Various activities that we carry out on a daily basis have been transformed because of artificial intelligence (AI).
In the financial world this this involves both banking and fraud prevention, but fund management and stock-picking are still a little reluctant to take the AI plunge so it seems.
While hedge funds will embrace AI to a point, they’d still prefer to keep it to a minimum. Sentient Technologies is an AI startup with a hedge-fund arm.
Co-founder of the company, Babak Hodjat says that if machine learning applications were left to their own devices they have a tendency to overfit, finding unusual patterns in the data that aren’t really relevant. So instead a more tactful way of deploying AI is needed.
READ MORE: 10 Applications of Machine Learning in Finance
READ MORE: The Big Problems with Machine Learning Algorithms in Finance
Sentient employs around two dozen AI experts and for that reason is able to continually research new techniques.
In the past three years it’s been expanding its AI applications to include both website optimization and online shopping services.
And just a few months ago the company launched a hedge fund to raise money from outside sources, whereas previously it only used its founders money.
Numerai is a slightly newer firm, founded in 2015, that also thrives because of its unusual business model.
Launching its first funding round this autumn, the firm uses AI to encrypt financial data.
It also runs competitions to entice others to join the machine learning craze.
These competitions are for anyone who can come up with the best machine-learning algorithm.
Already around 1,200 data scientists take part each week to win either cash prizes or cryptocurrencies.
The best algorithms are chosen to encrypt financial information back into financial data before it’s used to decide which shares to trade.
Cerebellum Capital is a hedge fund that’s been around since 2008.
In 2016, it began working on a fully AI-run equity fund, and earlier this year it was released.
This innovative fund uses machine learning to crunch data and compile strategies as well as gauging how effective these strategies will be.
The actual trading, however, is left to the humans.
So, while there’s some applications creeping its way into these sectors of the financial industry, there’s still an element of human interaction, and the reigns have not been fully released yet.
Source Economist
