right now is if you rewind a I don't know, 20 plus years ago, 25 years ago, Cisco at one point was the world's biggest company, and they were the router company, right?

Manish: Um, and now I mean, it's a great company today still, but far from what it was at its peak. A lot of those technologies became commoditized, right? And so Uh, I think that's the thing we should really be thinking about as we think about these chip companies. It's, it's, there's all this pressure right now, and everybody has insane demand.

Manish: Um, I, I wouldn't necessarily think that demand is going to last forever. I think there'll be demand, but I think right now we have pent up demand. We also have highly inefficient models that use lots and lots of resources. Right. And over time, my view is that a lot of these models that are being used and are going to become smaller, going to be just more efficient, and it's going to, uh, uh, reduce the need on GPU.

Manish: There'll be a need. probably always there, but I [00:16:00] also think there'll be models that are trained on CPU and other new innovations will happen. So I don't think that the hardware will eat the world. It's an important part of the ecosystem. Just like a Cisco router is an important part of how the internet runs, but I do not think it's going to be the dominant, dominant position.

Mehmet: So it's like, um, you know, the wow effect at the beginning and then with time. Yeah, similar to any, any piece of hardware. Absolutely. Now, you touched base a little bit on the AI infrastructure in the enterprise space. Uh, but can we like a little bit give some insights on like what critical components of this AI infrastructure, these enterprises, they need to focus on to boost the productivity.

Manish: Absolutely. Yeah. And I think we're in this moment where, you know, there are a lot of different opinions. I'll share you my perspective on how I think about. The infrastructure and where the big opportunities. So I think there are five layers as I see the application stack, there's the [00:17:00] UI layer, and this is the thin layer that sits on top of everything.

Manish: There's been a lot of companies that have been built here. Maybe in most recent memory, the Jasper AI. Companies like Harvey, which is focused on legal space, you know, there's these layers, which are sort of the user interface between, uh, in a very particular vertical, um, I don't think those companies are going to be, there's going to be many giant standalone companies unless they build bigger services around them.

Manish: So that's that UI layer though. Um, so I wouldn't necessarily be betting on that layer. The next layer is the deployment layer. I think that's a really important layer. And I'm biased because I've made some bets in that space as well. Um, and I think that's really important because if you think about the programming paradigm for how AI is deployed in the world, it's a big shift from traditional programming.

Manish: It's almost like the advent of CICD. It's just a different way to do things. And I think how you deploy these models, retrain them, uh, run inference, et cetera, is very different, uh, than [00:18:00] traditional engineering. And that's why I think that deployment layer is a really important one for enterprises to focus on.

Manish: Uh, because you want to empower your engineers and use those layers of abstraction to do more. And so that's why I think it's very interesting to allow invest in those types of companies as an enterprise and also as a VC, frankly, um, because who knows what model will win the day. Um, I think that's a long running debate and those models will evolve and evolve the open source world, et cetera.

Manish: Um, and also just new different ways. That, you know, we've talked right now, there's a lot of excitement around the model layer of LLMs, but there's lots of other deep learning models, which I think are really interesting as well. And that's that third layer. So UI deployment, the next layer down is model layer.

Manish: Um, I think that's a layer where enterprises will not start to have to really think, how do I bring in some of these skill sets so that I have my own proprietary models and that my team's able to manipulate and use those, because those become the secret sauce, uh, as part of sort of [00:19:00] my competitive edge.

Manish: The next layer down is the data layer. And I think a lot of enterprises have already invested a lot in that, you know, over the last 15 years, this idea of big data, I think everybody started to believe that and did that and started to capture that data. Um, I think one of the things that's interesting about data though, it's, it's kind of like oil, everybody says data is the new oil.

Manish: But like oil, oil is only good when you refine it, processes do something with it, turning the gasoline, as it sits in itself, it's not that useful. Same for data. So yeah, I think there'll be systems that help us refine, process, just do more with our data, whether it's within these AI models or elsewhere. Um, and the last layer, uh, is the hardware and compute layer.

Manish: You know, I think that gets a lot of attention right now as we were talking about, uh, but I don't think that's an area where a lot of enterprises, traditional enterprises should be investing. Uh, I think it's the abstractions on top of that, you know, the big tech company should, because they have the capital and resources to build that infrastructure.

