Gigaom AI Minute – October 5

Transcript

The world of high finance is a place that adopts technology very quickly and uses it as a competitive advantage. Artificial intelligence is no exception. Investors are willing to go to extraordinary lengths for any kind of advantage.

Since data travels through wires at nearly the speed of light, you might guess that your proximity to the building in New York where electronic trading occurs is irrelevant in the grand scheme of things. Far from it. Those who make the algorithms that buy and sell stocks think about it in different times. To them, the data travels at an irritatingly speed of one foot per nanosecond. Nowhere else is the axiom “time is money” more true than here, for those in the financial world know that if they get just a one millisecond advantage on their competitors, through faster computers or tighter code, it can be worth millions of dollars a day in extra profits.

Thus it should come as no surprise that the Wall Street tech nerds salivate when they think of all that is possible with artificial intelligence. Given that there is still a single room in the New York Stock Exchange that houses 10,000 computers, each buying and selling at speeds that dwarf the imagination, there is tremendous amounts of data for AI to sink its teeth into. AI, of course, loves data.

The financial world, therefore, gives us a glimpse of things to come. It is an area where we can no longer operate without the machines, and, here is the important part, there is no one who really understands what is going on. We just can’t keep up. There are an overwhelming number of transactions that happen each day, all triggered by the interplay of countless algorithms on innumerable computers, and all subject to potential mysterious and indecipherable behavior such as flash crashes. These events likely cannot be predicted, prevented, nor even understood afterwards.