In a fireside chat at DePIN Day Brussels 2024, Tom Trowbridge, co-founder of Fluence, sits down with Kyle Samani, Managing Partner at Multicoin Capital. The conversation spans the early days of DePIN investing, the Helium origin story, core principles of risk, reward, and scale in physical infrastructure networks, and how DePIN could disrupt some of the world’s largest markets—from telecom to AI and energy.
Early Bets: Investing in DePIN Before It Had a Name
Kyle shares how Multicoin was drawn to Helium back in 2017–2018, well before “DePIN” became a recognized term. Helium’s founder, Amir Haleem, envisioned a global, decentralized IoT network. But in order to scale, Helium had to solve a fundamental go-to-market problem: how to cover entire cities, not just individual campuses.
The pivot to crypto was born out of this challenge. Helium used token incentives to crowdsource hotspot deployments, giving early participants the upside in return for early risk. This became a blueprint for many future DePIN projects. As Kyle puts it, “If you’re building a wireless network with 10 nodes, it’s worth zero. You need to reach threshold scale to have any value.”
The Core Insight: Reward the Risk-Takers
A recurring theme throughout Kyle’s thinking is the concept of risk and how DePIN uniquely rewards early adopters. If you’re the first one to buy and deploy hardware, you’re absorbing the most risk and should be compensated accordingly.
This, Kyle explains, is what makes DePIN different from traditional Web3: the act of deploying infrastructure—not just buying a token or joining a Discord—is a real-world commitment.
Defining DePIN: Beyond the Buzzword
Kyle cautions that not every “decentralized” project qualifies as DePIN. For him, the most relevant definition centers on bootstrapping real-world infrastructure via token incentives that align supply and demand. The value is not just in decentralization, but in reaching operational scale through community incentives.
The Case for Passive Infrastructure
Multicoin tends to back DePIN projects that require passive engagement. That means you set it up once, and it runs. Think Fluence (deploy and forget), HiveMapper (mount the camera), or Helium (plug in the hotspot). In contrast, projects requiring active participation—like ongoing human labor or drone operations—are harder to scale and less capital-efficient.
Demand Matters Too
Token incentives may bootstrap supply, but without demand, it all falls flat. Kyle emphasizes that evaluating a team’s go-to-market experience and domain expertise is essential. HiveMapper succeeded because its founder Ariel Seidman had deep roots in the mapping industry and could close enterprise deals.
The Tech Convergence Behind DePIN’s Rise
Kyle argues that DePIN wasn’t possible even five years ago. Helium launched its own blockchain in 2019 because Ethereum couldn’t meet their needs. Since then, the emergence of performant L1s like Solana, cheaper hardware, and increased market belief in DePIN as a category have paved the way. “Now builders have confidence, and users believe it’s real,” he says.
On AI, Compute, and What Comes Next
Kyle sees DePIN and AI as deeply interconnected. The explosion of compute demand means there will be a structural GPU shortage for years. This creates opportunities for decentralized compute networks, both GPU-based (like io.net, Render) and CPU/serverless (like Fluence).
He believes future AI models will evolve to work on consumer-grade GPUs, opening the door for Xboxes and iPhones to become contributors to DePIN networks.
DePIN x RWAs: A Surprising Crossover
One unexpected application? Whiskey. Multicoin-backed BAXUS is building a token-incentivized data network for price discovery in whiskey markets. Anyone who submits pricing data for a bottle transaction can earn tokens. “It’s Deepin-esque,” says Kyle.
The Long Game
DePIN’s largest opportunities lie in markets no one dared to disrupt: telecom and energy. With $100B+ annual spend, even a small slice of these sectors could make DePIN massive. “No one has funded a new telecom network since Bell Labs. This is the first time anyone’s even tried.”