Manish: But if you were a fortune 500 [00:20:00] company that wasn't big tech, I would say that's a very expensive place to build. And the ROI is probably relatively low. 

Mehmet: Absolutely. I love the analogy of, of oil and data a hundred percent on that. And you know, it's like a kind of a coincidence. To your point, you put UI, you know, at the beginning and I agree with you because, you know, also something I started to notice, uh, you know, I worked in the B2B space for a long time.

Mehmet: I was sitting on the client side and I sit with the, on the vendor side. And now, you know, because I follow a lot of companies and what I started to notice that the UI in the concept of, you know, designing the beautiful menus and, you know, drop downs, it's like a little bit fading because everything's becoming like.

Mehmet: A chat based or even voice enabled thing, like for example, I can mention any example from a B2B space and say, instead of, you know, going and doing, finding the dashboard, you can just type, or maybe using your voice, Hey, [00:21:00] find for me the dashboard that shows X, Y, and Z, right? So the UI itself, I mean, still we need the design, how even the LLM model will show to us.

Mehmet: But I mean, it's became like. Like a de facto I would say, you know the way how we design it So 100 percent with you on this and and thank you for sharing that also manish now Talking about you know design and I know, uh, you know, you mentioned you you work In your early career at google now And you have the experience working on Google ads, the Google TV, the maps.

Mehmet: So how it's important to keep the human centered design philosophy, um, when designing, you know, the products, especially in the age of AI, because one of the feedback that we started to hear from people, yeah, but you know, like, it's, it's like kind of fake. It's kind of a, like we, it's AI generated. So how to do this balance?

Mehmet: Yeah. You know, keeping the product [00:22:00] really human centric versus, you know, all the, you know, the thought that people might have in advance about AI. 

Manish: Uh, I'm so glad you asked that question. I would love more people ask that question right now in Silicon Valley as they think about AI, because I think oftentimes.

Manish: The excitement ends up becoming around research paper or math, which are important, but I don't think that's the thing that's going to drive the best outcomes, uh, for consumers. Um, but it's the, it's kind of where the zeitgeist is right now. And when I think about a human centered approach to designing technology products, I think there are three components.

Manish: Um, one, and this is the framework I always like to use, whether it's when I'm teaching at Stanford or building or talking to my companies is, you know, One, is it feasible? You know, can, when you, when you think about building this technology, is it feasible? Does the technology even exist to build these things, right?

Manish: So if I want to build a teleporter, does teleportation actually exist? Can we figure out how to do that, [00:23:00] right? So that's the feasibility. Right? The second thing is viability, and this is kind of the business case, you know? So is it cost effective? Does everything have to be layered in gold and diamonds or something to be able to be, uh, used, right?

Manish: Is there only kind of one, uh, really rare element I have to go mine and it cost me a billion dollars to go be able to do this, right? So the viability, so feasibility, viability. The last thing is the desirability, right? And this gets to your question around feasibility. The human centered approach, right?

Manish: So what are those deep latent needs that a product is really solving for is in that, that's that desirability piece. When I think about those three components, if you were to imagine, uh, kind of a Venn diagram of circles, right? I think the most interesting innovative products are the ones that are in the middle where they think about feasibility, viability, and desirability.

Manish: At the same time and really lean into those aspects of human needs and those deep human needs, because I think the best [00:24:00] products that we see in the world today actually disappear from becoming technologies and just become part of our daily life. Um, you know, one example of my own lived experiences as somebody that was part of the team that built out Google Maps, you know, today, when people talk about maps, they often think about Google Maps, a digital map, right?

Manish: But when I was growing up. I thought about a map as a paper map or a globe, right? And so now that idea of the technology behind Google map is just a map. It's in the paper map is almost seen as antiquated, right? Um, it just disappears. It becomes part of our daily life. Even more true for televisions or like the radio, these things which were very novel, you know, 100 years ago or so.

Manish: And so as you think about AI, I think the same thing is going to happen where they start to, you know, Disappear. The best products will just become part of our daily ways. We do things and we won't even think about them. You know, the AI today that runs on our phones that let us use our capture our facial imaging to like unlock it.

Manish: That that has become ubiquitous today. People [00:25:00] don't necessarily think about us that as technology, it's just the way we do things. And so that human centered approach though, is I think how these companies today that are very, very focused on. The underlying infrastructure and the math and all that stuff will end up becoming great consumer products.

