A Life Building & Investing in Intelligence/David Yakobovitch -AI Trends, Jobs, LGBTQ+ Advocate ~ Spark of Ages Ep. 12
Mar 28, 2024Season 1Episode 12
Rajiv Parikh
Email us with any feedback for the show: spark@postion2.com Find more great content like this at: https://www.position2.com/ Mention us @Position2Inc in Comments!
SUBSCRIBE TO POSTION2 FOR MORE EPISODES!
Episode Description: Prepare to have your mind expanded by David Yakobovitch, a visionary in the realms of data and AI, as he joins me, Rajiv Parikh, to unveil the technological revolutions that await us in 2024. We're not just talking about advancements; we're discussing a seismic shift in how businesses harness real-time AI applications to reshape operations and customer engagement. David dissects how AI-first strategies are becoming non-negotiable for product development and how startups are successfully digging into ultra-specific niches. If you thought AI was just an add-on, this episode will make you think again—it's rapidly becoming the heartbeat of modern business strategy.
But why stop at business? Imagine a world where your workout coach doesn't just count your reps, but is bionic; where your vision isn't limited to what's in front of you, thanks to smart lenses. That's the future David and I explore, touching on how AI, bionics, and human augmentation are merging to enhance our everyday lives. We dive into the crucial tech breakthroughs paving the way, from chip production to Li-Fi connectivity, and even the emerging role of humanoid workers. As we navigate these waters, we don't lose sight of the ethical compass—responsible AI use and the nuanced role of prompt engineers are at the forefront of this enlightening discussion.
We round out our conversation with a look at the societal shifts accompanying these innovations. The rise of Responsible AI stirs a mix of excitement and concern—while we witness an explosion of tech roles, we also grapple with the reality of job displacement. But there's a silver lining: communities and governments are stepping up with reskilling initiatives. Furthermore, we delve into how Data Power Ventures is making waves by investing in data-first companies, and the growing influence of community and tech advocacy in shaping the tech landscape. Prepare to be inspired by David's commitment to technology education and his personal moonshot to invest in a thousand data-first companies, anticipating a data-rich future where innovation knows no bounds.
Show Description: In every episode, we’re going to do a deep dive with our guest about what led them to their own 'eureka' moments, how they went about executing it, and perhaps most importantly, how do they get other people to believe in them so that their idea could also someday become a Spark for the Ages.
Email us with any feedback for the show: spark@postion2.com Find more great content like this at: https://www.position2.com/ Mention us @Position2Inc in Comments!
SUBSCRIBE TO POSTION2 FOR MORE EPISODES!
Episode Description: Prepare to have your mind expanded by David Yakobovitch, a visionary in the realms of data and AI, as he joins me, Rajiv Parikh, to unveil the technological revolutions that await us in 2024. We're not just talking about advancements; we're discussing a seismic shift in how businesses harness real-time AI applications to reshape operations and customer engagement. David dissects how AI-first strategies are becoming non-negotiable for product development and how startups are successfully digging into ultra-specific niches. If you thought AI was just an add-on, this episode will make you think again—it's rapidly becoming the heartbeat of modern business strategy.
But why stop at business? Imagine a world where your workout coach doesn't just count your reps, but is bionic; where your vision isn't limited to what's in front of you, thanks to smart lenses. That's the future David and I explore, touching on how AI, bionics, and human augmentation are merging to enhance our everyday lives. We dive into the crucial tech breakthroughs paving the way, from chip production to Li-Fi connectivity, and even the emerging role of humanoid workers. As we navigate these waters, we don't lose sight of the ethical compass—responsible AI use and the nuanced role of prompt engineers are at the forefront of this enlightening discussion.
We round out our conversation with a look at the societal shifts accompanying these innovations. The rise of Responsible AI stirs a mix of excitement and concern—while we witness an explosion of tech roles, we also grapple with the reality of job displacement. But there's a silver lining: communities and governments are stepping up with reskilling initiatives. Furthermore, we delve into how Data Power Ventures is making waves by investing in data-first companies, and the growing influence of community and tech advocacy in shaping the tech landscape. Prepare to be inspired by David's commitment to technology education and his personal moonshot to invest in a thousand data-first companies, anticipating a data-rich future where innovation knows no bounds.
Show Description: In every episode, we’re going to do a deep dive with our guest about what led them to their own 'eureka' moments, how they went about executing it, and perhaps most importantly, how do they get other people to believe in them so that their idea could also someday become a Spark for the Ages.
Hello and welcome to the Spark of Ages podcast, where we're going to talk to game changers of all kinds about their big world-shaping ideas and what sparked them. I'm your host, Rajiv Parikh, and I'm the CEO and founder of Position Squared, a digital marketing company based in Palo Alto. So, yes, I'm a Silicon Valley entrepreneur, but I'm also a business news junkie and a history nerd. I'm fascinated by how big world-changing movements go from the spark of an idea to an innovation that reshapes our lives. In every episode, we're going to do a deep dive with our guests about what led them to their own eureka moments and how they're going about executing it and, perhaps most importantly, how do they get other people to believe in them so that their idea can also become a spark for the ages.
Rajiv Parikh:
This is the Spark of Ages podcast. Today, we are sitting here in Park City, Utah, where we are having our growth marketing conference and this is our third year of doing this where we have amazing industry leaders, investors, we have executives from folks who have started the companies, folks who are scaling companies in the go-to-market sense and it's our third year. It's just incredible. One of the great gifts we have today is we have David Yakobovitch with us and he is here to talk with us about some of his experiences and his in-depth knowledge, and his greatest strength is in the area of data and AI, and he's a practitioner he's an educator as well as an investor. So, David, so glad to have you here with us. Rajiv, thanks for having me.
Rajiv Parikh:
You're a general partner at Data Power Ventures. It's an early stage VC fund for the data economy. You're also a venture partner advisor at Legendary Ventures. It's a VC fund focused on consumer retail technology startups Firms like Lululemon, pinterest and Airbnb. That's not bad. Very small companies it's not bad. So you're the host of an AI and responsible AI podcast called "Humain. That's right, yeah since 2018. 2018. So you're also a fellow podcaster, so you can rate me as to how I do after this.
David Yakobovitch:
Tens across the board.
Rajiv Parikh:
I'll take it. I'll take it as long as I get a tip for my Uber. Rather, that's perfect. David, you're also an active keynote speaker, guest lecturer at universities at Carnegie Mellon and Columbia University. You also started as the AI policy ambassador for Google's Global Affairs, where you were the AI principal advisor, who stress tested Google's AI-first software and hardware products, and you supported partnerships for Google for startups in New York City with UC firms and accelerators. David, you have a fellowship from Columbia University in the area of AI, an MS from the University of Illinois, Springfield and a BA from the University of Florida. You're a state boy just like me. Go Gators, Awesome, I'm a New Hampshire boy. His extensive experience and expertise in data science, product management, artificial intelligence, machine learning and entrepreneurship make him an absolutely ideal guest for the Spark of Ages. So, David, welcome to the podcast.
