Sunday, October 2, 2011
Here's a post from Slate on an issue we've been following for a while (here, here and here). In my opinion, there's no doubt software will replace lawyers for many routine tasks (heck, it's already happened). Software can do some tasks better and cheaper than the average junior associate. Plus, lawyer-bots don't take vacations, don't steal clients and they'll likely make your malpractice insurance carrier much happier. But see this.
Imagine you've been served with a legal complaint. Your startup company makes a very popular widget, and your chief rival, MicroWidget International, is suing for patent infringement. If MicroWidget prevails, you could be out tens of millions of dollars.
You go to your in-house lawyer, who recommends that your company hire a patent expert at Moneybags & Moneybags, LLP. The next day, a brigade of Moneybags lawyers marches into your firm and outlines your options. Defending the MicroWidget case will be costly—in addition to the thousands of billable hours for the attorneys assigned to your case, you'll have to hire expert witnesses and jury-selection specialists, and pay for travel and court filing fees. The total legal bill will be about $5 million, give or take. But Mr. Moneybags, Esq., the firm's ancient senior partner, assures you it will be money well spent—he's worked on many of these cases, he says with a wink, and he's got a gut feeling you'll win this one.There's no easy answer. The legal industry is one of the few remaining outposts of the corporate world whose operations are dictated mainly by human experience. Basic questions that anyone would want to know before committing to a million-dollar case—How likely is it that I'll win? How good are my lawyers? Should I settle?—can't be answered with certainty. "There's a culture in the law around expertise," says Daniel Katz, an assistant professor at the Michigan State University College of Law who is among the vanguard of legal researchers working to bring empiricism and artificial intelligence into law. "There's a lot of human intuition, and people tend to think that whatever legal knowledge they have is uniquely human, and not subjectable to data and computers and automation."
Your other option is to settle with MicroWidget. You could pay a licensing fee of $10 million, which would be painful, but not fatal, to your firm—and it would allow you to go on with your business. What do you do?
Katz is working on something he calls "quantitative legal prediction." Thousands of patent cases are filed every year in the United States. There's a good chance, then, that MicroWidget's case against you shares some similarities with a bunch of those other cases. What if you could analyze the key features of MicroWidget's claim, and then see how thousands of comparable cases fared? "Lawyers will be able to say to their clients, 'Here's what we think your chances are—and based on 10,000 cases that are just like yours, here's what the computer thinks your chances are,' " Katz explains.