Manish: Um, I actually am not a big fan of the chatbot world, right? I think that I hope that really does evolve because I think that's such a limited interface in terms of What we could and should be doing with how people interact with AI. So I'm very excited to see the abstraction of all these, uh, tech companies that are getting a lot of attention right now to something even more simple, uh, the way Google basically made a very simple wrapper on a way to search the entire web 20 plus years ago.

Manish: I hope we can see that happening with these powerful technologies today and allow us to interact with them in much more sophisticated ways. Uh, they go beyond, uh, chat, just typing in chat, similar to what you said, whether it's voice or maybe it's multimodal, other things are just [00:26:00] much more natural. And I think that's when the exciting really things are going to happen because they're going to lean into my existing habits and let me just do more of those things.

Mehmet: Absolutely. And Manish, like you, because you mentioned now something and a question just popped up in my head, um, about, you know, differentiation. Now, some, some of my guests, you know, when, when we talked about AI and especially in the startup place, I mean, you know, so always we were thinking, okay, if the big guys can do this, should we, as founders, Like, should we be afraid?

Mehmet: Like, Hey, maybe, you know, like Google can make it like this. Microsoft can make it like this. So is this one of the differentiator that might let this startup actually grow, thrill, you know, and what other things you would, I would say, add to that to, you know, Make these founders not [00:27:00] get scared of the, of the big guys.

Mehmet: Right. So yeah, because they, they own the compute, they own the models, they own almost everything. So what should founders other than of course being like very, I would call it customer centric, human centric in the B2C space. So what other things they need to focus on to really succeed in this, Huge wave of, of AI startup that we are seeing today.

Manish: Yeah, it's a great question. And I think, um, I'll answer in a couple of different ways. One, I would really think about what your strengths are as a company, right? And if you are just trying to go head to head with one of the tech giants, um, I think that's a losing proposition unless you really understand Where you're at competitive advantages, because the scariest thing right now about innovation and what's happening in Silicon Valley is these large tech companies, which we've named are also relatively quick movers and adapt very quickly on.

Manish: [00:28:00] So they can copy ideas. They can learn from others very quickly. And they have at some level, as you said, infinite resource, infinite amount of talent that can go build and copy these things. Right. And I think you're going to see that behavior. I think you're seeing more and more of it. So that I think is something that's very real.

Manish: The way you defend against that is really understanding your strengths as a startup. Uh, I'll give you a historical example of this. There was a Google going back to my roots. There was a product many years ago called Google Video, and there was a product called YouTube, right? It was a separate company.

Manish: Uh, Google Video was very sophisticated infrastructure built by some incredibly smart people that had a streamed video on the web. YouTube was built by some very smart people that had a very product focused, very end user focused perspective. You know, famously, they ended up, even uploaded illegal content to their site just to get more usage of it.

Manish: Right. Um, and it was built on architecture and systems, which was far less scalable in Google, far more expensive. [00:29:00] Um, but YouTube won the day, right. And that is one of the, the biggest video streaming service in the world today. Google bought it for, you know, 1. 6, 1. 5 billion dollars many years ago and conceded defeat.

Manish: Um, and you know, it's kind of interesting that there's this kind of random startup that's That was far less capitalized resource than Google going exactly after the same segment of being able to kind of become the video streaming service for the masses, and they want, um, so, but I think they really understood what consumers were looking for, and frankly, they broke the rules where Google didn't break the rules, right?

Manish: Because they were worried, very worried about copyrighted content on that sort of thing. And if you think fast forward now to the world of a I do your question. I think that if you look at sort of how some of these large tech companies, uh, are dealing with, uh, AI creating consumer use cases, they certainly are being punished when there's all kinds of hallucinations and crazy things that are being produced from their models with Google Gemini and most recently in the news, um, [00:30:00] and they can't take the same kinds of risks as startups where startups can take those risks, but there isn't the brand damage and the kind of.

Manish: Uh, fear of losing. It's much more about we're playing to win mentality. So that's where they can then innovate much faster because they're learning loops are going to be much faster. Their cycles will be faster and they can innovate, uh, much more aggressively. Uh, now, if the, if the approach is like, well, we're going to have the technical innovation side versus the consumer product or really understanding the user side.

Manish: I think that would be challenging just given the scale of some of these large tech companies infrastructure. But if it's on the product side, I think that's a really interesting place to start. 

