I purchased this book for personal use.
Computers are everywhere. We knowingly interact with them many times a day, but they also influence us in unexpected ways. In Automate This, author Charles Steiner describes what computer algorithms can do and how they affect finance, music, medicine, and other areas of our lives.
Where the Money Is
Steiner begins with the story of Thomas Peterffy, one of the first Wall Street traders to use algorithms to generate and execute his trades. In 1987, Peterffy piped the NASDAQ feed to an IBM computer and let it generate and execute his orders for him. When a NASDAQ exchange representative gave him a week to change his system to enter orders one at a time using a keyboard, he had his engineers design a system that read the screen using optical character recognition technology, let his algorithm determine the order, and typed in the instructions using a robotic hand. Upon seeing the new setup, the NASDAQ inspector turned and left without saying a word. At the time Steiner wrote Automate This, Peterffy was worth $5 billion.
Peterffy's secret was to look for price differences for the same item in different stock or commodity exchanges. For example, his firm might be able to buy options for $35 on one exchange and sell them for $40 on another. This practice, known as arbitrage, is as old as commerce itself. The innovation was being able to use the automated data feed to identify and capitalize on these discrepancies in the fractions of a second it took the market prices to realign.
As the methods of Peterffy and other pioneers became more widely known, the price discrepancies decreased and the speed at which firms acquired and executed trades increased. One company, Spread Networks, installed a one-inch fiber optic cable from Chicago to New York that was significantly shorter than its major competitors' lines. The more direct path meant data could make the round trip from Chicago to New York and back in 13 milliseconds, 4 milliseconds faster than it could on other lines. When you're working with computers that can perform over three billion calculations per second, 4 milliseconds is an enormous advantage.
That said, the automated trading bonanza appears to be receding. An October 15, 2012 New York Times article quotes Peterffy as saying that the cost of technology required to keep pace with other high speed traders has increased to the point that it consumes much of profit. Add in the number of firms attempting high speed trading tactics and you end up with very small value fluctuations and limited opportunities for arbitrage.
Bots and Bach
Some music executives claim to have an ear for songs the public will love. Maybe some do, maybe they don't. What we do know is that an increasing number of music executives pay attention to the results of the algorithm used by the service Music X-Ray. The system uses an algorithm to analyze a song to determine whether it has the potential to be a Top 40 hit — the service lets songwriters upload their compositions so the algorithm can evaluate them. Music labels pay attention to the results and listen to songs the algorithm rates favorably. In several cases, the service identified songs that went on to become significant hits in Europe and elsewhere.
Identifying Top 40 hits is one thing, but one would think that composing music in the style of Bach or other classical masters must surely be beyond the capabilities of even the most sophisticated algorithms. One would be wrong. David Cope, professor emeritus at Cal Santa Cruz, developed an algorithm named Emmy (for Experiments in Musical Intelligence) that could produce works which were indistinguishable from those of Bach, Chopin, or Mozart. I use the past tense because Cope ultimately bowed to pressure from many in the music community and erased the databases Emmy used to generate its compositions. He had already begun working on another construct, so the loss wasn't total.
Steiner discusses other applications of algorithms such as medical expert systems, voice recognition, and speech processing. Automate This was written for a general audience and is in many ways a "personalities" book, meaning that the author profiles interesting people to put a human face on a complex subject. He describes machine learning, productive algorithms, and data mining at high levels, but spends much more time examining how those technologies affect commerce, medicine, customer service, music, and other fields. I enjoyed the book and think it will serve its target audience well.
Curtis Frye is the editor of Technology and Society Book Reviews. He is the author of more than 30 books, including Improspectives, his look at applying the principles of improv comedy to business and life. His list includes more than 20 books for Microsoft Press and O'Reilly Media; he has also created over a dozen online training courses for lynda.com. In addition to his writing, Curt is a keynote speaker and entertainer. You can find more information about him at www.curtisfrye.com.