zkPI and The future of the internet

As of 9 am PT, Tuesday Mar 12, 2024, Team Blockus has shipped zkPIs for various applications, from web3 “wallet/accounts” to NFT mint, burn, buy and sells. We are excited to continue to push the frontier of this new technological breakthrough.

A breakthrough 4000 years in the making

‍

What is the backbone of the internet economy? Undersea fiber optic cables? Switches? TCP/IP? These infrastructures enable the digital flow, connecting servers, data centers, and devices, facilitating the rapid exchange of information. But what use would this rapidly exchanged information have without trust? Trust that it had not been tampered with, or spied on, that the sender and receiver are who they say they are. Maybe Nobel Prize winner Paul Krugman wouldn’t have made his notorious prediction “...Internet’s impact on the economy has been no greater than the fax machine’s.”, if he thought for a moment about the importance of trust in an economic context. Ask a person on the street what Public Key Infrastructure, SSL/TLS, HTTPS, AES, and SHA are and you’ll get blank looks, even though their laptop, phone, or tablet relies on the standards and associated algorithms thousands of times a day. Without cryptography, there is no Amazon, Google, Facebook, or Netflix.  

‍

While modern cryptography has already been tremendously impactful, its most mind-bending subdiscipline, cryptographic proof systems, which, among other miracles, can prove truth while protecting information, is only beginning to be applied. The idea of secret messages is intuitive, with Egyptian monks in 1900 BC using secret hieroglyphics. Cryptographic proof systems, however, are anything but. In a two person interaction, one can prove (or “argue”) to the other that he knows a private fact which makes a public condition true, or even just that such a private fact exists, even though he doesn’t know it. In some cases, it’s even possible to create a proof artifact that will, for the lifetime of the universe, convince anyone who comes across it, aliens even, that the writer’s claim is correct. If that’s not enough, these proofs have amazing compression abilities, enabling massive computations to be verifiable in only a few hundred kilobytes. Quoting one of the seminal papers in this field:

‍

“In this setup, a single reliable PC can monitor the operation of a herd of supercomputers working with possibly extremely powerful but unreliable software and untested hardware.”

‍

Back when Babai, Fortnow, Levin, Szegedy published their paper in 1991, it was mostly dismissed outside of academia because of the limits of the tech at the time - somehow a single PC had to be deemed reliable in an unspecified way. We now have the perfect construct - a blockchain. Zcash, anonymous DAO voting, zkRollups, zkSyncs and zkEVM are familiar to anyone with exposure to Web3, but we believe that this is just the beginning. Eventually, the ZKP revolution will allow Web3 to not only completely swallow Web2.0, but to transform society as a whole. 

Predicting the future is hard (if you want to be detailed about it)

‍

‍

Imagine facing a huge mountain shrouded in fog, such that the peak is barely visible, but the path is not. There's a righteous path up, but because of the fog, we must meander in order to find it.  The exact details of the future state of Web3 are impossible to predict, but we can look for clues of things to come. 

‍

Take the example of SUI’s zkLogin. It piggybacks off the existing centralized identity models provided by Web2.0 giants such as Google and Amazon to significantly improve the onboarding experience for SUI. We at Blockus are obsessed with the gamer’s UX. Most games don’t care about the infrastructure underlying their games. They just want to have the same experience they are used to. Forcing them to install wallet software, and write down a seed phrase in a safe spot is a good way to turn a high-intent user into a no-intent user. For this reason, Blockus integrated zkLogin for Sui as soon as it was available. Worldwide the numbers of gamers increased by a billion in the 8 years from 2015 to 2023, and we think methods such as zkLogin will allow at least that number to be onboarded to Web3 gaming in the next 4 years. 