Mehmet: So it's all about agility and you know, the, the, um, Um, the possibility of, you know, also shifting, trying multiple things without, uh, damaging the brand.

Mehmet: So, so this is, I 

Manish: think that's why big companies don't innovate as well because they have a lot of bureaucracy, you [00:31:00] know, I often call it the physics of large companies, right? Like it's just, you know, you go from building a fighter jet to a giant tanker and it's just slow moving. 

Mehmet: Absolutely. Absolutely. Now, if we want to look at the AI space from, you know, of course you, you, you, you, you run a VC firm.

Mehmet: So what's your perspective on, on, on AI and what would be the role of In, you know, having the investment portfolio. So what, what are you seeing in that space? Uh, yeah, 

Manish: absolutely. So I think, you know, every, every startup it's, it's, it's almost like, you know, in 20, 2007, eight, when iPhone launched, every startup became like mobile startup.

Manish: Right. And now I think you see everybody becoming an AI company. Right. Um, I think there's some companies that are going to be core AI vets, you know, Or building new technologies that will fundamentally shift how everybody can use ai. And that goes back to that kind of [00:32:00] deployment level of companies. We have a couple in our portfolio like Union AI and, and Yurts, uh, which have done exceedingly well.

Manish: Um, I think there's other companies though, as they think about a broader portfolio of ai, which you're basically gonna use AI to supercharge what they do. Their core experiences will be about the consumer or doing some task, but AI is going to be a way they have, uh, leverage their data in a new and innovative way.

Manish: Um, I think, you know, the best example probably of that is if I think about what TikTok is and how that company has just created such an addictive. Experience by building by really leveraging their data and really being consumer focused. It's pretty incredible. I think there's gonna be a lot of companies that are like that, where they are really kind of a I first companies really are thinking about what is the core data that we need to collect and how do we quickly.

Manish: Learn from that data, [00:33:00] deliver a product experience that is really exciting for users, and then pivots and evolves quickly depending what user sentiment is. Um, I think the place that we'll probably see the most innovation, frankly, um, from companies that aren't kind of core AI companies is in the gaming world.

Manish: Uh, I think those companies are often the first to pioneer really interesting things. Uh, I was very fortunate to have a, uh, be involved with a gaming company. Uh, they got acquired last year for five building, excuse me, five billion by, uh, the, uh, a Saudi group gaming group, savvy gaming. And so, uh, you know, I think those types of companies though have the DNA to pull in technologies very quickly and have the user experience and real time nature of it to do pretty interesting things.

Manish: So that's where I would look. For innovation in terms of where AI will come out first. Um, but across the board, I think a lot of companies are seeing the power of what these, uh, large language models can do. And I think it's going to move, it's going to shift not just to LLMs, but more to deep learning [00:34:00] models in general.

Manish: Um, and people are going to start to build infrastructure as they build their companies to really understand, okay, my deep data modes are what. And how do I really think about that in a new and different way? Um, so I think it's an exciting time for us as venture capitalists. Uh, not just because of what's happening with the AI world, but the speed at which products will be adopted, you know, and I think there's a lot of infrastructure, so to speak in the world because of social media, that two folks in a garage in Silicon Valley or Dubai could build a product and it could touch a billion people within a few days, You know, and that's a really exciting thing.

Manish: Uh, for me, 

Mehmet: it's, it's, it was never like as easy as you mentioned, like before, as this time, you know, like, yeah. So to your point now, out of curiosity, Manish, like I asked you about like, what do you look into the AI space from VC perspective? I've seen it somewhere. Like, I'm not sure if you're [00:35:00] doing it or like you plan to do it, but I've started to see also the AI being used actually to evaluate these companies and evaluate.

Mehmet: So is this, is this really a space where we can really rely on AI? Um, to, to judge maybe a pitch deck or a business model, because I know some start, I know a startup, but actually they do the opposite. So they help founders using AI to build the business model, you know, build the financial projections and so on.

Mehmet: But from the other side of the table, is it something you think it's feasible? 

Manish: I don't think we're there yet. Uh, I, I do think that a lot of, uh, my junior. Team members here are writing their internal memos using chat GPT. So I think that is definitely happening. Uh, and I think they, I think we, we do use it as a tool.