David Yakobovitch:
Wow, so excited to be here, Rajiv.
Rajiv Parikh:
So great to have you. So, if you were to talk about the data and AI landscape, what are the key opportunities that you're seeing? Landscape, or what are this, the key opportunities that you're seeing? Like you saw what's happening at some of the biggest companies saying, already, google with a, it was at the magnificent seven, maybe maybe it's the, the slug c6 now because they dropped tesla from that, but, um, you are looking at the landscape and technology, but you're also seeing the startup activity. So what are, like you pick the three of the, the three trends that you're seeing, the three notable things that you're seeing.
David Yakobovitch:
Yeah. So I think 2024 is definitely a breakout year for data and AI, but this started well back into the early 2010s and, as we know, 2017, the transformer came about and the last 14, 15 months is when LLMs hit the market. So I think one of the big trends for 2024 is to see applied AI truly in practice. 2023, a lot of companies were experimenting with chatbots. They're trying to do standard summarization, categorization with analysis, but now it's going to be moving more towards that real time.
David Yakobovitch:
We spoke about earlier that not only Google's doing, but a lot of other note-taking apps in real time, insights in real time, extracting value in real time. I think a lot of applications are going to move there. I think that's one area. Number two, I think there's going to be a lot of super niching down into verticals. So when startups are building products, you might not say I'm going to build for finance verticals. So when startups are building products, you might not say I'm going to build for finance but I'm going to build specifically for financial control, for regulators around loan approvals and finance right.
David Yakobovitch:
If you drill down into multiple levels, that's a better way to build products for an industry that needs change and get those use cases right. Beyond that, I think the third area is well, we used to say that every company was a technology company, and then every company became a software company in the last couple of decades. Software is going to eat the world. Software is going to eat the world. Well, guess what? Now AI is eating the world. Every company is becoming an AI first or AI powered company. So I believe in the next few years, every product is going to have some extension where here, use this chat, gpt integration or generate this image, or here's your legal briefing. Let's summarize the value, and so I think it's actually table stakes for companies these days to build in different AI first features, not for the sake of FOMO and keeping up with the Joneses, but to ensure that your product is giving the most value for your customers.
Rajiv Parikh:
That's really amazing. So the three things that you talked about are real-time products, getting super niche with data, and then digging into AI and making that an essential part of what you do as a company. That's right, is that, fair to say, awesome, so as part of that you've looked at many layers of the stack. Do you want to describe some of them in Raymond's terms?
David Yakobovitch:
Yeah, so the AI industry is always evolving, but what we have today is very much called the modern data stack, and this modern data stack basically means layer one there's some sort of infrastructure. So the infrastructure could be APIs, it could be databases, it could be where your data is stored and then how it's accessed. Then, often layer two you'll have different software to manipulate the data. These could be tools like Airflow, etls, kafka, pipeline or any of these companies to help massage, manipulate, transform your data. Then you go to layer three. Layer three is where you get these insights. So these could be the dashboards, the analytic tools building out to get the value that we talked about in email or through chatbot to surface those real-time results. And then layer four is actually the application layer. So if you want to actually interact with it as a customer or enterprise, you're going to have applications on your phone that do on-device chip management or let you interact back and forth like with a chatbot.
David Yakobovitch:
So I think those are the four layers. I think most companies are kind of still between layer one and three. Layer four is, of course, the hot, sexy AI that we're all moving towards, and I think we're going to see more of that this year, as every enterprise company saying we want to be AI folks, we want to have these applications. A few years ago we didn't know how to because the databases weren't up to par, Nothing was real time or the data was too expensive to manage. But now all those prices are going down, so companies are ready to go, is there like an order of magnitude price reduction?
David Yakobovitch:
So we're seeing the reduction in both price and then time for results. So things that used to take five seconds now could take 10 milliseconds. So we're seeing the speed order is 100x plus. Prices have generally come down 60 to 70% from a few years ago. So what used to maybe cost $100,000 to build a chatbot might cost $30,000 these days. So you know, it's still not cheap right?
Rajiv Parikh:
This is software. It's not 10x, yet it's not 10x, it's not 10%. It's not like the thing that really drove the internet or drove us into computing is how a mobile phone was a supercomputer. That is greater than what Apollo 11 used, right, so it's not that yet, but it's getting there Right.
David Yakobovitch:
And there's some areas that are just as expensive. We have the GPUs from the video, like these H100s and A100s. You can't even get a hold of their national assets. You see companies like Mark at Meta saying we're going to buy $10 billion of H100s so we could do our real-time inference. So put in the big order so that you can get first in line. Yeah, so we're looking at a lot of chip companies that are trying to optimize, solve LLMs faster, compete with NVIDIA right, there could totally be more players in the space. That's amazing.
Rajiv Parikh:
So there's a lot of growth there, so there's a lot of investment going towards the infrastructure players. Nvidia, or NVIDIA, it's the chip right, so the chip layer is really high and there's all the compute layers. And actually I was talking to a friend of mine who's an executive at one of the larger companies and they're talking about how they're signing multi-billion dollar deals for infrastructure and the interesting part about it is that they're signing those deals even before the applications have been built, that many of the applications, especially the enterprise ones, are still proof of concepts. Is that what you're seeing?
David Yakobovitch:
Yeah, so we're seeing today a lot of these companies signing large LOIs.
Rajiv Parikh:
Yeah, and I think it's what LOI is A letter of interest A letter of interest or letter of intent.
David Yakobovitch:
So you see some companies saying you want a billion dollars of purchase orders and they want to be first to it. First mover advantage can be everything I mean. Think back to OpenAI and ChatGPT. Right, they were the first mover advantage. Here we are more than a year and a half since the launch of these modern LLMs and they are the dominant player in the world. So sometimes that first mover advantage can give the company the opportunity to win a large market share.
Rajiv Parikh:
It's amazing, right. It's a whole field in itself, right, because you can go back to Google and say they weren't the first search engine. I remember back at the time in my days at AltaVista, we were the first search engine and Google went and thumped everyone. They were the 20th or the 30th, and Facebook's realm right, fenster, was way earlier MySpace was way earlier and then Better One came along.
Rajiv Parikh:
But First Mover can be huge here because of the advantages of mindshare, right, and I think one of the things we could definitely talk about is how OpenAI grabbed everyone's mindshare by being a consumer application that everybody can relate to, so they're getting more data to improve their engine.
David Yakobovitch:
It's a virtuous cycle.
Rajiv Parikh:
Yeah.
