If computers can beat chess and Jeopardy champions, then they will soon take over other aspects of our lives from driving cars, such as in Nevada, to reviewing documents in litigation. Although my reasoning is invalid, it’s important to note that Nevada opened its highways to robot (computer-driven) vehicles, not its urban areas. To date, the New York and Boston metropolitan areas have not followed Nevada’s lead because driving in unpredictable and heavy traffic is not for robots or computers alone. And like heavy traffic, complex e-discovery is not for computers alone, but they can help humans make it more efficient and cost-effective.
Computers use binary numbers for input and output. They receive or read bits of input in zeros and ones and output the same. Granted, they do this very, very fast in consistent, repetitive manners that make them efficient candidates to review large sets of documents and determine whether or not some documents are relevant, or not, to a set of litigation facts. For example, Anne Kershaw and Joseph Howie found that when e-discovery providers used technology-assisted review with predictive coding, they were more cost-efficient than humans in a “brute force linear review.”