Knight Capital’s recent $440 million computer-glitch trading loss is another chapter in the long story of machines gone wrong. Still, this is…crazy.

First, some background: Knight Capital is a brokerage firm specializing in computer-driven, high-speed, and high-frequency trading. Knight’s computers scan dozens of public and private marketplaces simultaneously, using complex algorithms to buy and sell millions of shares in minutes, earning a fraction of a cent off each share. They’re not alone in this approach—firms that use “algos,” as they’re known on Wall Street, account for more than half of all stock trading.

On Aug. 1, Knight Capital introduced new software to take advantage of a change in regulations. As anyone who has read Frankenstein or watched Terminator and 2001: A Space Odyssey could have predicted, the software malfunctioned, erratically buying and selling millions of shares. In perhaps the stupidest side note of the whole mind-bogglingly stupid affair, the computer lacked a “kill switch.” The runaway machine went on trading for thirty minutes. By the time they shut it down, Knight Capital had lost $440 million.

Before we get worked up, let’s relish the delicious irony of a trading company that built its entire business on nanosecond-quick transactions having to watch its own machine run rampant for thirty minutes.

But now let’s get worked up. Because this is more than the fact that greedy technocrats lost nearly a half-billion dollars in a span of minutes. The fiasco caused wild swings in a vulnerable market, affecting more than 100 companies. And, more so, these wild swings are recurring with frightening regularity.

The Knight Capital fiasco was the third stock-trading debacle in the last five months. Before these was the “flash crash” of 2010, when, triggered by another errant algorithm, the market lost trillions of dollars of value in minutes. High-frequency traders like Knight have contributed to a stock market, according to The Wall Street Journal, “so fragmented and fragile that a single large trade could send stocks into a sudden spiral.” And, as with the mortgage-backed security disaster of 2008, regulators have been slow, inefficient, or apathetic in keeping pace with the highly complex and dizzyingly fast market system.

Of course, one way of reducing market volatility is to pare back high-frequency trading. One way to do this is to force firms to pay a small levy on each trade. In fact, our own Peter DeFazio introduced a bill containing a financial transaction tax of three-hundredths of a percentage point on each trade. Though critics argue that this would reduce trading, DeFazio (rightly) says, “Somehow we built the strongest industrial nation on Earth without algorithmic trading.” After all, these traders aren’t investors, they’re speculators—and their speculative transactions provide little if any benefit to the real economy. The machines may have taken over Wall Street, but we can still pull the plug.

(This article first appeared in the Sept. 5, 2012 issue of The Corvallis Advocate)