David Yakobovitch:
And it doesn't have to be all closed right. We're seeing other players, like Meta with Llama 2 and Llama 3, building an open source as well, so I think there's a lot of room for multiple players in the space. You even see out of Europe, like Mistral, building your own small model so you and I can go into our Windows or Linux computers and just run our own chatbot for free. So I think a lot of this technology as we saw how databases got commoditized over time large language models will be commoditized, but it's going to be the unique data modes. It's going to be how you can perform analysis and applications and people always want to pay for managed service products that package and make it easy to build your own.
Rajiv Parikh:
That's a huge area of interest, right? So there are companies that are getting billions of dollars in funding between OpenAI and Anthropic and friends that cohere, so there are folks who are getting billions of dollars funding. There's Llama, which is Facebook's open source version of what is an LLM language model, and then there are models you can literally run on your laptop.
David Yakobovitch:
Mistral right.
Rajiv Parikh:
Help us think about why you kind of hinted at it. I can run Mistral on my desktop, because what would I try to do with it as an entrepreneur or a buddy or a company? What would I do with that? Versus going to open ai and using their apis, their tokens, to build an app yeah.
David Yakobovitch:
So you know, when you look at these, like chat gpt4, 3.5 turbo, these models are huge. They're like trillion plus parameters. If you or I were to download it, it would take all the storage in our computer and if we try to run a command, it would just take an hour or two. We just don't have the compute and processing power, which is why we're, in that case, running on Azure's servers to get all the access to all these great chips to move really fast. Now the question is where do we see the future of AI assistants moving? And I think there's going to be three approaches. Number one is well, we're all going to have our own AI-powered assistant. These super-powered assistants mean we have to make it accessible for everyone and a model like Mistral. It's so small. It's like less than 100 gigabytes, which is small in 2024. It's very small, yeah.
Rajiv Parikh:
No big deal, you get on my iPhones, that's right.
David Yakobovitch:
Like 250 gigabytes, even some of them are one terabyte these days yeah, so you can download it and just simply classify articles based on that locally in your machine. So I think it's further increasing the accessibility for everyone, so that every device will have AI-powered assistance and they'll be able to use local GPU and local TPU or MPU or whatever kind of device we're working with. So I think in the short term, we're going to have these super-powered assistants as they commoditize. Over time, that's going to move into hardware. We're going to be at a point in the next couple years where, whether I have a pin on my shoulder or a smart contact lens or the Apple Vision Pro, all these are having real-time AI. I think this is where we're moving into bionics that are augmenting the human condition, and so we're going to be moving more there. We might even have a space in a few years where I have different augments on my arms that if I would do CrossFit, I could actually do heavier weight training and just put them on while I'm working, so in, so not.
Rajiv Parikh:
I don't need access to the entire, you know. I don't need access to thousands and thousands of processors for that. They've taught, they've. Whoever may build that bionic cross-fit trainer has optimized. Uh, they call it. What rag and retrieval that's right generation um version of it that just has data about what they consider best practices, data about me, data about what other people have done, but they've condensed it to something that can be on a local device.
David Yakobovitch:
Yeah, there'll be censures there that will see if some of my muscles are overstressing and it'll help compensate. So I think we're going to get to that place pretty soon. And then the long-term vision for these AI power assistants of course, we see Tesla, Humanoid and other companies out there exploring that space. Right, we're going to be able to say, like, if you're the Amazon warehouse, you need to lift 50-pound boxes 300 times a day, or can we have a machine support us there and have these humanoids that work alongside us?
Rajiv Parikh:
So it could be humanoids or it could be a sister of ours, I mean, but some of those movies were already. You're wearing that, so skeleton, and you're building, you're doing booms or you can have a robot potentially doing yeah and to do that we need a lot of chips produced.
David Yakobovitch:
Uh, we need to reduce the cost of compute and get that real time. You can't depend on going to a cloud server and then you're like oh, the machine got stuck waiting for the data. But we have a lot of new technology coming out. There's Li-Fi coming out. There's YR, which is going to replace Bluetooth in a few years. There's a lot of great new technology, but these take years and years to go from proof of concepts to fast enough for the industry adoption.
Rajiv Parikh:
So there's also underlying technology that's being improved to Bluetooth, to YR would be something that's more about the interconnection between devices at short distance. That's right. That's different. That takes longer to deploy than something that you deploy in the cloud. So why would you, one minute, use ChatGPT to get information? Because it's like you were saying, some of them are connected to Bing or other real-time search engines. And then why would you use Perplexity? Why would you then switch to Cohair's version? Why would you switch amongst them.
David Yakobovitch:
Well, I think you know, being only a year and a half into the LLM craze, I think it's worth, as consumers, trying out the different products. They're all going to be different experiences. They're all trained in different data sets. So if you ask the specific question about like how do I play pickleball from scratch, and you ask that to OpenAI versus Cohere, you might get a different response based on the data.
Rajiv Parikh:
Do you have an example of one where one was better than the other?
David Yakobovitch:
It varies wildly with not only every question that you ask but the intent of your question. And so we're seeing, actually, the rise of a new career opportunity called prompt engineers. We're seeing people who are getting hired full-time for companies to literally figure out how to best write the prompts so that the LLM engine gives you the best result back. And so that means when you're generating images. If you say, for example, I want to have a city of a hotel in Park City covered with snow and cows running throughout the field, well that kind of image may generate a different result when we are right now.
David Yakobovitch:
Where we are right now versus At the top of a stable in an office. This is so true, but depending on the language that you use, because these large language models are based on languages and the English intent it could be different for each model. So I think that's number one, a different reason to try out different ones. Number two, it's the speed of them. So if you're paying premium for one of these models, you might get faster results than others. I know open AI and anthropic have these really long context windows so you're able to run chats for days and days. You might not be able to do that with other models Right. So what used to maybe take me 10 to 20 hours, I could do in two to three hours. Or instead I could still do 10 to 20 hours but do 200 hours of productivity, so you could reduce the time it takes or you can get that much more accomplished to accelerate your work.
Rajiv Parikh:
I did an AI marketing conference Not the one that you and I did, but we did one on privacy, security and ethics and I basically went into one of the engines and asked what questions should I hear the people who are coming? What questions should I ask? What questions are they likely to answer? Give me tests that can help me overcome these issues. Are coming, what questions should I ask? What questions are they likely to answer? Give me tests that can help me overcome these issues. And then, uh, and I asked him a bunch of questions and I shared it with the people I'm presenting with and they're like oh, this is great.
David Yakobovitch:
so we saved hours of time and effort we still got great answers from people, but we move them along much faster yeah, you could see the questions, you could prompt the questions, and it doesn't mean the model is going to give you everything. And so that's my biggest caution to people is about responsible AI. Don't assume that the results the model tells me about these math functions is correct. Use my brain, think about this, ask the question again a different way.
Rajiv Parikh:
So you get the citation, you go check up on it and make sure you're not getting a hallucination, you're not getting something that's potentially false. Is that the notion of responsible AI or is there something more? Is it more about ethics and bias?
