Alice Liu: Jeremy, thank you for sitting down with me today. I’m excited to dive into your background and your work at RISC Zero. Let’s start with your journey - how did you first get involved in Web3 and specifically in the ZK space?
Jeremy Bruestle: Thanks, Alice. I’ve been an entrepreneur and technologist for most of my life. I actually started programming when I was around six years old, and I’ve always had a passion for mathematics. Throughout my career, I’ve focused on bridging emerging mathematical research with real-world applications.
I first started following ZK around seven or eight years ago. The concept of verifying computation with a single proof fascinated me, but at the time, the performance just wasn’t practical. The programming models were also incredibly difficult to work with. While working at Intel after my previous startup was acquired, I closely followed advancements in the field - especially @EliBenSasson's work on STARKs. That was the moment I saw ZK getting closer to being viable. I started experimenting with GPU implementations to understand where the performance could go.
From there, I began discussing these ideas with my co-founders - both of whom had worked with me in my previous startup. We had this pivotal realization: what if we could emulate a processor using a circuit, allowing us to run normal software inside a ZK environment? That became the foundation of RISC Zero. Once we realized that a performant zkVM was possible, we went all in, raised funding, and started building.
Alice: That’s incredible. You’ve had a diverse background - not just in crypto, but also in machine learning, compiler workflows, and bioinformatics. How has that influenced your approach to building in the ZK space?
Jeremy: A big part of my work across different industries has been recognizing when an emerging mathematical technique is ready for real-world application. That experience directly shaped how I approached ZK.
For example, when I first read Satoshi’s whitepaper, I was impressed by the theoretical breakthrough but skeptical about Bitcoin’s scalability. I initially underestimated how strong the adoption curve would be. Similarly, early ZK systems had enormous potential, but they lacked the usability and efficiency to be practical. My approach was to solve that gap-taking something mathematically promising and turning it into something developers could actually use.
Alice: Over the years, how have you seen the ZK space evolve? What challenges do you think have been overcome, and what do you see as the biggest hurdles ahead?
Jeremy: The biggest historical challenge was performance. Early ZK systems were slow, but we’ve seen tremendous improvements both in proof systems and hardware acceleration. Today, generating proofs is orders of magnitude more efficient, and we’ve reached the point where proving off-chain is often cheaper than executing on-chain.
Another major hurdle was usability. Writing polynomial constraints over finite fields isn’t intuitive for most developers. With zkVMs, we’ve made it possible to run conventional programs in a ZK environment without requiring cryptography expertise. That was a crucial step.
Looking ahead, I think one of the biggest challenges is simply awareness. Many developers still assume ZK is too difficult to work with or too expensive. The reality is that today’s zkVMs are not only practical, but in some cases, they’re already the most cost-effective solution. The challenge now is making sure developers realize that.
Alice: That’s a great point. When you think about the long-term trajectory of ZK, what does the “ZK endgame” look like to you? Do you see ZK remaining dominant compared to other cryptographic primitives like MPC or FHE?
Jeremy: I believe ZK will become a fundamental computing technology, just like public-key cryptography did. Long-term, it will be integrated into nearly everything. The ability to verify computation with mathematical certainty is too powerful to ignore.
As for other cryptographic primitives, FHE is promising, but it’s not yet practical at scale. It’s still in the research phase where ZK was years ago. MPC has its own set of trade-offs, and while it will have niche applications, I don’t see it replacing ZK in most cases. The efficiency and versatility of ZK make it the dominant choice for verifiable computation.
Alice: That makes sense. Shifting gears, let’s talk about RISC Zero’s zkVM. What was your original vision when you founded RISC Zero, and how has that evolved over time?
Jeremy: From the beginning, our mission was to make ZK practical. Initially, that meant tackling performance and usability. We built the zkVM so developers could run conventional software without needing to understand ZK cryptography.
Now, we’re focused on expanding ZK’s reach across all chains. That’s where our Boundless initiative comes in - a universal proving system that enables verifiable computation across different ecosystems. We want developers to be able to leverage ZK seamlessly, regardless of which blockchain they’re building on.
Alice: Boundless sounds like a game-changer for interoperability. Can you elaborate on how it enables cross-chain composability and what impact that could have on the fragmented blockchain ecosystem?
Jeremy: Absolutely. Boundless provides a global proving infrastructure that any chain can leverage. This means developers can verify computation off-chain and bring proofs back to their preferred blockchain, whether it’s Ethereum, Solana, or another network. This helps solve issues around liquidity fragmentation and interoperability because proofs can be trusted across different environments without requiring a shared consensus mechanism.
Alice: That’s exciting. Another major component of Boundless is decentralized proving. What are the advantages of decentralizing the proving process, and are there trade-offs compared to centralized provers?
Jeremy: Decentralization is critical for security and censorship resistance. A fully decentralized proving network ensures that no single entity has control over verification. This is particularly important in Web3, where trust minimization is the goal.
That said, there are some cases - such as private computations - where centralized proving may still make sense. However, even in those cases, we’re exploring privacy-preserving techniques like trusted execution environments (TEEs) and client-side proving. Over time, I expect most use cases to shift toward decentralized models.
Alice: Looking ahead, what’s next for RISC Zero? Any upcoming developments or announcements?
Jeremy: Our top priorities are launching Boundless and continuing to enhance our zkVM. We’re rolling out a major update with improved performance, larger memory support, and new cryptographic accelerators. We’re also onboarding developers into our Boundless testnet and will be expanding it soon.
Long-term, we’re focused on making ZK computing a fundamental part of the blockchain ecosystem - and beyond. We see applications far beyond Web3, from verifiable AI to secure cloud computing.
Alice: That’s a bold vision, and it’s exciting to see it coming to life. Thanks so much for sitting down with me today, Jeremy. Looking forward to seeing how RISC Zero evolves in the coming years.
Jeremy: Thanks, Alice. I appreciate the conversation and am excited for what’s ahead!
For the latest updates on Jeremy's work, follow @BruestleJeremy and @RiscZero on X.
