Dear, Wall Street: Silicon Valley is increasingly coming for you.
A machine intelligence system, dubbed Emma AI, is starting a fund that hopes to outsmart the humans and computers that make a living trading stocks. It’s part of a wave of tech startups aiming advanced machine learning at financial markets.
Automation is not new to Wall Street. But Shaunak Khire, Emma’s creator, claims his system differs from current finance computing — high-frequency trading and “quant” data science — because its system of neural nets takes into account a more complex set of factors affecting stocks, like management changes or monetary policy in Europe, that other programs miss.
“This is not algorithmic trading,” he said. “This is literally replication of an analyst.”
Emma will start trading stocks from pharma giant GSK and Tesla along with U.S. Treasury bonds. Previously, Emma went head-to-head with financial news writers, testing its ability to replace the profession of yours truly.
It picked a handful of stocks six months ago and Khire said they’ve yielded more than 30-percent growth since. He’s keeping the name of his company under wraps while he awaits certain patents involving bots and AI, he said.
Other startups are applying AI to financial trades. Those include Aidyia, run by eccentric roboticist Ben Goertzel, and Sentient Technologies, which has raised formidable money and hired top engineers from IBM and Apple. IBM’s Watson began partnering with Citibank four years ago.
Khire notes that other startups haven’t disclosed their performance. And he is also extending an open challenge on AI performance to Watson, which he and several others in the machine learning world dismiss as underwhelming.
Some proceeds from the Emma system trades will go to Watsi, a healthcare non-profit that came out of the well-known Y Combinator startup incubator. “There could be some common areas to work on down the line eg: use of bots/ai for personalized, affordable healthcare,” he said of Watsi in a follow-up email.
The primary purpose of the trading now, Khire said, to test the strengths of AI on the selected stocks and bonds, although he stressed that also has commercial applications.
“How do you have an AI that — for lack of a better word —can ‘think’ on the input?” he asked.
The startup is in the process of closing a private funding round, Khire said.
This article originally appeared on Recode.net.