David Yakobovitch:
Responsible AI. There's a lot to unpack. One of them is around the hallucination right. Ensure that you're giving fair Responses to an audience member because, look, when we're chatting here, you and I are going to research it, but like is my mom going to honestly research it? She's going to take it as true face value yeah.
Rajiv Parikh:
So as an investor, how do you think about it?
David Yakobovitch:
Yeah, well, I think what's exciting about what OpenAI did is they unlocked the kimono for AI. It used to be this very hands-off topic.
Rajiv Parikh:
It blew up in our imagination.
David Yakobovitch:
Right Now, hundreds of millions of users can use a text prompt editor and get real results. It's not, like you know, yesterday's AI technology when we think of the Jetsons so one that accessibility creates the possibility for consumers to adopt the products and that means more people are going to build in the ecosystem and it becomes that virtuous cycle.
Rajiv Parikh:
Right. So now you can um, if you're, if you're a budding entrepreneur or even an executive or a developer at a company, you can now work off of the marketplace, in with open ai or some of the other players right, and build your own applications yeah, and I think that's the moment that open ai is having this year.
David Yakobovitch:
They're having your apple app store moment by launching uh earlier, earlier in the start of 2024, these GPT stores, right.
Rajiv Parikh:
Do you have two or three favorite startups in that area?
David Yakobovitch:
Oh, you know one of my friends, reuven Cohen. So he's actually he calls himself an ethical prompt hacker, so he spends all his time building out different prompts and he's built applications to like automate his calendar and like just look at all of his emails daily and triage which ones he should respond to. I tell a lot of people in my life today, as an investor and product leader, that I actually don't spend much time in search. I spend most of my time in chat interfaces. I'm spending my time between perplexity, between coherent, anthropic, open AI, llama, mistral, falcon Almost everything I don't search.
Rajiv Parikh:
David, with all the stuff that's going on with AI, there's a lot of legitimate fear about, especially since it's happening so quickly. There's a lot of legitimate fear about everyone losing their jobs. If I have AI as my co-pilot tutor, why do I need tutors? Maybe I don't need teachers, maybe I don't need folks who can write code if it's writing code for me, if my prompts are just in basic English, prompts are written for me. Or why do I need to get my PhD to do something if the system already is a PhD? Or even in a more straightforward sense, I'm looking at loan applications and I'm filing them away and the system's doing it all for me. So there's jobs at every level, and especially at the more educated levels that are now being affected. What's your thought on this? I mean, it's quick, it's jarring, it could be tremendously challenging, exciting for someone who loves technology, or it could also be scary.
David Yakobovitch:
The fears are real. They're definitely real, Rajiv, and the example you shared about our AI-powered tutor. After OpenAI launched ChatGPT, Chegg a public company spoke about its earnings and one quarter later they were down 98%. Everyone just completely abandoned their products. So you've seen real jobs being lost every day because of AI, because of these co-pilots being in the world, and the challenge is figuring out how does policy manage with private tech change and there's always a gap. And today we're seeing the EU Commission launching AI policies, Two of our portfolio companies, one of them, Vera AI, that does responsible AI for startups. Liz O'Sullivan spoke to the Biden administration to actually propose stronger regulations and policy. Another one of our founders, Francesco Marchionne from Applied Excel in New York. He spoke to Boris Johnson's administration in the UK also about why we need stricter policies and regulation Early in 2024, I also spoke to the information about this. With responsible AI innovation, I think we have to move forward faster in policy because it lags behind and jobs will be at risk.
Rajiv Parikh:
So, but you can't regulate something that's happening in technology, right, but you absolutely can. So in Europe.
David Yakobovitch:
Several countries have banned self-checkout lanes at grocery stores. They're completely banned. They're forcing and allowing people to keep jobs for like a universal basic income.
Rajiv Parikh:
So maybe provide a transition point to it.
David Yakobovitch:
It could be, and some of that transition is reskilling.
Rajiv Parikh:
It's not like New Jersey, with you know they still forced to fill your gas, whereas everywhere else you just do it yourself.
David Yakobovitch:
Yeah, new Jersey, we're not going to pay for your taxes. Okay, fireworks, go in the East Bay. You know, look how quickly that could change. But I think it's going to require a lot of reskilling, upskilling initiatives. This takes the government to come in. They're going to have to support. What are the new jobs going to look like? And there are going to be new jobs. When we think about 10 years ago, what were the top jobs? Lawyer, doctor, maybe software engineer, a couple others. As of 2023, 2024, top 10 jobs Data scientist, software engineer, site reliability engineer, maybe prompt engineer, application developer, front end, back end it's all software jobs.
Rajiv Parikh:
So people have to be re-skilled and what governments can help and our community can help is by helping people re-skill themselves.
David Yakobovitch:
Right and we can't make it the onus of you and I to reskill. There has to be support from society on that and new jobs will come about. But these are real fears and they do need to be validated. People who are struggling it off, they're living on another planet, I mean. Most recently, the World Economic Forum, early on, said 40% of jobs are at risk, and I think that could even be an understatement, right. It's just what is the timeframe for those job shifts and changes?
Rajiv Parikh:
In my company, we're already seeing a 30, 40% increase in productivity in using various AI tools for everything from developing videos, creative writing code. The easy answer is just to let folks go. The better way to go is say you know a lot of things, why don't we do a lot more? Or do things in more depth or be more personalized about how we create these campaigns. Let's go hyper-personalization. You really need those folks to drive all these things.
David Yakobovitch:
Yeah, I think you could have more leaner teams now, but it doesn't mean you have to reduce your force. It means with your current teams, you can augment and do more and have more productivity per employee and generate more revenue. And so my aspiration is that companies would be forward thinking rather than short-mindsided on cutting the bottom line.
Rajiv Parikh:
So now, over the last few years, you've been building and investing in data products for businesses. So when you were at Google, you were been building and investing in data products for businesses. So when you were at Google, you were working on data products. Can you explain to the listener what that means?
David Yakobovitch:
Sure, so I think in the last 10 years, data products have evolved. Any software that we work with today has multiple surfaces. Those surfaces could be things like email, chatbots or actual reports, and often that data is static. But what if that's real time? What if you're having pipelines where you're getting real analysis and insights for executives, for sales leaders or for operations to make business decisions? So at Google, we call data products that whole culmination of data intelligence, sales intelligence across all those surfaces.
Rajiv Parikh:
So what's an example of that? So you weren't in the Google. Everybody knows Google for Google search, right, and it's historical. But it can also be pretty much real time, especially Google News and tweets and that kind of thing. But this is something different.
