Hedge Fund Launches Currency for Open Source AI Network

Back in the eighteenth century, Adam Smith identified what he called the “invisible hand” of the free market. This was a
characterization of the naturally emerging equilibrium that transpired from the actions of a multitude of self-interested individuals.

Now a new virtual currency, the first from a hedge fund, is set to play a key role in perhaps conjuring a new invisible hand. This synergy is expected to emerge, as with Smith’s version, from a multitude of individuals – but this time from collaboration, not competition – driving value for both the individuals and the company. With marquee backers, this “open-source” hedge fund has attracted notice, and, if it succeeds, it could introduce a revolutionary new ethic into ultra-Darwinian Wall Street.

The company is San Francisco based Numerai, founded by South African technologist and Cornell maths major Richard Craib, who is CEO. The currency is Numeraire, a cryptographic token intended to incentivize the modelling of the Artificial Intelligence (AI) that directs the hedge fund’s trades. On June 21, the Numeraire smart contract was deployed to Ethereum, and more than 1.2 million tokens were sent to a distributed network of 19,000 data scientists around the world. Already trading successfully for over a year now, the company is refining its approach and hopes the Numeraire will bring them further along the path to establishing a fully decentralized, software-powered hedge fund.

According to Craib, the Numeraire can bring an open source ethic to finance, potentially delivering the kind of network effects already harnessed by companies like Uber and Facebook. “Finance is totally competitive right now,” he says. “There’s no reason why you’d want to help your friend with a hedge fund if you have a hedge fund. But if you both held the same cryptocurrency, and it went up in value together, maybe there would be more collaboration.”

The company has far more data scientists working on its investment strategy than the biggest name hedge funds, and everybody involved has the incentive to recruit even more talent. What makes this hugely concerted effort possible is the combination of bleeding edge technologies and practices, principally the confluence of data science, cryptography, artificial intelligence and crowdsourcing. Each week, Numerai sends encrypted stock market information to its partaking data scientists who then work to create algorithms to isolate patterns, staking Numeraire upon their models according to their confidence in them. So far, around 40 billion predictions have been received, with 100 million added every day.

“We want to make a system where our incentives are aligned with user incentives,” says vice president of engineering Geoffrey Bradway, “so if we do really well, the users do really well, and if the users do well, we do really well.”

The staked Numeraire is held in the smart contract while the model’s performance is assessed, the amount staked contributes an extra data point regarding user confidence in the prediction. If the predictions perform well, the user earns ether, and his Numeraire is returned. If unsuccessful, the user loses their Numeraire which is destroyed. Each week, all who participate receive a portion of new Numeraire minted that week. Craib comments, “Over time, the Numeraire will perfectly incentivize the users to do the one thing we want them to do, which is to make the fund go up.” While Numeraire has no stated value in itself, it is predicted to trade on secondary markets. Likely buyers will be successful data scientists who wish to stake more in search of bigger ether rewards. However, even those simply holding Numeraire, without making bets, will see its value grow if the fund succeeds and secondary demand increases.

Yet more conservative voices suggest Craib’s vision is the result of overthinking. “Their model seems overly complicated. It’s not clear why they need it,” says Michael Wellman, a University of Michigan professor who specializes in game theory and new financial services. “It’s not like digital currency has magical properties.” Wellman suggests sticking to the old method of simply recruiting the most talented data scientists.

Visions are always a gamble, of course, and imitative replication a surer bet. If Numerai can truly harness the power of the network, we might see a rising tide of collaboration which will, as the proverb goes, lift all boats. Then, collaboration may outcompete competition itself.