Block Talks by The Blockopedia: Crypto & Web3 Podcast

George Margis of 0G Labs argues AI agents require a blockchain trust layer for security and scalability

24 min Listen to the original Episode page
AI AgentsBlockchain InfrastructureDeFiHealth TechReal World AssetsVenture Capital

Summary

George Margis, leader of ecosystem partnerships at 0G Labs, identifies a lack of trust in AI agents' handling of capital and data as the primary barrier to adoption. He proposes that blockchain provides an immutable audit trail while Trusted Execution Environments (TEEs) verify model integrity. Margis highlights current use cases in DeFi trading and consumer health monitoring, predicting future growth in agent orchestration and real-world asset tokenization. He notes that 0G supports over 2.7 million live agents and offers significant infrastructure and funding support to builders.

Key takeaways

  1. 1
    0G Labs utilizes Trusted Execution Environments (TEEs) to provide hardware guarantees that LLM model weights and biases remain untampered during inference and fine-tuning.

    George Margis · 3:15

  2. 2
    The 0G community exceeds one million members, providing significant co-marketing and activation opportunities for projects launching on the infrastructure.

    George Margis · 10:38

  3. 3
    The 0G foundation manages an $80 million ecosystem fund and has raised over $300 million total, allowing for active investment in projects even during bear markets.

    George Margis · 11:18

Episode sections

  1. 0:00 – 2:03

    Intro and Guest Background

    George Margis introduces himself as the leader of ecosystem partnerships and investments at 0G Labs. He states that the primary barrier to AI agent adoption is a lack of trust in how agents handle capital and sensitive data.

  2. 2:03 – 4:02

    The Need for a Trust Layer

    Margis argues that blockchain provides a distributed ledger audit trail to prove on-chain actions. He explains that 0G uses Trusted Execution Environments (TEEs) to verify that LLM model weights and biases were not tampered with during inference or fine-tuning.

  3. 4:02 – 6:37

    Current Agent Categories and Use Cases

    The speaker identifies DeFi as the primary vertical, citing trading agents, portfolio monitoring, and transparent market-making solutions. He also highlights consumer applications like 'Best CEO' for marketing automation and health agents that monitor wearable data.

  4. 6:37 – 9:18

    Future Growth Areas

    Margis predicts growth in agent orchestration layers, health monitoring, dynamic gaming environments with AI NPCs, and the tokenization of Real World Assets (RWAs) such as equities.

  5. 9:18 – 12:21

    Builder Support and Infrastructure

    0G provides a decentralized compute network with over 150 LLMs, a private storage network, and free compute credits for early-stage teams. The ecosystem also offers co-marketing and access to an $80 million investment fund.

  6. 12:21 – 15:43

    Bridging the Gap with Centralized AI

    Margis notes that less than 4% of US households use AI services, citing UX friction as a major hurdle. He claims 0G simplifies this with fiat on-ramps, email sign-ups, and automatic model routing to reduce complexity.

  7. 15:43 – 19:30

    Active Models and Network Scale

    The most popular models on 0G are DeepSeek and Zero GM (a customized Qwen 3.6.6 variant). Margis states the network hosts over 2.7 million live agents and is horizontally scalable to support millions more.

  8. 19:30 – 24:08

    Founder Advice

    Margis advises founders to solve real problems rather than building for the sake of being founders. He emphasizes the importance of validating product-market fit, iterating based on user feedback, and having a clear utility for blockchain integration.

Claims worth verifying

George Margis leads ecosystem partnerships and investments at 0G Labs.

George ·0:17

How to verify: LinkedIn profile of George Margis or 0G Labs team page. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

0G Labs has raised more than $300 million.

George Margis ·11:38

How to verify: 0G Labs press releases or crypto funding databases like Crunchbase/CryptoRank. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

0G Labs has an ecosystem of over 350 projects.

George Margis ·4:02

How to verify: 0G Labs ecosystem page or press releases listing portfolio companies. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

0G chain has 11,000 TPS and super cheap gas fees.

