Billionaire Cliff Asness downplayed the potential of machine learning and artificial intelligence to revolutionize investing.

“I’m a little bearish on the whole thing,” Asness said at the 2nd Annual Evidence-Based Investing conference in New York on Thursday. “Machine learning is basically a way to find more and more patterns. It can be used wisely.”

Asness said his firm is mainly using machine learning to improve its trade executions and figure out when to roll over futures contracts, instead of using it for core investing decisions.

The portfolio manager also shared his ranking of different quantitative investing strategies.

“Gotta love the market factor. My favorite [among] the others is still the value factor. It has the best story,” he said. “My least favorite is the small-cap factor. Evidence is weaker than the other ones. And the story, why should small beat large? It is a risk or illiquidity story.”

He added investing in momentum or stocks that are performing well is another one of his best quant strategies.

Asness is founder and chief investment officer of AQR Capital Management, a leading quantitative investing firm. AQR has $208 billion of assets under management, according to its website.