David Yakobovitch:
Yeah. So one thing that's well known in the industry is that, outside of IBM, google is now the number one company for the most amount of internal tools managed of any company. There's literally tens of thousands of tools that are all in-house built. If you think of tools that are external, like Asana or like mondaycom, google has a version of every single tool built and maintained. Part of that's the proprietary edge of ensuring that you're building something that's secure and transparent and keeps the data locked in one central place, but also it's got the customization for your customer needs. Specifically, a lot of the internal tools I worked on included CRMs, email integrations, live chatbots with sellers so all these tools we're seeing in the industry today. We have our own versions of these in-house.
Rajiv Parikh:
So an example of a situation where you need a real-time data product if you're in sales or marketing, what would that be?
David Yakobovitch:
Yeah. So think about when you're bringing on a new customer to Google ads. You start with some sort of acquisition channel and to get them to onboarding and to start spending actively and building in the Google ecosystem, you have to make sure you don't churn that customer. And in the small, medium business area of Google ads, we see millions and millions of new inbound every single year. So if you could have real-time insights for sellers, who are basically account managers, to understand when there's a greater propensity of a customer to convert or they need more support, or they don't need that support, or they need an insight that will help you, help your customers. And so today we have all these real-time dashboards that every single seller can view for all their accounts and extract insights from different sources and then either share them directly with the client or use them to better manage their productivity and their performance for attainment and measurement.
Rajiv Parikh:
Right. So they literally can be on a call with someone that's advertising on their Google network and, as they're going about the discussion, if there's information that's coming up about a competitor or a potential better ad that they can run or a different technique they could use in managing your Google account, that's going to be in front of them.
David Yakobovitch:
You'll either have insights as a collection, before all your new calls, of all the old calls and summaries, but then in real time you could also get those insights, and of course, we see a lot of platforms building those. Today, google does most of that in-house.
Rajiv Parikh:
Like is that part of the 20% of time that Google engineers have, or that's something that's part of their normal project?
David Yakobovitch:
Right. So it's incredible to think that even today, almost you'd say 80% of people at Google are engineers. Right, there is a lot of the business org of product managers, partnerships, dev support, but engineers are full-time focused on these internal tools, yeah, so, for example, these ones that I work on, we have thousands of engineers managing the entire google ads ecosystem, both internal tools and an external facing, like pmax and others that you and I pmax right yes performance max right, which is the one google tool that they tell us will rule them all.
Rajiv Parikh:
That where it will, based on the goals that you give it, it'll automatically manage your ad spending across display, video and search right. So one of the things my my folks like to play with uh note to everyone. It doesn't work in every situation, which is why you hire an agency. So you uh made a pivot in your life where you were in, you were working, you know, a job in a company, right? This was when you were in education, and then you went and started a venture firm. You just woke up one day and said, hey, let's write, let's write checks as a venture firm.
David Yakobovitch:
I think it's always been. There's always a transition point. It's always been like a dream of me to like coach a lot of founders and support them, and at a couple of my previous startups I was at Galvanize and Single Store At Galvanize it was already my fourth startup I was at and I was like I want to provide some more support and that's where I started doing these direct checks 10, 25k checks just getting my feet wet Starting feet, wet playing with the ecosystem.
Rajiv Parikh:
Yeah, I was an EIR Figuring out what kind of founders are really going to be good.
David Yakobovitch:
That's right. Yeah, I mean, I was an entrepreneur in residence at Techstars in New York City, so I was seeing founders cutting checks into their portfolio and then at one point I realized well, hey look, I come from an immigrant family. I'm first generation. I came from lower middle class, worked my way up Like I don't have an unlimited checkbook.
Rajiv Parikh:
So you had empathy for them.
David Yakobovitch:
I had empathy for founders and I said, well, how can I further support founders with bigger checks? And that led me becoming that new generation of syndicate leads, like we see with Alex Patis and Morgan Schwank and others who are leaders in the space. I was following in the footsteps of them as giants. Of course, doing that now fast forward many years. I'm probably known as one of the top 25 syndicate leads out there in the industry. When the pandemic happened and when the non-recession recession of the return from COVID, a lot of these leads left, but I'm this person who's very persistent and disciplined.
Rajiv Parikh:
I've stuck with it over the years. I think it's because you know the core of these things. You understand the power of the data. Data is the new oil, so it doesn't matter, right? You know that, regardless of a recession, these things are going to be persistent.
David Yakobovitch:
That's right and by doing that I was able to upsize my checks from 1025K checks to 100 to a half a million dollar checks. And it's been amazing to support founders because often when you're pre-seeder seed, a VC firm doesn't want to cut a two, $3 million check. So having these party rounds, having these operator VCs which is who I became supports founders that they can build their early milestones and then get the VC funding.
Rajiv Parikh:
You know, at the time you were at Google. You're also, at the same time, you're a venture capital partner. Why did Google let you do that?
David Yakobovitch:
Yeah, so my story about getting involved in the VC ecosystem started well before Google. I spent over 10 years working for five different startups as a tech operator. Two of these were like the Series A type, two were Series B and C, which got acquired, and one was late stage pre-IPO. Two were Series B and C, which got acquired, and one was late-stage pre-IPO. Along that journey, I got this infectious bug that I wanted to support founders, to coach and advise them, and I started very early on, around 2018, 19, writing my own direct checks.
David Yakobovitch:
That grew into launching my own syndicate SPVs, launching my own emerging fund, which became a portfolio of 35 data, ai, deep tech startups between New York City, silicon Valley and Seattle. So when I joined Google, it was just an extension of myself. I'm someone who loves community community first principles, supporting founders and so, outside my core work at Google, part of my 20% capacity is I was mentoring startups, going to Google for startups, for Accelerator, supporting the accelerators in New York City like Antler and Techstars and ERA, and for me, that gives me a lot of life and a lot of energy supporting startups. So of course, I couldn't give that up and now I've moved even closer in that space.
Rajiv Parikh:
I think that's amazing that you were able to continue along with that opportunity. It's because most of the time when you join a company like Google, they don't want you to do that, and that's amazing that they gave you room for it. So you may have a fund that's of a certain size, and let's say it's $10 or $20 million, which I'm not saying that's where you are but you can extend that fund by doing what they call these syndicated rounds or they call it SPV, special purpose vehicle, that's right.
Rajiv Parikh:
Where, using some software like AngelList, you can literally say oh, do you want to write a $10,000 check or a $20,000 check to support this round? And so you're the lead, right? So, david, your fund may be the lead no-transcript.
David Yakobovitch:
Sometimes they don't want to let a founder know that they're investing, but they want to be part of that ecosystem. Other times a lot of investors want to be passive. They don't want to go direct on the cap table because there's a lot of process to manage and, as a syndicate lead in my early days I'm acting as a fund manager and of course now I am a fund manager, but it was managing that process through and through for hundreds and hundreds of investors.
Rajiv Parikh:
It's really helpful. I've done a bunch of these of like the venture firm, like I might be an investor in a seed stage fund, and then all of a sudden, one of the startups they want to do a bigger round, they want to hold their place what they call pro rata. So they're like, okay, you can jump in for some minimum amount and I really like it, so I'm in and I can play in it and not have to deal with all the disclosure and being in meetings and being listed on the cap table, like you're saying so.