George Margis ·4:22

How to verify: 0G Labs technical documentation or blockchain explorer metrics. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

0G community has over one million members.

George Margis ·10:38

How to verify: 0G Labs social media metrics or community announcements. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

Less than 4% of US households use AI services.

George Margis ·12:26

How to verify: Pew Research Center or similar demographic surveys on AI adoption. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

0G network has over 2.7 million live agents.

George Margis ·14:45

How to verify: 0G Labs dashboard or blockchain explorer statistics. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

0G Labs raised $35 million in a pre-seed round in March 2024.

George Margis ·17:45

How to verify: 0G Labs press releases or funding announcements from March 2024. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

0G Labs has three papers accepted at URIPS.

George Margis ·18:45

How to verify: URIPS conference proceedings or 0G Labs research publications. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

Michael Heinrich worked at Bridgewater, Bain, and Microsoft.

George Margis ·22:47

How to verify: Michael Heinrich's LinkedIn profile or biography. For recent events, verify 2026-08-15 through 2026-11-15. Use another period only when the speaker explicitly names it.

Primary source

Do your own research

  • Verify 0G Labs' total raised amount and fund size via their official website or recent press releases.
  • Check the live agent count and TPS claims on the 0G network explorer or official documentation.
  • Review the technical whitepaper for 0G's TEE implementation and how it differs from standard blockchain security.
  • Investigate the specific AI models hosted on pc.zerog.ai and their licensing terms.
  • Assess the regulatory landscape for AI agents managing financial capital in key jurisdictions.
  • Look for independent audits of the 0G network's security and performance metrics.
  • Compare 0G's gas fees and transaction speeds against competitors like Solana or Ethereum L2s.
  • Examine the terms of the $80 million ecosystem fund to understand eligibility and vesting schedules.
Full transcriptexpand
0:00
George Margis: Do not build something and then figure out what the problem is. If you ask anybody in a crowd how much capital would you trust your AI agent with, it's not gonna be a significant portion of their net worth. It's because they just don't trust these agents. I lead ecosystem partnerships and investments at Zero G or Zero Gravity Lab.
0:17
George: There are multiple AI agents working really fine without the need of blockchain.
0:23
George: So why we need the blockchain?
0:26
Unidentified speaker (UNKNOWN): So when
0:26
George Margis: AI agents are working with your capital or working with your sensitive data, how do you trust that that data is not being abused?
0:34
George: So how do you think uh we can bridge that gap?
0:37
George Margis: It is very difficult to use AI on-chain. Um, it's the the friction is there on the payment side, on the sign-up side.
0:45
George: The Uh what advice you want to give to a builder who is building right now on AI front, but not using the blockchain?
0:52
George Margis: A lot of uh founders just become founders just to be founders, and it's like that's not the right approach.
1:16
Unidentified speaker (UNKNOWN): So,
1:16
George: hi guys, my next guess is really exciting. Uh, his company has raised more than 300 million dollars, and the thing, uh the concept on which they are working is really, really cool for this time. They are building a blockchain for AI agents. So, let's try to know more from From George.
1:35
George Margis: Hi, thank you so much for having me. My name is George Margis. Marguis. I lead ecosystem partnerships and investments at Zero G or Zero Gravity Labs. And I'm happy to be here. Zero G, what we're building is the trust layer for AI agents. And we have this layer one blockchain, as you mentioned. It's EVM compatible. We also have a decentralized compute network and a
1:55
George Margis: decentralized storage network. And we vertically integrate that whole stack to power the next generation of AI agents on chain.
2:03
Unidentified speaker (UNKNOWN): So,
2:03
George: George, uh, my simple question would be there are multiple AI agents working really fine without the need of blockchain. So, why we need the blockchain? Like, what's the need?
2:15
George Margis: Yeah. So there's a few needs, right? So on the first aspect that we're focused on is the privacy and trust aspect. So when AI agents are working with your capital or working with your center. sensitive data, how do you trust that that data is not being abused? How do you trust that that capital is safe? Right. If you ask anybody in a crowd how much capital would you
2:35
