Watson Comes to Law


The latest news from the world of technology suggests that advances in computation may disrupt the legal profession sooner and more broadly than I had thought. Students at the University of Toronto recently designed a new legal search tool, winning a competition for the best use of IBM’s newest computational resource, Watson.  Specially designed and programmed,  Watson challenged the best Jeopardy players in the world in 2011 – and won. IBM, however, was not aiming at world Jeopardy domination but at making money by invading other more lucrative domains. And it has already spun off a division to exploit Watson’s technology in fields as varied as medical diagnostics and aerospace engineering.

Wisely, IBM has also begun university competitions to interest students in designing new uses for Watson.  The result from Canada is Ross, an application  expressly designed for legal research.  Computerized legal research is itself nothing new, having begun over forty years ago.  Today, Lexis/Nexis and Westlaw are better known than any single law firm. But Ross has two advantages over the kind of computerized legal search most of us have known.

First, Ross boasts a higher signal-to-noise ratio.   In response to a query, Ross will not merely deliver a lot of information, some of which may be relevant.  Instead, Ross provides targeted information and may even provide a direct answer, such as how many days are needed to file a case.   Second, like Watson, it provides an estimate of its certainty of the answer.

The designer calls Ross “Siri for lawyers,” which seems apt.   As Gary Morganthaler, an early backer of the technology behind Siri, the Apple search engine, put it, the future of search is “to deliver the information you want, not in a million blue links, but in one correct answer.” That is no less true of legal search than search in general.

Ross is not a lawyer.  It still needs someone knowledgeable about law to make good use of it.  But as Ross and similar applications are perfected they will make search more efficient and may make it much easier for a senior lawyer to dispense with the services of a junior gofer. This process is often the trajectory of technology: it strengthens the position of the better, more knowledgeable people in the field because they need less low-level assistance.

So Ross isn’t going to displace many associates tomorrow or even next year. But within a decade or a little longer these kinds of programs will transform the legal world. All law students now must consider how to prosper in a world of legal machines.  And their law professors should do the same.