The practical application of machine learning to e-discovery, commonly referred to as predictive coding, has begun to move from just a debated topic, to an applied technology. As developers of machine learning explain the efficiencies of their processes, law firms and review companies have the same natural reactions that most do when learning about an industry changing technology. These reactions include the initial confusion of how this complex technology works, the doubt of putting decision making into the hands of computers and the concern of potentially having jobs replaced by the new technology. As discussion of the projected impact of this technology has begun to find its way into the mainstream media, it is the jobs impact issue that seems to be getting a lot of ink these days, perhaps highlighted by the concerns of an economy that is still seeking traction.
Take for example the recently released book – Race Against The Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy, by Erik Brynjolfsson and Andrew McAfee. In a Yahoo! news blog interview of McAffee, the research scientist for MIT’s Sloan School of Business suggests that a busboy might have less anxiety about job prospects than a lawyer due to advances in technology. This is a comparison that seems a bit far-fetched and one that preys directly on the fears of an uncertain economy.
But the attention focused on jobs impact is nothing new, back in March the New York Times ran a story headlined “An Army of Expensive Lawyers Replaced by Cheaper Computers.” And while the story did discuss the benefits of machine learning, it is also implied that the technology was putting lawyers out of work by eliminating them from the document review process. In practice this is not true. The intention of machine learning is not to replace lawyers, but rather to reduce the burden on lawyers as they deal with an overwhelming deluge of data.
Servient has developed applications and technologies that greatly reduce the time and cost of document review. It is critical to understand though that our learning technology, Predictive Review, is based on the input of skilled legal knowledge. We don’t eliminate the lawyer and their knowledge of the case; in fact it is intrinsic to our process. What we do accomplish is to speed the process and free the legal team from the burden of reviewing documents that are irrelevant to the case. A key economic benefit is that we reduce costs that can be passed on in the form of savings to the client; an opportunity that we feel will ultimately provide the firm that uses it with a competitive advantage in the marketplace.
Rather than focus on the fears that a troubled economy may foster we suggest that a more holistic discussion of advanced learning technology and its place in the e-discovery process be taken. We would begin our dialogue by stressing that the developers of e-discovery machine learning technologies are not removing lawyers from the document review process. Machine learning technology, such as our Predictive Review, can free up legal talent from the burden of reviewing documents that are irrelevant to the case and focus on the documents that are relevant. The result is a gain in productivity and increased intellectual availability, which are arguably very positive economic impacts.
When one considers the impact that advanced technology has on the legal profession, it is helpful to ask whether reviewing a mountain of completely irrelevant email is truly a lawyer’s job; or is it instead more aptly described as a wasteful increase in the cost of litigation. The advanced technology that reduces the burden created by the mountain of irrelevant documents involved in today’s electronic discovery does not do away with lawyer jobs. Instead it helps to preserve the economic viability of litigation.