George Margis: trust your AI agent with, it's not going to be a significant portion of their net worth, you know? And there's a reason for that. It's because they just don't trust these agents. And so we believe that fundamentally solving the trust layer for agents will enable us to actually use agents in a much more meaningful and significant way. And so what do we do with that, right? So, how do we
2:55
George Margis: solve the trust issue? Well, the blockchain is great for trust, right? What it does is it is a distributed ledger technology that actually proves what has happened on-chain. And there's an entire audit trail. So you can go back and say, hey, this transaction occurred, or this agent did this, X, Y, and Z. And we take it a step further. So we're not just a chain. We
3:15
George Margis: we have our own compute network. And so what that means is we're able to actually prove the model weights and the model biases in these LLMs and we're able to see whether or not any of that was tampered with. We can actually prevent this tampering of these models during the inferencing and fine-tuning process by leveraging trusted execution environments or
3:35
George Margis: TEEs. And so what that means is no matter what model you select, uh if it's running through the TEE, we know that there was no external manipulation. There is a hardware guarantee, and then there's attestations and proof that the model did what it was supposed to do and there was no manipulation.
3:53
George: So can you name some uh like uh two or three categories where uh your team is building or supporting the AI agents like for the first zero?
4:02
George Margis: Sure, yeah. I mean it's in across various verticals. I would say the the DeFi or decentralized finance industry is where most people are building the AI agents in our ecosystem. We have a large ecosystem over 350 projects building with us. Some are leveraging our storage, some are leveraging our compute, some are just leveraging the chain because it's super fast and it has 11,000
4:22
George Margis: TPS and you know super cheap gas fees and all of that. And so the the verticals are
4:27
George Margis: diverse. And I would say the DeFi space, we have trading AI agents that are using the compute network on 0G to power their intelligence. And so it provides insights and recommendations and monitoring of portfolios. Um we even have agents that are actually contributing to the market making uh uh uh
4:47
George Margis: solution. And so as we know, market making is a is a big issue in crypto where uh the interests may be misaligned from the market maker with the project or with the with the community token. And so if agents get control of that, there is that aspect of, hey, we can we can program the agents to act in the best behavior for
5:07
George Margis: the project's needs. And then it's transparent. All the actions, all the wallet balances, all the the movements are are are try tr are traceable.
5:15
George Margis: We even have uh agents that are that are helping people with their daily lives. So you can set up a team you know of multiple agents. You have we have this product called Best CEO. And what this product does is it it's like an all in one team platform for agents. And what it does is it it provides marketing insights and SEO geo optimizations. Uh
5:35
George Margis: It actually has agents that go out and crawl the websites and social socials and gives you insights on how to maximally drive marketing efforts. And then it actually executes on that. And so it'll actually, you know, change your website, it'll change the keywords and all these things and execute for you. So we're empowering builders, we're empowering the community to
5:56
George Margis: go the extra mile because we release products and infrastructure that actually empowers the agents, which become more useful in our network. It's not the same for other other chains or other ecosystems because they're not focused on, you know, the trust or the verification aspect of AI agents. That's what what Zero G is doing. And so people have more confidence using agents on our
6:16
George Margis: chain versus other chains.
6:17
George: So, George, uh as you are supporting new uh projects or AI agents, so what are the next uh categories or like uh you can name some vertical. verticals where uh you think the next growth like in next three months or six months will be towards the uh like for the AI agents, excluding the
6:37
George: payments part because that is already there. Yeah. So, like what are the fields you are seeing, like as you mentioned about the agent which works for marketing, the AEOGO. So it's a traditional work, it's a real work which uh people are doing. using Hermes or OpenClaw or Claude. So in these sort of scenarios, because they have
6:57
George: use cases, if we can do something uh using blockchain on that. So what's your as an ecosystem head? So what's your plan for that?
7:05