Manish: These, these AI to kind of find data, but to, to do, to replace the venture capitalist, I think we're not there yet. Um, [00:36:00] it could happen. I think this question for me though, is really. Less about the models and the technology, but more about the data, you know, often what we do is series a investing very early stage investing and, you know, we look at deals from a perspective of people market product and then the investment to deal itself.

Manish: Oftentimes you're judging people. And I think that there's such an art form versus a science, you know, because you're looking at their life history, uh, challenges that they've overcome. Oftentimes they may be a first time CEO. I think that stuff is very difficult to capture in data that can just be.

Manish: Dumped into a model that can tell you good or bad, you know, over time, maybe that does happen because we all live in social media and there's more signal that's coming around how and who somebody is. Um, but I think we're pretty far from that reality today, you know, I think can help supplement decisions, but I don't think it's going to replace it.

Manish: That decision, you know, maybe for later stage investing, you think about public [00:37:00] stocks and that sort of thing. There's a lot more data out there. I think there's probably more advantages using AI in those situations. 

Mehmet: Manish, you just mentioned actually, what do you look for in, in, in startups? So of course the market, the product, and there's the, the.

Mehmet: The founders, right? So market, you know, as you mentioned, there are data points. Usually they have the dead homework and they bring that in front of you as a VC product, same thing, more or less. Now, out of curiosity, when you want to evaluate. The founders, what are like the traits you look for usually? And I'm asking this question, hopefully for fellow founders to listen to this and think about, you know, usually how a VC would evaluate you as a founder.

Mehmet: Forget about the Product forget about, you know, the, the, the market. So you can tell us about that. It's 

Manish: [00:38:00] a great question. And, you know, I think those four dimensions, by the way, are what most VCs look at found most look VCs look at investments by they just wait them differently at my firm and my team. We overweight on founders and market, which we, the product's going to evolve a lot.

Manish: Most products from series eight IPO have changed over the time. And what we look for is what we call founder market fit. First of all, you know, there's product market fit, but we look for founder market fit. So I tell the founders, you know, tell me a secret. Tell me the unknown unknown about a market. I often joke, if I know more about a market than a founder, I'm not investing.

Manish: You spend every minute of your day thinking about this. So tell me that secret, tell me that you need quirky insight. The thing that you see that nobody else sees, and then let's debate it. Let's talk about the validity of that validity of that. Excuse me. We also look for founders that can build teams, right?

Manish: They have something about. that they can pull people in with them. No, no founder is a solo. If you're only as [00:39:00] good as your team, right? So can they pull in people around them to help them go on this journey? Not because they raise a lot of money, not because they have a fancy brand associated with them, but because of who they are.

Manish: And that's really critical. Because, you know, so many great people and great talent in the valley have in the world, frankly, have so many choices. And you really want people that can kind of build that quote, unquote, cult of personality around them, that people really believe in who this individual is, and this individual team of managers are that are leading the company.

Manish: And also this broader vision of where they want the company to go, right? Because startups are frankly, Irrational things to do. They're very hard, most likely won't succeed. And it's a long journey. And so you want somebody that you enjoy working with and you're just going to want to follow along on that journey.

Manish: So really looking for that. So tell me a secret, can you build a team and then can you go and raise capital or, and, and, and start to like have that momentum in a company, you know, but you're going to [00:40:00] basically going from zero to one and can you have somebody that can really kind of. Uh, create that kind of momentum to become a juggernaut.

Manish: And so we like to look for those things. There's many different ways we look for those things. Um, sometimes it's oftentimes about the life journey though. What has this person done in life? Have they been, uh, great in what they've done? You know, were they the first person in their family to go to college?

Manish: That's an incredible achievement. You know, have they done something special in some part of their life? Right. Uh, do they really understand, uh, like the, the, the markets they're working in and have some deep insights. Those are the types of things we really like to look for for founders. Um, there is, I don't think there's just like one template though, because it's about that founder and going back to that market.

Manish: So those are some of the things we look on the founder side on the market side. Then, you know, is this person the right person to prosecute this opportunity in that market, you know, that context really matters. Um, and oftentimes I get excited when founders. Uh, tell me a story about a market and make me re imagine what that [00:41:00] market is.

Manish: Right? Like famously, maybe like, like Uber re imagining the taxi or Kareem in your part, you know, re imagining what is a taxi cab, right? What does that mean? That's really, really exciting because then you go, not just from being a market taking company, but a market making company, you're changing the whole shape of the market.

