Last week, the inaugural Tokenomics conference in Paris brought together a unique blend of computer scientists, economists, and mathematicians with a program focused on peer-reviewed cryptoeconomics research. The first of its kind, this conference sought to educate attendees and celebrate the impressive academic attention being given to this interdisciplinary field by a number of esteemed researchers. The program featured original presentations from over 30 international academic researchers.
Most notably, the conference was the first professional academic venue bringing together economists with computer science researchers and practitioners, two group that don’t frequently have the opportunity to interact with one another. While many of the presentations were highly technical, they were designed to be accessible to all attendees, no matter their background.
The two day conference was divided into 6 different research themes: Consensus, Smart Contracts and Applications, Initial Coin Offerings, Cryptography and Security, Cryptocurrencies, and Incentives. Each theme consisted of a panel discussion among relevant researchers. Presenters hailed from over 20 international universities including Columbia, Cornell, University of Chicago, Brown University, University of Toronto, NYU Business School, King’s College London, University of Sydney, Queen’s University, UC Berkeley, and more.
Keynote speeches were given by Amr El Abbadi of the University of California Santa Barbara, William Cong of University of Chicago, Dahlia Malkhi of VMware Research, and Catherine Casamatta of Toulouse School of Economics. All presented their original cryptoeconomics research.
The conference ended with an awards ceremony for the “Asseth — Kaiko Prize for Research in Cryptoeconomics,” to award the “best paper” as selected by the organizing committee with a €1,500 prize to encourage further research. The winning paper is titled ‘A Smart Contract Oracle for Approximating Real-world, Real Number Values’ written by William George and Clément Lesaege of Kleros Cooperative (link here).
Overall, Tokenomics 2019 proved a success and was a great first initiative in bringing together researchers from diverse fields in an emerging research sector. For more information on Tokenomics, you can visit their website at tokenomics2019.org.
Winning Paper Summary: A Smart Contract Oracle for Approximating Real-world, Real Number Values
In order to be useful, many smart contracts require access to information that does not natively live on the blockchain. For example, miners cannot determine who won an election, whether there was a flood in Paris, or what the price of ether is in US dollars. This information may be necessary to execute a smart contract for prediction markets, insurance, or financial contracts, but cannot be natively derived on-chain.
To respond to this challenge, the authors construct a general-purpose, decentralised oracle that can estimate the value of a real-world quantity that is in a dense totally ordered set.
Blockchain protocols generally don’t have access to off-chain information, which limits the existing potential of smart contracts. The need for oracles to feed smart contracts with off-chain information is an on-going discussion in the blockchain community. One problem involves the inevitable requirement of trust that must be given to whichever application powers the oracle. For example, an oracle that relies on a trusted third-party website is vulnerable to insider attacks where manipulated information deriving from the website could be used on-chain.
A contrasting approach is taken by decentralised oracles, which do not depend on any trusted third party with special privileges. Decentralised oracles involve setting up a type of economic game, whose incentive structure produces correct answers.
The authors outline the incentive structures of a number of existing oracle models, including Chainlink, Maker Dao, Truthcoin, Augur, Gnosis, and ASTRAEA. They then propose the existence of a decentralised, general purpose oracle that can decide between binary propositions. This means that it is able to produce true answers about the external world as a “yes” or “no.” Then, they propose extending such a binary oracle capable of producing an element in a dense ordered set. Their research builds upon existing oracle research to propose a decentralized oracle incentive structure that can estimate real number valued quantities.
Read the full paper here.