George Margis: Yeah, I mean, we're we're really excited to see the various use cases of AI agents. Um, I would say we're finding useful work being done by agents now. So in 2023, when when ChatGBT was properly released, it was just, you know, a glorified Google search engine, right? You ask a question and it gives you an answer. The
7:25
George Margis: past few years, what we've seen is uh actual intelligence in these systems. And these agents are able to actually execute on complex, multi-level multi level tasks. And so what we find is that there are teams in our ecosystem that are saying, hey,
7:39
George Margis: um, you know, let's create this agent orchestration layer and bring these agents and do useful things for people. You know, we have, for example, health agents that monitor all your
7:53
George Margis: lab results and your medical histories and your doctor visits. And then these agents, you know, give you daily reports, you know, because it monitors your Aura ring and your Apple Watch and all these things. right? So continuous monitoring of your health. Um, we have agents that are in the gaming space. And so gaming is not something that you you think of with agents, but um what they're doing there is
8:13
George Margis: they're they're creating a more dynamic environment where a single player can actually feel like they're in a multiplayer situation. Uh, and the NPCs or non-playable non playable characters actually have dynamic content. They're actually talking uh with you back and forth, and there's actually not a pre-written script. Uh there's also dynamic generation of maps in the game, you know, because of
8:32
George Margis: AI. Um we're seeing it in RWAs. So in tokenized equities, yeah, that's going to be coming onto the zero G chain. Uh we have uh a lot of interest from builders to pair AI agent tokens with these uh tokenized RWAs, these actual stocks. And so people want to speculate on the AI agint agent as well. And so the entire
8:52
George Margis: industry is just growing at such a fast rate. Uh Zero G's goal is to make sure we have the piping and the pipeline for for anybody to come in and actually actually build and and and and really take advantage of this um growth.
9:05
George: So uh as a ecosystem uh leader or head of ecosystem, what are the uh support activities which uh from your side a builder can uh get if he's building?
9:18
George Margis: Yeah, so I mean there's a lot of things we do. So I think the first thing is we provide the best infrastructure, the best in-class infrastructures at Zero G. Uh our compute network has an inbuilt router, it has inbuilt privacy features. Um, you know, we have basically a decentralized and private open router, right? It's called ZeroG Zero G Private Computer. Uh, it's accessible at PC.
9:38
George Margis: pc.zerog.ai. And there anybody can come in with one-line API code. They can come in and integrate and they can use the their favorite LLMs. We host over 150 LLMs on that platform, either hosted or integrated. And uh many of them are have privacy as well. And so we provide the compute network. And for teams that are in the early stage
9:58
George Margis: of funding, we provide free compute. You know, so we subsidize it, we give them compute credits and things like that. And if you ask any AI builder what's their number one cost these days, it's it's compute. So we we reduce that burden for them as much as we can. Um the second thing we do is we have a good storage network as part of the infrastructure. And so if people care about data sovereignty, if they care about encryption of
10:18
George Margis: their data, if they're dealing with sensitive data, healthcare data, you know, uh financial data that's you know high, high, high value and high risk, um they're very happy to use our. storage network because they know it's it's stored locally, it's not, it's not um, you know, it's encrypted, and then there's no chance that others can get access to it and steal it. Um, in addition to that, you know, we are a
10:38
George Margis: large community. The zero G community is huge, over a million plus uh you know, uh uh uh members, uh maybe more than that. I'm probably under underestimating. Um and so we provide a lot of co-marketing, we provide a lot of ecosystem activity. activations for the teams, right? Like if they launch a project using Zero G Infra, you know, we do a lot of marketing push for these guys and we say, hey,
10:58
George Margis: Zero G community here wants to experiment with some new cool applications. And then we take these applications and we introduce it to them. Uh in addition to that, I lead investments at Zero G. And so if the teams reach a certain level of idea validation, if they reach a certain level of traction, um, we invite them to present at our investment committee meetings. We have it
11:18
George Margis: quite often. And the foundation has a large ecosystem fund. There's $80 million in there. And we are actively investing in projects. So even in the bear market, Zero G was actively investing in projects and supporting these founders. And in addition to that, we've raised a lot of money, $300 plus million dollars. And so we have you know a large investor network in that.