David Yakobovitch:
I think that's a great way to magnify yourself.
Rajiv Parikh:
Thank you, thank you, but now you're doing a lot of this in New York right.
Rajiv Parikh:
And maybe you could correct my thinking about New York. I love New York, but I grew up in New Hampshire and the reason I went to Silicon Valley was because I saw such an open mindset. It may have come from the social tolerance that people had for each other, the openness that it had. I felt, like in the East Coast a lot, that the answer every time I had an idea the answer was no, but here's a hundred things wrong, whereas in Silicon Valley it was like yes and and, let's riff together. In the end we may come to the same answer, but there was an openness to discussion. Have you found that things have changed?
David Yakobovitch:
New York's a very special place and, for those who don't know, a couple of years after college, in 2014, I was at an inflection point. I was at this like quarter life crisis and I was deciding I want to leave Florida, do I want to go to San Francisco? Do I want to go to New York City? I actually spent almost three months living in San Francisco in 2014. I was like I'm going to try this out and I was like hosteling it up and like meeting people and going to the tech events, I was like drinking the Kool-Aid.
Rajiv Parikh:
Not just tech, but social life. Wise, I mean very open to everything that you want to do, right, I mean, it's just amazing people that are it's a huge amount of tolerance yeah, that's right. And acceptance yeah, and it was Not even tolerance acceptance.
David Yakobovitch:
It is acceptance right, and California is always known for being one of the most progressive states out there, and it was a great experience. But I think for me then what was lacking is the focus was way too much over-indexing on tech, and I truly wanted to be in a city that I believe is the center of the universe, where you get a little bit of everything. You can work hard, play hard and you get every industry. I think also I felt a little homesick for my family, my parents. Before I was born they moved to South Florida, but I have over 50 cousins in the Northeast between Connecticut, new Jersey.
Rajiv Parikh:
New York. That must be great.
David Yakobovitch:
And I wanted to get close and get to know my extended family, my grandparents, my aunts, my uncles. So I think, with those indications, I made the jump to New York. You know, crashed on a friend's couch right, got my way started Thinking back to New York. Crashed on a friend's couch got my way started Thinking back about New York's evolution over the years. I agree with you, back in 2008, 2009, tech was not a thing in New York. I think it started to get evolved around 2010 to 2012. You had FFVC, union Square Ventures a lot of these firms that built the foundation of New York Ventures a lot of these firms that built the foundation of New York. Of course, we know, as of 2022, almost a quarter of venture dollars in the United States goes to startups in the tri-state. That's awesome. Half is still Silicon Valley and the other quarter is the rest of the country.
Rajiv Parikh:
Yeah, like you I'm a big fan of. It was like oh, there's investment going in New York and Austin and Boston all over the country and I'm like this is great, this is what we want. We want entrepreneurs everywhere building great companies. It's the right. You can't just say, well, one area is the center of the universe, as much as you want it to be, as much as you think you want it, it's actually better for the whole network that there's multiple yeah, I think so from a tech perspective and so we've seen that.
David Yakobovitch:
And I think so from a tech perspective and so we've seen that and I think the other decision that broke me to go to New York versus SF I'm a very proud advocate in the LGBTQ community Awesome, and I was trying to find my identity in San Francisco.
Rajiv Parikh:
And there's too many advocates in San Francisco, there's so many advocates.
David Yakobovitch:
And I made friends and I met people and, as you said, it's a very open culture. But at that point in my life I was like I think New York would provide the support, the nurturing I needed.
Rajiv Parikh:
It's also a very supportive place, right yeah.
David Yakobovitch:
And so that you know, of course, I found my tribe and tribes in New York over the years and today I spend a lot of my focus as a community builder outside of tech, also for the community, you also for the community. I put on events that are very much non-party focused, which are about networking for people in finance and tech and bringing together the community, whether that means on the VC side with folks like the Gangel's team, or working with Out in Tech, where I mentor founders who are coming from diverse backgrounds, who maybe struggle to get the meetings with a VC and be like oh yeah, you're an LGBTQ founder, I'll take that first meeting, no problem, let me give you some feedback, let me give you some support, so they have a person that's been in their world that can put out their hand and help them up.
David Yakobovitch:
Yeah, I think it's super important to always have people who are advocates on your side, and so that's been a real pleasure. And now, being 10 plus years in New York, I found my community, but I'm also a community leader much in the side of product, in the side of LGBTQ, in the side of just general community events, putting them on, and it's exciting to bring people together let's have an amazing group of friends.
David Yakobovitch:
It's a lot of fun. It's a lot of fun as long as you start with fun first. It's not about are you building the next unicorn We'll all get there. You know, one of my biggest success stories in New York is backing an LGBTQ founder. I actually backed David Stein. He built Ash Wellness out of Cornell Tech and 500 startups Started supporting before the pandemic. Now we support them every round through their seed extension that are in Series A and they build these remote diagnostic testing kits where you and I could do our blood or urine testing at home and then send those results out and find out. You know, hey, we're hiv negative and let me get my prep, or you know I'm just someone who needs to get whatever of 150 plus biomarkers. So it's really great to support the community. And, um, you know I talk the talk and walk the walk.
Rajiv Parikh:
You definitely do, I think, jim Cascade. In one of our earlier podcasts we talked about how quickly this has happened. So if you look at how long it took for TV to be adopted, or how long it would take to get 25% of Americans to use something like that, it would be 40 to 50 years. Electricity was like 60, 70 years. Your mobile phone would be seven or eight years. This is like a year or two.
Rajiv Parikh:
And it's because the cloud is already there, the device is already there, it was really quick to deploy, so there's push and pull. There's some things that you can deploy in the cloud that you can get super fast. There's some things that need every device to have it. It has to be part of the infrastructure, so they move slow. These are tools that make you feel like they're God right, I mean it's you don't like it's coming down from the heavens, because it sounds so intelligent.
David Yakobovitch:
Yeah, and you know, the other day I was chatting with my parents and they were scrolling through their Facebook feed and they said, oh, isn't this true? Didn't you do this? And I said, no, actually this post and that post had nothing to do with each other, but Meta surfaced it for you, and so it's incredible the gap in tech enablement we're having between generations today, and we need to provide more support for silver tech, which is our aging population, and I think a lot of companies are not being responsible enough to support they're thinking of the zillennial and the millennial and us, you know, gen Y and we need to provide better safety mechanisms so that wrong actions are not occurring. So one, it's about that fairness. I think beyond that also is about bias right, so that wrong actions are not occurring. So one, it's about that fairness.