Manish: And that's really exciting as a tech investor. 

Mehmet: Absolutely. You know, I always tell. You know, people who comes to, to me and say, Hey, I have an idea. I say, what's your story? Right? Yeah. So, yeah, you know, because okay, you might have very cool product, maybe it's an innovative product, but if you don't relate it with a story, it would be hard.

Mehmet: Forget about talking to investors, talking to customers, because you know, you need to visualize this to, to, to people. Now, I'm a little bit biased to first time founders, Manish, and because, you know. I didn't have this experience of being a founder in the sense of a founder myself, but you know, I tried, [00:42:00] I'm passionate about startups, read a lot of stories and I've, I've seen, you know, majority of the time, the first time is the hard time.

Mehmet: And then you see these guys that they go after and they become serial entrepreneurs as we call them. So the second time, third time for them, it would be like a piece of cake. Now. For the sake, for the first time founders who I appreciate a lot, I can see the sweat, blood that they put to get things done.

Mehmet: What are the pitfalls that they have to avoid when they reach out to someone like you Manish, like, or any investor out there, what they should avoid? 

Manish: Absolutely. And I, by the way, I love first time founders and I love serial founders, right? I think, I think both are, uh, both there's history of both that have changed the world and done amazing things.

Manish: I think that for first time founders, though, um, I would say that, you know, when they reach out to a VC, let's talk about the reach out itself. I think the best way to reach out to a VC [00:43:00] is often through the VC's CEOs. Whenever my CEOs, um, in portfolio companies tell me, Hey, check out this person over here.

Manish: She's amazing. I always listen. So if somebody cold emails me, you know, unfortunately, what I'd like to say, respond and read every cold email. I don't. But when somebody I know, particularly one of my CEOs who I trust is saying like, this is amazing. You should look at it. I will always take that meeting. So for a first time founder who wants to reach out to a VC, I think that's really important.

Manish: The second thing I think that's important is for a first time founder really understand who you are. who that VC is. I feel like it's, you know, it's, there's, there's so many different, it's like flavors of ice cream. There's so many different flavors of ice cream, strawberry, chocolate, vanilla, what have you, you know, every flavor is not a match for you.

Manish: Just like you, you know, every engineer is not necessarily the right engineer for a particular company. Um, every VC is not the right VC for a particular company. [00:44:00] You're looking for that person that can really understand who you are and understand your point of view. Uh, so, you know, at my firm, we have three kind of core values that drive us.

Manish: Kindness, truth, and hustle. And kindness is one where it's really important, not only is it more fun to be a kind person, but from a culture of kindness, you can experiment, learn, become your best self. I think in a culture which is very rigid, it's difficult to do that. And so for first time founders, It's often about figuring out who you are as a CEO and who you are as a founder, because you haven't done it before.

Manish: So you really want to partner with folks that can help you become that best self. And hopefully that best self is somebody that can go build a multi billion dollar kind of business and company that changes the world. Um, you want somebody that can really support you along that journey. They're not going to do the hard work for you.

Manish: Um, that's your job, you know, you have to really do that hard work, but. Our job is to help you get there by becoming your best self. So those are the types of things I think founders should really research of [00:45:00] like, okay, who do I pattern match to? And spend a lot of time talking to the venture capitalist about how they work with their CEOs.

Manish: What has it been like? What is it? Where do they succeed and where do they don't? Because no VC can do it all. And you want to really figure out what does this person bring to the table? And is what they bring to the table what I really need? 

Mehmet: Absolutely, Manish. Like these are like really great advice from your side.

Mehmet: Now, the final, maybe it's traditional question I would ask as well. So we talked a lot today about AI. But because I'm sure you see and, you know, you receive a lot of pitches all day long, other than AI, you know, what do you think, you know, some of the emerging technologies that Would be I would say going in parallel with the AI because I believe AI will be everywhere.

Mehmet: This is my own belief There will be nothing without AI So but what what are like some of the technologies or emerging technologies that will go hand by hand, you know in [00:46:00] parallel with with AI In the coming few years. I will not ask 10 years from now because Things are moving very fast. Who knows? Who knows?

Manish: Um, so I think another trend that's really interesting to be watching right now as AI goes and develops is what's happening with observability. So how systems interact, not just the coding of data, but how systems interact with each other. But the health of my code, the health of my system, the observe, this category of observability is really important.