‍

A skeptical reader might find this example unconvincing in a number of ways. First, it still relies on the Web2.0 giants, so how can Web3 supplant Web2.0? Second, it’s just a different login method, what’s the big deal? Addressing the first criticism, we note that Amazon started out as an online bookstore, which relied on companies such as FedEx and UPS to deliver its products. Now, it’s competing directly with those same companies through Amazon Shipping. With respect to the second question, we believe that zklogin will be one of the first mass adoption consumer facing zero-knowledge API, ZPI for short.

‍

ZPIs are revolutionary because they replace trust in the messenger with trust in the message.  When files are stored on Google cloud, there is implicit trust that Google will keep them intact and available. When files are stored on with Filecoin, Proof-of-Replication and Proof-of-Spacetime is used to achieve this trust. A data vendor with only a Web3 presence can prove that data was collected and processed correctly, at both the record level, and, using recursive zkSnarks, at a dataset level. More humorously, they can even leave the data publicly accessible, but with the caveat that those who want verify the data’s correctness must pay them for an interactive zero-knowledge proof session. Why might someone require such a proof? Simply because after the advent of generative AI, it will be dangerous to simply trust plausible looking datasets. 

‍

Speaking of AI, the myriad potential crossovers between AI, Web3 and ZPIs is too long to list. We’ll just leave it at the most obvious: verifying digital photos weren’t created by an AI. It’s also easy to see how the photo example extends to Internet-of-Things data and its processing. Certain types of oracles might even be replaced with ZPIs which aggregate signals from many IoT devices to achieve consensus of some external measurement. The IoT example is in turn suggestive of spatial computing gaming applications. Choosing what information to reveal and what to hide would obviously be of paramount importance in such applications, since they deal with physical location, and, as PokemonGO has shown, it matters from both the entertainment and safety angle. 

Some possible applications of zpi’s

We can think of a number of novel applications of ZPI which might appear in the gaming space in the short term. The first obvious application would be small puzzle games. The contract code would consist of a zkp verifier, while the gaming application itself would be hosted on client side. It’s obvious how something like a sudoku puzzle would be amenable to zkps, but sudoku itself isn’t interesting since simple hashing could also be used to verify solutions. More interesting would be puzzle games that are more open ended, in which the goal could be to get a blockchain recorded high score. In a similar vein, simple side scrollers intended for speed running, such as “Getting Over It with Bennett Foddy”, could potentially be cast in a zkp verifiable form. Although a Web3 “flappybird” might be a commercial success, we think the real impact of such games would that they are more amenable to younger individual developers, onboarding them directly into Web3 game development.  ID Software’s John Romero credits the success of that company to its founders having had many years of game development experience prior to starting it. 

‍

More ambitiously, some of the ZPIs that we expect to eventually find their way into the real world can start being experimented on in a Web3 MMPORG. What if a game had a sort of “Dark Forest” mechanic? Imagine something like EVE online, but with droplet weapons from the Three Body Problem.  Survival horror on a galactic scale. Scarce resources, long set up times for resource exploitation, lopsidedly overpowered offensive weapons. Space combat relegated to carefully executed decapitation strikes. And of course, no safety zones. Any sort of coordination would have to be done very carefully, revealing as little information as possible to counterparties, since you wouldn’t know if you could trust them. 

‍

For higher-end games, ZPIs can be useful in situations where the element of surprise is a useful mechanic. Consider a situation where a player might want to hide an in-game NFT they own for some period of time. This might be the case in games that involve a dueling mechanic. The developer might also want to make the item available across multiple games, without forcing the player to reveal the item. Furthermore, if the player, at any point, wishes to acquire the actual NFT by interacting with the blockchain, then a proof-of-purchase NFT would be a potential solution. A careful reader might think that the previous scenario could also be solved with standard cryptography, but we believe that the zero knowledge approach would enable a more uniform, maintainable, solution. 

‍

Keep in mind the photo of the mountain, and that it’s hard to predict exactly which application involving ZPIs will take off in the short term. However, Blockus will continue to lead this field and support new use cases as game studios unleash their creativity...

‍