Ethan Bloch was in junior high school in Baltimore during the dot-com boom.
For his bar mitzvah — the ceremony that welcomes 13-year-old Jewish boys into adulthood — Bloch received $7,000 in cash.
It was 1998 and, like so many amateur traders at the time, he plunged his wealth into the stock market, mostly software and telecommunication names like Lucent and Nortel.
He quickly tripled his money. By age 15, it was all gone.
"This knocked me over the head and left a burning curiosity that I still carry today," said Bloch, now 31, from the San Francisco headquarters of his financial-tech start-up Digit. "I realized I didn't know s--- about how any of this was working."
Now Bloch is playing into another trend that's taking over Silicon Valley: Machine learning.
Using the combination of massive data sets, exploding compute capacity in the cloud and a host of analytics tools, entrepreneurs are training computers to make increasingly sophisticated decisions. Eventually, the experts say, we'll land at true artificial intelligence, where computers are smart enough to train computers.
And it's going to upend the traditional banking industry.