Math Olympian in Shadow of John Nash Tries to Solve Blockchain, AI Trust Dilemma

As a researcher delving into the cutting-edge field of technology, I am currently tackling a complex challenge that lies at the crossroads of two rapidly expanding yet intricate domains: blockchain and artificial intelligence.

As a analyst specialized in decentalized AI technologies, I’d rephrase the announcement from Hyperbolic as follows: “At Hyperbolic, the startup where I serve as leader for the past two years, we’re introducing a new solution to bolster trust within decentralized AI networks. We call it ‘Proof of Sampling (PoSP)’. Launching Thursday, this protocol aims to tackle the challenges that have long plagued these networks.”

I was established in the year 2022 with the partnership of Zhang and Yuchen Jin. He earned his doctorate degree in computer science from the University of Washington prior to our founding.

The new protocol’s concept was developed collaboratively with researchers from Berkeley and Columbia University, as stated by the team. This innovative approach merges mathematics, computer science, and economics. By implementing “sophisticated data gathering techniques and game theory,” Hyperbolic aims to encourage honesty and reduce computational costs in decentralized networks, according to their press release with CoinDesk.

In a conversation with CoinDesk, Zhang, aged 28, expressed his belief that Proof of Stake (PoS) represents the next advancement in validation processes for decentralized systems.

As an analyst, I would rephrase it this way: “Initially, it was believed that there was a sole method for carrying out verifications using consensus. However, over time, we uncovered alternative approaches such as optimistic proving and zero-knowledge proofs.”

He introduced PoSP as a versatile solution that can be implemented not just in AI but also in rollups, a kind of layer-2 blockchain technology, and actively validated services (AVSs) such as EigenLayer, which are secured by staking protocols.

An unreviewed research study penned by Zhang and his team about the Proof of Sampling Protocol was uploaded to arXiv, a Cornell University-hosted platform for distributing scientific papers prior to formal evaluation, on May 1st.

The design in the paper is based on the concept of a “Nash Equilibrium,” which is a well-known game theory concept named after John Nash. Nash was a mathematician who studied at Princeton University and became famous for his work, as depicted in the 2001 Oscar-winning film “A Beautiful Mind” directed by Ron Howard and starring Russell Crowe.

Here’s a figure from the paper illustrating the architecture:

In the rollout, Hyperbolic unveils “spML” as a new feature: an optimization of Probabilistic Spuriousness Model (PoSP) tailored for artificial intelligence validation.

In the press release, Zhang stated, “Using the fundamental concepts of PoSP, SpML develops a verification method that is not only quicker and more secure but also cost-effective.”

Now they just have to prove it works in practice.