11:38
George Margis: And we're able to make introductions for these teams to these investors. So it's infrastructure, it's marketing, it's investment. I think those are the three most important things for any founder.
11:48
George: So uh like there is there is a use case for traditional AI tools like chat GPT, Cloud and everything is there. And the on the other side sign we are uh providing blockchain as a test trust layer for AI agents. So there is a gap. There are like let's suppose ten people using Cloud and Chat GPT and rest
12:08
George: hardly one people is using. So how you think uh we can bridge that gap? Or like there was a Chat GPT moment for AI that happened and then everyone was using AI.
12:21
Unidentified speaker (UNKNOWN): Yeah.
12:21
George: So what could be the that moment for decentralized AI or like
12:26
George Margis: Yeah. So I I think, you know, the gap right now in usage. I mean, I think there's less than four percent of, you know, I'll speak in the US perspective, less than four percent of US households actually use AI services, which is crazy, you know. Like, I mean, if you see the the value that it can bring for you, um majority of America is not even In capitalizing on that,
12:46
George Margis: right? So well that in a way that's a good thing because there's massive opportunity for the builders. right? Um, but I would also say that it means it points out that there is a problem, there's a fundamental gap. So it's honestly very similar to the initial days of the crypto, uh, where people just couldn't understand the technology. They didn't understand uh,
13:06
George Margis: you know, the hash rate and mining and the creating your own seed phrase and wallet, private keys, all that. Um AI is kind of there. It's much easier. to use an LLM versus, you know, creating a private uh key. Uh, but I would still say that the gap is, especially when you talk about blockchain, it's the UI UX experience. It is very difficult to use
13:26
George Margis: AI on-chain. Um, it's the the friction is there on the payment side, on the sign up side, the friction is there on actually integrating your various AI harnesses and uh the actual storage networks and the compute networks. and even on the routing side, right? Like how do you choose the best model for your task? Um, you know, people
13:46
George Margis: use open router and then people use other services, but it's actually quite difficult. Um and then if you talk about inferencing, fine-tuning and training, doing that on the blockchain, you know, there is, you know, a couple extra steps that have to occur. And so we make all of that very easy. Um, you know, even in our in our PC, in the private computer, you know, we have fiat on our app, so you
14:06
George Margis: don't even have to touch a crypto wallet. Uh you can sign up with your email. Um and on the on the routing side, we automatically route your request through the best model for you. So if you are uh less price sensitive, we'll give you the high performance models. If you're more price sensitive, then you know it can route to the to the smaller models and uh less expensive ones. Um in addition to that, you know, our
14:26
George Margis: storage network is very easy to use. You can just drag and drop and it's pretty easy. Uh and then the chain itself is super fast and super cheap in the gas fees. And so if you are a crypto person and you need to have a large network of AI agents, for example, if you do it on other chains, it's actually expensive. It could clog up the RPCs, you could have drop transactions. But the way
14:45
George Margis: Zero G architected our chain is it's infinitely horizontally scalable. And so what that means is we can support a network of millions of agents. And in fact, today we have over 2.7 million agents on Zero G live. Trading agents, RWA, gaming, whatever. They're all there. 2.7 million plus. And it's just growing at a rapid rate. And
15:05
George Margis: our network has no issue, right? The chain is purposefully built for AI. It's a different from what you see in other ecosystems where they're pivoting to AI. or they're putting the AI tagline on top of them. From day one, Zero G has been focused on AI and and now we get to to benefit from that because now we get to actually onboard the ecosystem projects and they get
15:25
George Margis: to experience how we re-architected the entire infrastructure to support a large scale of AI agents.
15:33
George: And uh because there are these many AI agents active on Zero G. So uh just tell us about the models which you use and what like you are providing to the ecosystem.
15:43
George Margis: Yeah, I would say the most popular model right now is it's there is there's two, it's almost a tie. You can look at our stats and you'll see it live on a dashboard. It's between DeepSeek and Zero GM. So Zero GM is is uh a a model that's created by Zero G Labs. Uh we actually are a Frontier Lab at this point. So we took Alibaba Qwen three 3.6 point six. and we