Manish: We've spent a fair bit of time in that space and I think, you know, perhaps most famously Datadog might be the, you know, it's like a 50 billion company, the most famous publicly traded observability company out there today. I think that's really interesting because Right now, we're at this interesting moment in time where system complexity is increasing at an incredible rate, and we're going to have to be able now to understand what is happening with these systems from a more sophisticated place because they're going to start breaking in funny ways, they're going to start changing in funny ways that we don't understand the ghost in the machine, so to speak.

Manish: I [00:47:00] think that's a really interesting sector that is going to get attention and continue to get more and more important over time, particularly because AI. It's just a different way to be deployed technology and use technology. Um, so I'm very, very excited about that. Um, I'm also very excited and I don't see many, uh, the next generation of consumer oriented companies.

Manish: I know we've talked a lot about enterprise and I love the enterprise. I think that's really where the value is going to happen over the next two or three years, but a little bit thinking about, okay, over the next five to 10 years, what is the next step? Snapchat. What is the next TikTok? I think it's really exciting moment because you can do a lot more as a lean team, uh, than ever before, as we were talking about early on.

Manish: So I'm very excited to see what companies come out of that. And oftentimes those, those companies are started by first time founders because, uh, they are coming with a fresh pair of eyes to it and are really just thinking about user behavior. 

Mehmet: The time will show us, of course, indeed, but to your point about [00:48:00] observability, I think in the enterprise space, especially Um, you know, I follow a lot of, of trends also as well.

Mehmet: Yeah. And it's becoming one of the major ones, especially with all what's happening with the AI, because funny enough, AI, it makes a lot of things easier, but at the same time, it has its own complexities as well. So let's, let's see how things will, will move on, but you know, we, we covered a lot today. And I really appreciate all the feedback that you gave if you want to leave us with maybe a final thought that You know before we we finish and how people can reach out to you.

Mehmet: Oh, 

Manish: absolutely So I think the final thought I think is Right now, I think we're in a really unique time in human history, uh, where, you know, I think there's going to be some amazing companies being built because if you look back to 2007, 2008, you know, [00:49:00] you had kind of a bad economy, you know, some crazy things were happening in the economy, but you had the emergence of this new technology, the smartphone, right?

Manish: And so many amazing companies were built Airbnb, WhatsApp, et cetera, Uber. And I think the same kind of thing's happening right now, but even bigger scale where you have. AI is a new platform becoming more accessible, kind of crazy economy in many different parts of the world, global economy, that I think is a really interesting combination for some amazing companies to be built.

Manish: So I'm very excited about that. And then to get in touch. Email Manish Patel. Oh, sorry. Actually it's Manish at NavaVC, M A N I S H at NavaVC. And then, as I said, for first time entrepreneurs. Our ceos go to get to them to get to me 

Mehmet: great I will make sure I will put the email address in the show notes Thank you very much manish, but just as a final thing you mentioned and i am happy you You mentioned this about people Please please please don't listen to the negative things you hear out there because to manish's point [00:50:00] the great companies that we know now airbnbs and uber and the others were I mean built during what the great financial Depression crisis whatever you want to call it and the same thing.

Mehmet: I think it's happening here. I'm just I was reading a report the other day, you know, like the the number of startups that actually started during this phase And it looks like very positive from projection perspective It's it's happening now in in the past two years after covid during covid, you know So just focus on your passion and go do it just go do it This is what this migration or my my my thought about this Manish, really, I appreciate the time that you spent today and all the insights you gave, uh, whether it's around AI and the investment part of the AI, what are you seeing from Trent's perspective?

Mehmet: I really appreciate that. And this is how usually I end my episode. This is for the audience. If you are a first time listener here, thank you for tuning and passing by. I hope you liked. this [00:51:00] episode. If you did so, please don't forget to subscribe and share that with your friends and colleagues. And if you are one of the loyal followers that keep, uh, uh, you know, sending me their notes, their suggestions, thank you very much for doing this.

Mehmet: And please let, Any thoughts you have, you know be sent to me because I like also to see if I can fix anything If there's anything I should do better I would love to hear that from you And of course if you are interested to be a guest on the show, you know how to find me mainly on linkedin you can reach out to me there or you go to Uh mehmetcto.

Mehmet: show and you see all the contact details Thank you very much for tuning in and we will be again in a new episode very soon. Bye. Bye