David Yakobovitch:
I think beyond that also is about bias right. So if my mom does a search versus a woman in Iran, does a search right Are they gonna get results that are personalized and are safe for them, or are they gonna get data that could be completely wrong and could actually be offensive? And that's why, when we think about the political election that's coming up in 24, just earlier in the year, sam and the OpenAI team specifically said we are banning all political questions. Any questions you ask this year, openai will say I cannot answer For any local elections, state elections, federal elections they do not want to be involved in misinformation and disinformation campaigns.
David Yakobovitch:
These models today for generating images have gone so good. If you and I want to use mid-journey, we could literally generate an image of Obama and Trump shaking hands in 10 seconds, high-fiving each other, high-fiving each other, doing all the hands. Making it look like a real.
Rajiv Parikh:
You put it in like that, did they sign an agreement?
David Yakobovitch:
No, they never did that. That's fake news. Yeah, but if you're like that did they sign an agreement? No, they never did that.
Rajiv Parikh:
That's fake news, yeah we can do a lot of things and that is one of the biggest concerns about this election and any election going forward. Let's go to something more back in your daily life, which is now you're back at Data Power Ventures. It's an early to mid-stage, early to Series A fund Like $100 to-200 million. Is that the fund? I'm sure you're managing hundreds of millions. The fund itself, over multiple funds, is managing quite a bit. What's your go-to-market for a fund like this? We're sitting from where I am in Silicon Valley. There's lots and lots of great funds. How do you differentiate?
David Yakobovitch:
Data Power Ventures specifically, is focused on that data first principles. We actually launched back in 2020 something called the Data Rider or Data Pledge, similar to when Act One Ventures launched the Diversity Rider.
David Yakobovitch:
And this is telling all of our companies that we invest in you. We promise to be good fiduciary of data access and we require you as founders to do that as well of data access, and we require you as founders to do that as well. Similar to some other great programs, we have experts and residents all across different CXO stacks and levels. So founders that were investing and backing as pre-seed or seed can get that support before they become scale up to series A. So really that sweet spot's coming in very early, offering that technical support, offering that product marketing support, because often these founders they're brilliant, they could build anything in technology, could build the next Mistral AI, but if you don't know go to market, if you don't know how to build the business model, innovation it's just a pipe dream.
Rajiv Parikh:
I think one of the greatest things beyond a lot of times, people in their mind just think it's a check. And it's not just a check. And it's not just a check right, it's people who are going to help you make a difference. So your strength is that you focus on data, you and you help them with all the other things. They may be great technologists and they have great ideas. You're guiding them on how do I take that product, how do I get it to market? How do I do my product market fit? Maybe eventually go in and talk about their go-to market fit. You're helping them. You're going with them in depth.
David Yakobovitch:
Yeah, because when you think about companies that are in this classic pre-seed and seed stage pre-seed companies don't often even have a product yet and see they have a product, but maybe it's 100,000 revenue, maybe a million revenue. They have not hit product market fit, so they really need a lot of support. These teams are often still four to ten people and the best support is outsiders, right outsiders who don't have necessarily a strong opinion about your product and can tell you well, this is what I'm seeing in the industry and have you thought about it? With this other approach?
Rajiv Parikh:
right. So it's really helpful to have that community with you. You bring your team that to them. So let's talk about what drives you, why you got into this field. Was there just a youthful thing that sparked you about technology? Was?
David Yakobovitch:
there something you did when you were younger, I'd say back in middle school I did math competitions Just for fun. Just for fun, you know how fast could I solve math by hand. Or could I use the TI-84 calculator to do some statistics?
Rajiv Parikh:
That's a super high-tech calculator. We love those right.
David Yakobovitch:
It was I mean those were amazing calculators.
David Yakobovitch:
And that led me to state and national competitions and RML and all these top math competitions. Now, these days, our LLLMs can do all this for us. So just earlier in 2024, there was a research report that the Math Olympia was solved by some LLM. So students like myself and others, who dedicate years and years to learn the most advanced proof concepts, now it's all solved. Should you still learn that? You should still learn it, I think. But I think education is going to shift. It's no longer going to be this space where you and I manually learn something at a slow level. We're going to move to a level of education where educators can tailor education for each student. So perhaps you have students with learning disabilities who can learn a little bit slower in the class, get support with large language models and other tools to help get them up to speed. But do you feel like there's a?
Rajiv Parikh:
concern, like if you don't actually walk through how some of these things are. You know your multiplication tables and getting into algebra and learning about derivatives and trigonometry. If you don't go into that depth, it may be hard for you to understand some of the concepts of variables being a substitute for a hard answer right.
David Yakobovitch:
I think you can still do all of it. It's just going to look different. Today we're going to be solving math on our ipads and we're going to have an ai assistant that verifies if you did it right or wrong or provide a different solution and give you that feedback the teacher or ta would provide much quicker. So, and give you that feedback that the teacher or TA would provide much quicker. So you'll have the middle of the pack, like most people who learn and only want to get to a certain level. But then the people who want to go for PhDs they can accelerate through the classroom and not be held back because they're like I'm super fast in math, so let me do. They can catch up faster and you can get teachers. I want to do 12th grade calculus in eighth grade.
Rajiv Parikh:
Yeah, sure you can do it now, why not? That was the original problem with the Khan Academy, right, yeah, to accelerate you through that when you were doing these math competitions in middle school. What's the journey look like from that? It sounds like you have a nice, supportive family. Yeah, was there a person in your life that said you know, I want to be like that person. I want to that person. This is a way that I get into this technology in Florida. So what took you there? Was there a professor, parent, mentor that inspired you?
David Yakobovitch:
Yeah, so growing up actually my dad, we're an immigrant family.
David Yakobovitch:
And he was very much into electronic and circuits and when he moved over from Israel to New Jersey he started doing TV repair, BCAR repair back in the 70s and when he made his way down to Florida he launched his own shops At one point had multiple locations, dozens of engineers fixing all kind of products. This is the analog days, of course, and for me, in elementary, middle school and high school I would go to my dad's shop and I would literally take apart circuit boards and remove capacitors and fix tubes for tvs and vcrs and audio kits and this is how I got very hands on into the precursor to robotics and really my dad inspired me so much about this I'd look at you already comfortable.
Rajiv Parikh:
Back then your father was doing it. You go to the shop. You weren't like the one that in in some random house, taking apart the tv and and get pissing off your parents.
David Yakobovitch:
You're actually the parent that was like oh no, come take this apart with me yeah, my dad would say, hey, like I'm obviously fixing these ones to put, you know, uh, bread and butter on the table, but oh, this one, it's okay if we don't fix it for a few weeks, but if you do this, david, and then you fix it, I'll also give you $10 or $20 if you fix it, father son playing with these toys and getting paid for it and getting paid for it. What better middle school incentivizing?
Rajiv Parikh:
thing. That's amazing. I love this. And then later on, as you went through you also, you know you were first in education, right, and then you went into investment. So how'd you make that move? Yeah, so what sparked you for that move?
