In this episode of the EUVC podcast, Andreas discusses, with Richard Muirhead, Managing Partner at Fabric Ventures, the transformative potential of Web3 and AI.
Richard shares insights into how these technologies will reshape the internet, the power dynamics between big tech and decentralized communities, and Fabric Ventures' role in this evolving landscape. We also discuss ethical considerations, the development of decentralized finance, and the challenges and opportunities facing venture capital in Europe.
Fabric Ventures is a 125M EUR fund, with 250M under management, anchored in Europe around Paris & Zurich, but investing globally. They are actively investing in WEB3, which has become a global phenomenon, and they have incredible investments in their portfolio so far, such as Sorare, Ramp, Near, Immutable, Polkadot, Flashbots, and many more.
Watch it here or add it to your episodes on Apple or Spotify 🎧
This episode is brought to you in partnership with Zero One Hundred Conferences, which organizes LP-GP networking events for PE & VC players in various European regions with a global outreach. In the last 8 years, they have hosted 50 events with 1700+ speakers, 6500+ investors, and 12000+ attendees.
The upcoming 0100 Conference Mediterranean presents a unique opportunity for private equity and venture capital LPs and GPs to meet and develop meaningful relationships.
Attendees will include major industry firms such as the European Investment Fund (EIF), Tikehau Capital, Vencap, Arcano Partners, Amundi-Alpha Associates, P101, United Ventures, Merseyside Pension Fund, and many more.
22:32 Challenges and Opportunities in Web3 Adoption
26:59 Disrupting Centralized Powers with Web3
29:20 The Future of AI and Web3 Integration
33:43 AI Safety and Ethical Considerations
35:28 The Case for Open Source AI
37:19 Challenges of Open AI Systems
37:59 Business Models in Open AI
40:48 Token Economics and Incentives
42:35 Decentralized Systems and Network Effects
50:35 Operational Challenges in Venture Capital
53:57 Building a Successful VC Firm
57:34 Advice for Emerging Managers
01:05:04 Concluding Thoughts on Venture Capital
✍️ Show notes
Is Web3 x AI ‘a thing’, and does it matter?
In the near future, AI will service our every need safely and with integrity…or will it? Exactly how these systems are built will determine the answer.
1/ There are FIVE key inputs:
Compute
Talent
Data
Models
Governance
For decades, AI has flourished because of the open research, development & contributions from academia, governments, and companies around these inputs. Let’s double-click on a couple of examples.
2/ Data
ImageNet was a clear catalyst for the deep learning boom; Current state-of-the-art models are mostly trained by scraping the open web. The dataset below was used to train GPT-3 (@jimmy_wales)
3/ Models
Building in the open has resulted in extraordinary innovation from the global community, from AlexNet (UToronto, ‘13,), alphaGo (@GoogleDeepMind), Deep mind, 2016,), Transformers paper (@Google, 2017), to GPT-1 (@OpenAI, 2018), and more recently, the Llama models (@Meta, 2024)
4/ Talent
A remote-first ecosystems - fascinating innovations such as running Llama on an iPhone using MLX, or an independent researcher writing 84 new matrix multiplication kernels for llama (https://justine.lol/matmul), making it magnitudes faster.
5/ A victim of their own
But now, a victim of their own success, exponentially rising compute costs have forced open startups to sell equity to the massive web2 compute providers. With new owners, they have turned away from their open origins.
6/ Governance
They argue, of course, that the move from open to closed is necessary - only they can do safety properly, and others cannot be trusted with open models - “Open Models are harder to control and put guardrails on… I’m not saying we should ban it outright but we should look carefully at their implications as they grow.”
It conveniently ties into the tech titans’ business model, controlling the means of production and selling access to the outputs. Yet the pitfalls of Web2 are already well understood.
Corruptible governance
Deliberate and Accidental Bias
Lack of Transparency & Privacy
Profit over utility
Single points of failure Hence: “don’t be evil” > “can’t be evil”.
Not only is this better, it’s necessary. Today AI looks like simple chatbots, but in the future it will need to cater to increasingly complex requests ranging from geopolitical national security matters to custom models that adapt to specialised, localised computing environments.
Call to BUILD a better path. One that maintains the open, permissionless & productive innovation AND also delivers compelling & sustained economics for participants.
A ‘Can’t EVER be Evil’ architecture.
3 biggest learnings in venture
You shouldn’t look for deals or do any analysis. They should come to you and they will either feel right or not and any more time wasted is time not spent seeking the ‘next’.
Few people outside of operators have earnt the right to advise founders.
It’s about what could go right / not wrong. It’s about committing not omitting. It’s about ‘just the right side of crazy.’
Advice for young people in the industry
Go join a Series A funded company and work in product management, or even better raise seed funding for your own.
Just start ‘cutting cheques’ …however small…into projects that excite you.
Interview the founding teams and their advisors and publish this as X storms, podcasts, mirror posts or whatever.
Tips & tricks for emerging VCs
Find your team (hint: you probably already know and trust them)
Pick an ‘edge’, now more than ever you need to be distinct, and align every aspect of the firm with it.
Court your anchor. This is more about matchmaking than money-making if you have picked your edge well.
Most counterintuitive learning
For a long-term business, short term results weigh disproportionately. Especially in Europe.
Results count, but brand matters.
Top venture investors must embrace significantly more risk than the average founder…each time they place a ‘bet’.
Consensus is a cancer. Conviction and dissent is where the returns lie.
📅 Upcoming virtual events
From time to time, a podcast is just not enough. Check out our roundtables and live events below.
Gender diversity in VC isn’t just a buzzword—it's a power move that drives innovation and profitability. 🤘 Balancing the scales in the investment world means more creative solutions and stronger economic growth. Ignoring this is so 2010s. 🕰️
AI is flipping industries on their heads, from health to farming. Get the scoop on what’s real and what’s just sci-fi. With a vertically specialized micro VC, a Robotics CVC with major acquisitions behind him and a generalist who just can’t stay away from the space, you’ll be in the know after listening to this conversation 💸
This roundtable is a must for GPs looking to get an edge using AI. By tapping into AI-driven analytics and predictive modeling, VCs can unlock hidden trends and streamline their operations. Plus, tackling early risk assessment can save portfolios from tanking 💥. It's about staying ahead in a cutthroat world.
🗓️ The VC Conferences You Can’t Miss
There are some events that just have to be on the calendar. Here’s our list, hit us up if you’re going, we’d love to meet!
On decentralizing the future - Web3 x AI and why does it matter, with Richard Muirhead, Fabric Ventures