16:03
George Margis: added a lot of features to it and made it much more efficient, much more uh, you know, the cost per token is much, much better. And so the Zero GM model is is like the number one most use because it it it's much, much cheaper than using Qwen. It's much cheaper than DeepSeek and all of that. I would say the third one after the those, it would be GLM. Uh people
16:23
George Margis: love GLM. And then there was a spike with with Fable. and then now Opus and things like that. But I think it's kind of cost prohibitive for the most most of the builders, you know. And so we have a whole ranking list. You can check it out at pc.zero G uh slash stats and you can see the the consumption. But honestly, it it it it rotates, you know. I uh at
16:43
George Margis: one point there was the DeepSeek moment, everyone's using DeepSeek, C, then there's the minimax and everyone's using minimax. It just rotates across. And so we're grateful that. we can just, you know, integrate all of them and let the community decide what they want. Yeah.
16:55
George: So uh this is all about work. So from the point of view of George, I just want to know like your journey with Zero Z Lab and in this Web3 ecosystem.
17:05
George Margis: Sure. So I'm I mean I've been in the crypto web three space since uh 2017. Um I I participated in all the the ICOs, initial coin offerings back in the day. Um Uh I created a venture fund with my brother and a few friends uh called Newfield Fund. And you know, we were investing in a bunch of projects uh um from 2019 onwards, uh, you know, and we're very happy with
17:25
George Margis: that. Uh and then, you know, in uh 2024, you know, I was actually working at Stanford and I met Michael Heinrich, our CEO, and his wife Ada. Uh, they were presenting at an AI and Web3 conference in Stanford, and uh I'm I I saw his presentation, I said, This is amazing. So I went up to him immediately. I said, Let me invest. Uh, and then he said, uh,
17:45
George Margis: maybe, because at that point, March 2024, Zero G had already raised $35 million in a pre-seed round. So he wasn't looking for any more capital, but I tried my best and convinced him and you know, I had a lunch with him and his wife, and it was an amazing time. And you know, they let me join the cap table. Shortly after that, I said, Look, um, you know, the way we work as a as a VC fund is we like to
18:05
George Margis: support our portfolio companies. Um, and I'm happy to join as an advisor for free, just just just support. And I loved so much the team, you know, Ada is is is is a is a machine. Um, and then and Michael is like a seasoned operator with, you know, f several unicorns behind him. Uh his nickname is actually Unicorn, which is crazy. Um, and so I was like,
18:25
George Margis: this is an amazing team. I need to get deep. deeper and deeper. So joined as an advisor for free, then joined part-time in the business development side, and then later joined as a partnership lead. And then now I'm running the investments arm. So I just got deeper and deeper into it. I love what they're doing. They're solving a fundamental problem in AI. And they have a very strong tech team led by Ming
18:45
George Margis: Woo, Wu, the CTO and co-founder. And so the the infrastructure is enterprise grade. It's top level. And these guys are pioneers in research. We have three papers that were just uh accepted in URIPS, right? It's one of the biggest conferences for AI. Um, we're releasing more and more products that uh and and research that the rest of the the ecosystems in
19:05
George Margis: the blockchain space are are you know using. So so we we push the envelope forward and then others take that technology and and use it their own their own way. And so I wanna want be to at the seat of innovation. And that's why I'm at Zero G.
19:17
George: Cool. So, George, my last question would be uh what advice do you want to give to a builder who is building right now on AI front, but not using the blockchain? So, what would be your advice to him?
19:30
George Margis: Yeah, so I mean the first thing is to take a step back and make sure you're building something that is solving an actual problem. You know, a lot of uh founders just come become founders just to be founders. And it's like that's not the right approach. Or perhaps they're trying to escape, you know, the monotony of a nine to five job. And I understand that, right? I I was a doctor and so I was I had a six to six job, not a nine
19:50
George Margis: to five, right? Six a.m. to six p.m. So I know that. I understand that we all want to escape that monotony monotony at that sometimes, but that's not enough, right? You have to build something that has real value, you have to build something that solves a real problem, a pain point for people. Once you identify that, that's gold. The second thing to do is to say, okay, I identified a problem. Am I uniquely capable of
