Alex Mirran leads business development at Lilypad, where they're on a mission to make powerful computing accessible to everyone. With a background in finance and computer science, Alex worked at companies like John Hancock before getting pulled into blockchain through his interest in Bitcoin.
His work at Lilypad is part of a bigger shift in how we think about computing infrastructure. Just as Airbnb transformed unused living spaces into a global network of rentals, Lilypad is creating a marketplace where idle computing power can be put to work. Their platform focuses specifically on AI and research applications, where the need for processing power is growing faster than traditional data centers can handle.
This episode of Contrarian covers everything from how everyday people might earn passive income with their hardware to why the speed of light matters in cloud computing, and what it takes to build infrastructure that could compete with big tech.
Below are five highlights from our discussion about the future of decentralized computing. If you find them interesting, you’ll want to check out the full episode right here:
Trust Through Verification
"When you have a model and you have training data and you have inputs and outputs, there is a big incentive and a big reason for that data to be verifiable... We're not doing training data, but there's others that are doing training data. We do verifiable inputs and outputs... And then of course, a verification that the compute happened."
When you're running critical AI models you need to be sure the computations are done right. That's why Lilypad is starting with biomedical researchers, who need both powerful computing and complete confidence in their results. Every computation comes with cryptographic proof that it was done correctly and the data wasn't tampered with. While they're focused on biomedical research now, this same approach to verified computing could be valuable across any industry where accuracy and trust are non-negotiable.
Blockchain Made Simple
“They don't want to set up a Metamask wallet and go get some tokens... they just need a service that works and they need it low cost. They need it to be available all the time.”
The Lilypad network makes it easy to deploy containerized processes (LLMs & more) to a global GPU network with blockchain verified compute jobs. With the upcoming SaaS product, complexities around AI model deployment, crypto payments, and more will be abstracted away with a user friendly tool.
The GPU Crisis and AI's Growing Appetite
"Even at Lilliput, we have a Google Cloud instance that we use as part of our development environment and it has a bunch of GPUs. And a lot of the times, I don't know, probably 60 to 70 % of the time we'll spin that thing up and it'll say, sorry, compute resources are not available."
The AI boom even has big companies with deep pockets struggling to get the computing power they need. Meanwhile, there are thousands of powerful GPUs sitting idle in gaming rigs around the world. Lilypad is building a marketplace to connect these unused resources with the researchers and developers who need them. This helps break down walls that are keeping smaller teams from competing in AI development. Even Stanford University, with all its resources, has more PhD students than GPUs available for research.
Your Hardware Can Make Money While You Sleep
“I'm seeing people take it upon themselves to become their own little cloud computing company... just like how people will buy up five Airbnbs and they become a mini hotel service, people are running a file coin storage miner. They're running a Lilypad GPU compute miner... It's their side hustle, and it turns into almost like a decentralized computing business."
Think about how much time your gaming PC or high-end laptop sits there doing nothing. Now imagine it making money during those down hours. That's what's happening with Decentralized Physical Infrastructure Networks (DePIN) - people are turning their spare computing power into passive income streams. Once you connect your hardware to the network, everything happens automatically - the platform finds the customers and directs the computing jobs to your machine. Unlike driving for Uber or managing an Airbnb, this is truly passive income - your hardware does all the work while you're off living your life. Smart contracts ensure everyone plays fair, with automatic payments and verification of work done.
Solving the Speed of Light Problem
"Every cloud company and really every internet user faces a physics phenomenon called the speed of light. You can't go faster really than speed of light... And so if data is being stored in San Francisco and you're calling that data in Indonesia, there's latency and there's a fundamental sort of limit to how fast that data can move."
Big tech companies solve this by keeping copies of data in data centers around the globe - if you're in Indonesia, you'll access a copy stored in Singapore rather than waiting for it to come from San Francisco. But what if instead of huge data centers, we could use the millions of computers and devices that are already out there? By tapping into devices that are already close to users, decentralized networks might end up with more points of presence than even the biggest cloud providers. This distributed approach could make AI and other compute-intensive applications faster and more accessible worldwide.
