The Web3 + AI Daily: Science of Blockchain Conference 2026
The most interesting Web3 + AI talks from this year's Science of Blockchain Conference.
Hello, everyone, and welcome to all newly joined subscribers!
This is the 77th edition of The Web3 + AI Daily - your definitive guide to the intersection of blockchain and AI. Today, I’m sharing key takeaways from the Science of Blockchain Conference whose ninth edition recently concluded at Stanford University / Stanford University Department of Computer Science.
The SBC isn’t your typical flashy crypto event with yoga classes and spa treatments. Rather, it’s one of the most technically rigorous and research-focused events in the blockchain ecosystem. The program was curated by Dan Boneh, Tim Roughgarden, Dawn Song, and Ari Juels, with Ethereum, Protocol Labs, Interchain Foundation, DFINITY Foundation, Polychain Capital, YZi Labs, and BNB Chain joining as sponsors.
The research featured this year spanned consensus mechanisms, censorship resistance, DeFi, blockchains' response to the quantum threat, attack vectors, and zero-knowledge proofs. I selected a few talks that I think every Web3 + AI enthusiast will find particularly intriguing.
Thank you for being here! Let’s dive in.
New Consensus Paradigms
Proof-of-Useful-Work: Can AI Be Used for PoW?
Rafael Pass took the stage to discuss Proof-of-Useful-Work used by Pearl Research Labs, and whether the "useless" computations featured in the Nakamoto Consensus as a security guarantee for Bitcoin can be replaced by actual AI operations. I told you about Pearl Research in my recent article covering Web3 companies commoditizing AI inference:
Let me give you some context: Proof-of-Work (PoW) is a Sybil resistance mechanism used in Bitcoin and other blockchain networks, involving miners solving complex cryptographic puzzles in a competition to win the right to mine the next block and get the block reward. The goal is to ensure that every participant has real “skin in the game” and cannot simply create countless identities to improve their odds of winning. However, solving these puzzles require significant computational resources, while the solutions themselves have no value beyond securing the network.
This raises the obvious question: instead of letting all that computation performed by PoW go to waste, could we design puzzles whose solutions produce something of real-world value? The answer is 'Yes,' since a recent breakthrough made it possible to use matrix multiplications, the exact operation used in LLMs workloads, as a useful proof of work.
Moreover, as Pass demonstrates in his lecture linked below, the Proof-of-Useful-Work (PoUW) mechanism implemented by Pearl Research Labs does much more than that:
Pearl Research Labs uses its token inflation to subsidize AI workloads, much like Bitcoin uses BTC inflation to pay for the wasted computation. The result is lower-cost AI operations, which in turn stimulates AI demand and, ultimately, expands the market for AI.
“In that sense, PoUW becomes a finance vehicle for AI workloads. As a result, AI inference providers are incentivized to promote token adoption and better technologies. The better technologies and the higher token adoption you have, the more discount you get on AI and the bigger the market becomes.”
So, here’s another example of how the synergy between blockchain and AI can create meaningful social, economic, and financial benefits:
“In conclusion, adding AI workloads to a blockchain with a well-designed PoUW leads to more socially valuable computation.”
Space Fabric: A Satellite-Enhanced Trusted Execution Architecture
Filip Rezabek took the floor to present SpaceComputer | Orbital Root of Trust and its Space Fabric framework for a satellite-enhanced trusted execution architecture.
I must say, the whole idea behind data centers and AI computations moving to space is a bit out there for me, but if you’re a fan, you’ll find this talk intriguing.
In general, a trusted execution environment (TEE) is a secure, hardware isolated area inside a processor that protects code and data from the rest of the system. In the context of AI, they’re used as a mean to deliver data confidentiality and security. Rezabek’s lecture showcases how a satellite can be used as a TEE.
As it turns out, performing computation in space still relies on trust and lacks verifiability, as it faces handoffs between operators, platforms, and ground segments. On the other hand, once a satellite is up in space, physically accessing it becomes very difficult, which makes it attractive from a security standpoint. There’s also the aspect of its jurisdictional neutrality.
However, TEEs come with their own shortcomings, including the possibility of physical attacks and significant trust assumptions involving both hardware manufacturers and software suppliers. Space Fabric aims to address these challenges.
Listen to the full recording to learn more.
Zero-Knowledge Proofs
Self Protocol: Lessons from Deploying zkPassport at Scale
The regular readers of my newsletter know that the question of how we can verify and prove who’s human on the internet is one I consider quite urgent. That’s why I enjoyed the talk of Marek Olszewski presenting Self Labs and their zero-knoweledge-powered identity solutions.
Self Labs uses government-issued identity documents as the basis of its zkPassport:
Leverages the most secure attestations of who you are: biometric passports and IDs.
Why? Because they’re difficult to obtain, unique to individuals per issuing country, and machine readable. There are billions of people all around the globe who are currently in possession of biometric IDs.
The way Self Labs turns biometric IDs and passports into a zkPassport is the following:
Your mobile app scans your ID document;
The mobile app sends the scan to a TEE (because generating a zero-knowledge proof (ZKP) on your mobile phone would take too long;
The TEE generates a ZKP;
The TEE sends the ZKP back to you, and deletes your data.
This process guarantees that neither Self Labs, nor any third part has access to your credentials. It also allows you as a user to prove a number of circumstances, without revealing your name, date of birth, or ID number. You can easily certify that:
You’re over a specific age;
You’re from or not from a specific country or a set of countries;
You’re not on the OFAC sanctions list.
As Olszewski recounts, one enormous challenge Self Labs had to overcome was supporting every single biometric ID format in existence, because governments around the world use different hashing functions and digital signature schemes in the ID documents they issue. Check out the recording to learn more.
Thank you for reading! I hope you found this article insightful.
If you’d like to keep up with the latest developments at the burgeoning intersection of blockchain and AI, I invite you to subscribe.
And if you find value in my writing, consider supporting the newsletter. I’d greatly appreciate it.
The Web3 + AI Book Club is live on Fable. We’re currently reading Karen Hao’s ‘Empire of AI.’ Follow the link below and come join our literary community.
I'm looking forward to connecting with fellow Crypto x AI enthusiasts, so don't hesitate to share your thoughts here or reach out on LinkedIn, Instagram, or Firefly.social.
Disclaimer: The content of this newsletter is provided for informational and educational purposes only. Nothing contained herein should be construed as financial advice or as a recommendation to buy, sell, or hold any of the companies or assets mentioned.
Please don’t take my views at face value. Instead, do your own research (DYOR), think critically, and share your perspective so we can challenge ideas, learn from one another, and arrive at better conclusions through thoughtful discussion.