20:10
George Margis: solving this problem? Okay. And if you are, do it, right? Go out, build it. You have AI agents now that can be your co-founder and CTO. So even if you're non-technical nowadays, days you can be a solopreneur, right? Uh, but if you're not capable of it, you have to ally yourself with other ecosystems or or builders or accelerators that
20:30
George Margis: that can help you get that level of expertise and domain expertise to build a great product. And then once you align yourself and you have this team, build the product and continue to iterate. So what happens is founders say, hey, the product is not perfect yet. We're going to delay the release, delay the release. I don't think that's the right approach. The approach should be go ahead, release the product. And if there's issues, keep iterating and
20:50
George Margis: then have customer feedback, detailed customer feedback. Always talk to your top 10 customers and then say, hey, what can we do to improve? Or we noticed you didn't use this aspect of our platform. Is it not useful or do is it too complicated? Whatever. The user feedback is super important. And so do that and keep iterating the product. And then the product will evolve over
21:10
George Margis: time. The last advice I would say is pay attention to the market. Um, If you're going to enter the blog, blockchain space, first thing is don't just add crypto or blockchain to it just for the sake of adding crypto blockchain. Uh it's very transparent for seasoned VCs and investors. Uh they they know exactly what you're doing when you do that. And you say, hey, uh, you
21:27
Unidentified speaker (UNKNOWN): know,
21:27
George Margis: we we're adding crypto for this aspect. And they're like, you really don't need crypto for this. You don't really need to tokenize your product, you know. Um, and so there should be a real token utility that you can communicate to. any investor. And you should assume the investor will will say that's not needed. And so you have to say, no, no, no, it is needed. So like for Zero G, there were investors, we raised $300 million.
21:47
George Margis: There were some investors, right, that's came and said, why do you need a Zero G token? And then we said we're a layer one blockchain. So the Zero G token is our gas token, right? At a minimum. Then there's validators to secure the network and they need to stake zero G tokens. And then we have a compute and storage network. So you have to pay with for that with Zero G tokens. So we had a bunch of reasons. We had to actually list it out for
22:07
George Margis: exchange listings and for investors. Same thing for founders. Be ready with that answer. And then the last thing is you know, if you want to raise capital, you know, make sure that you present your pitch deck to all your friends, all your family as many times as you can. The pitch is so important. If If you've been in the room with Michael Heinrich, our CEO, when he's pitching zero,
22:27
George Margis: G, it's a masterclass, right? It's there's no one who's better at fundraising in the entire blockchain space than my CEO. And um I I asked him, how did you get to this level of expertise? Well, I mean, he worked at at Bridgewater, he worked at Bain, he worked at Microsoft. So he has the credentials. He's studied at Stanford and Harvard and all of that. But that's one aspect.
22:47
George Margis: Even for a seasoned veteran like him, he's always iterating and practicing on his pitch. Always. He goes through it and he presents it to himself, presents it to his wife, presents it to people, and he's keep keeps on working on it. So a lot of founders come to the investment committee and they didn't perfect their pitch. And honestly, we just have a five
23:07
George Margis: minute window to learn about a project. If your pitch is not crystal clear, it you very likely you won't get the funding. You know, you have to. really understand how to present it in a very convincing manner, very clear thought process. And then you have to make it seem like if you don't invest, it's your opportunity m amiss, right?
23:27
George Margis: And so if you take all the advice I give you, I still think the most important advice for a founder is solve a problem that people actually care about. Do not build something and then figure out what the problem is. Find the problem first and then solve it.
23:42
George: Thanks, Lord George, for your time. And uh we have the next uh speaking slot at the stage. So we'll wind up this year. And we'll love to sit with you for a long session on our virtual podcast after this week. Sure. And then we can discuss more. That sounds
23:57
Unidentified speaker (UNKNOWN): very good. Thank you so much.
23:59
George Margis: Thank you. Appreciate it. Thank
24:00
Unidentified speaker (UNKNOWN): you.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights