Tristan Greene* says DeepMind’s becoming a talent-churning startup factory.
Three of the developers responsible for DeepStack, the first AI system to beat humans at heads-up, no-limit poker, have left their posts at DeepMind to form a new startup dedicated to dominating the stock market.
The new venture, called Equilibre Technologies, will employ algorithms to pick stocks and cryptocurrency.
Up front: The stock market is a nebulous and challenging environment to try and “solve” with AI, but the team’s leadership appears more than up for the challenge.
Martin Schmid, Rudolf Kadlec, and Matej Moravcik, the company’s founders, all worked together at IBM and DeepMind.
There’s something to be said for keeping a successful team together, but the existence of this startup does raise some questions about whether DeepMind is bleeding talent, or simply elevating it to the next stage.
There isn’t a lot of information to go on concerning Equilibre Technologies at this time, but we can be pretty sure they’ll stick close to the vest when it comes to developing algorithmic trading technology.
Background: While they’re certainly not the first to attempt to demystify the stock market with artificial intelligence, they’re taking a slightly different approach than their competitors.
The typical stock-picking AI tries to “guess” what’s going to happen next in the market based on historical trends.
But the Equilibre team’s known for creating incredibly complex algorithms to succeed at solving problems without a lot of information.
They combine game theory with artificial intuition to gain a theoretical advantage over other computer-based or human trading methodologies.
According to the DeepStack research paper, the challenge of winning at poker isn’t all that different from successfully playing the market:
Poker is the quintessential game of imperfect information, and a longstanding challenge problem in artificial intelligence.
We introduce DeepStack, an algorithm for imperfect information settings.
It combines recursive reasoning to handle information asymmetry, decomposition to focus computation on the relevant decision, and a form of intuition that is automatically learned from self-play using deep learning.
In a study involving 44,000 hands of poker, DeepStack defeated with statistical significance professional poker players in heads-up no-limit Texas hold’em.
Quick take: Nothing in tech is guaranteed, but this feels like a license to print money.
The founders’ history and connections give a solid bedrock to build out from.
And, per this report from CNBC, there’s no shortage of VCs chomping at the bit to throw money at Equilibre Technologies.
As the old saying goes: teach an AI to win at poker and you’ll eat for a day, teach an AI to dominate the stock market and you’ll be set for life.
*Tristan Greene covers human-centric artificial intelligence advances, politics, queer stuff, cannabis, and gaming for The Next Web.
This article first appeared at thenextweb.com.