David Yakobovitch:
so so, doing all this math in middle school and high school, it led me in college. Actually, I originally went for math and physics. That's why I was going to college you did what you were good at.
David Yakobovitch:
I did what I was good at, but I quickly said don't do theory or applied. And that's what led me into information systems, statistics, um data analysis, and in college I caught the bug for startups. That's when startup weekend became a thing, yeah, and so I went to every hackathon I could go and very quickly I realized I think I'm kind of good at these hackathons and I would, you know, win awards for, you know, best spoken or best product. Working with big teams at the university. I enjoyed it.
Rajiv Parikh:
I was even at cryptography competitions, so the competition was not just you doing your building something, but then you had to explain it. That's why you got good at actually explaining it to folks.
David Yakobovitch:
That's right. I think that was the precursor to education and doing a lot of this tech enablement.
David Yakobovitch:
And post-college, you know. Then I started doing these jobs, both at large companies and then startups, and the common thread is I was always the tech enablement guy. You know, I'd work with the data science teams, the software engineers, the sales teams and help them learn SQL, help them learn Python, help them learn Excel. Naturally, I became decent at it and then went to two of the large boot camps in the space General Assembly which got bought by Adeko, that's right.
David Yakobovitch:
And galvanized by Stride, and so I was running these enterprise teams for them, these motions with large hedge funds, quant shops, telcos, teaching non-technical people how to prompt, but to prompt with sql or to prompt with python, of course. Now we're in the days of prompting with large language models right.
Rajiv Parikh:
Yeah, well, how I got through college was teaching my roommates how to do the electrical engineering work. So just because because I was doing it, I got much better at it. Of course I did pretty well, I got, I got to be. It helped me do better on tests and probably kept me away from a lot of trouble. So especially since I was a frat guy. So now you're talking about LLMs, you're talking about data, you're talking about technology. Is there something either from before or after, that we normally wouldn't know about you? That's like a favorite invention or technology.
David Yakobovitch:
Okay. So I think everyone is both founders you know you're like LLM so I think, as founders, operators and investors, we all have what's called an anti-portfolio that startup you wish you got hired for, the one you wish you launched or the one you wish you funded and probably it was around 2012, 2013,. I, in my hackathon phase, went to Miami and formed a team with some incredible engineers and we built this product called Gorilla Cab. This is just when Zimride was launching. We basically built Lyft in an over-the-weekend hackathon and we had a moment there. We won first place for this at&t hackathon. We said should we launch this as a startup? Uber didn't even exist, lyft didn't really exist and we're like we think we could do it, but the team really didn't come together. Of course, that was a missed opportunity. Lyft and uber became multi-billion dollar startups.
Rajiv Parikh:
You were already in this multi-sided network I've always been in a startup.
David Yakobovitch:
one One of our hackathons was creating earbuds that you could press pause and play and change songs. This was years before the Apple AirPod Pro came out, so we've always been really early. I've always been this bleeding adopter in technology, both the hardware side and the software.
Rajiv Parikh:
And then over time, as you invest in it, pick at the right moment you can actually get, get it somewhere. So is there like a personal moonshot that you'd have?
David Yakobovitch:
personal moonshot. I think the biggest moonshot I have is investing in a thousand data first principal companies. I truly think we're at the beginning stages of unlocking the world with data. If we think, in the next 30 years, as we get to 2050 plus, there's going to be millions of satellites in space, we're going to be having stations on the moon, we're going to be having this uh, yr and li-fi and all these data sensors moving everywhere. We're going to probably have dozens of these devices we each own with our endpoints. There's going to be thousands of startups that need to build the infrastructure and build the tooling around that, and I want to support founders who are very ambitious about building quick products, constantly pivoting, understanding the market and truly building for the next data economy. That's awesome that's awesome.
David Yakobovitch:
So your life and your moonshot and it all comes together really nice hmm, yeah, I've really enjoyed looking back, right, I'm like my early days of math, into education, into building products and now investing, and I'm excited at the crossroads I'm at. You know, a lot of my experience has been as a early stage investor right, these pre-seed data AI companies, but now spending my time between data power and legendary ventures. Legendary has its own early stage practice and its growth practice and so, and are you teaching them cult?
Rajiv Parikh:
are you teaching these founders about building culture too?
David Yakobovitch:
I think culture is uh essential. I've uh so, having backed so many founders. I've seen it all in my short existence working in bigger companies. Yeah, I've seen educating founders break up, you know turnarounds, the solutions, you. These things are normal and you need to have VCs who are there to support you through both the good times and the challenging times. That's awesome.
Rajiv Parikh:
So now say you're in a parallel universe, right, and you weren't so into math and science and data. What would you?
David Yakobovitch:
be doing Gosh. Well, I do love community, you know, in New York I do put on events for VCs and founders. Well, I do love community. In New York, I do put on events for VCs and founders. And also it's so important to me because I think grassroots is how you do everything in life. You have to start with your customer, you have to listen to them, and so for me, when I've done anything in community, I always ask people what's your feedback, what do you want to see? And that's how I build it Because, honestly, my biggest life motto is don't hold strong opinions about loosely held ideas.
David Yakobovitch:
My opinion does not matter about almost anything in life. So I'm really trying to get that feedback, get all these data signals from the community. But if I could do it over again, I truly wish I would have gone all in on robotics and hardware and building software there. I love it. Today I still dabble in code and things when I can, but there's only so much time and it's always trade-offs in prioritizing what I build. So that's an awesome sense that I get from chatting with you that you have that ability.
Rajiv Parikh:
You're at this conference and you, like I talked about, we're meeting with entrepreneurs, technology executives. You've met with folks who are board members of public companies executives of public companies Any thoughts about that kind of community that you've been involved with over the last day?
David Yakobovitch:
Wow. Well, first, the caliber of talent is incredible. I was telling close friends before coming to the conference. I said you know why me Like. This is just an incredible group and selection of people. So, first, like honored to be part of such a great group. Also, everyone is just so open with their ideas, hungry to learn new things, hungry to share best practices in the industry, and I think that's what community is about. When you're masterminding, when you're bringing everyone together, it's collaborating. And who knows what's going to come out of this week, right, Maybe new startups, new ventures, new collaborations. For me that's fun. I tell everyone, let's make sure we have fun, let's build community, Get to know each other personally on relationships first, and then I'm sure we'll do some business things over time as well.
Rajiv Parikh:
I I really hope that happens. That's my my intent on it, just like you, that one of the things I love about you is you have this real deep affection for people who create, create community and who build each other up and drive success, and so that's the thing I care about the most is if we will all make money.
Rajiv Parikh:
Money's out there for people to make, but it's really about the connections. You have amazing people, so I'm so glad that you're here with us to do this and glad that we can also take the time and do this podcast. So thank you so much, david, for for joining us thank you for thanks for having me.