I’ll assume that you’re aware that Watson, IBM’s artificial-intelligence “question answering machine,” bested two human competitors on “Jeopardy” this week. (If not, check out this New York Times story.)
It didn’t take long for this event to acquire a legal angle, in the form of this National Law Journal article by Robert Weber, IBM’s general counsel. In it, he suggests that the new technology used by Watson could greatly help lawyers digest the flood of data now available to us.
Here’s the bit that caught my eye (emphasis added):
Just think of what this can mean for our profession.
Imagine a new kind of legal research system that can gather much of the information you need to do your job—a digital associate, if you will. With the technology underlying Watson, called Deep QA, you could have a vast, self-contained database loaded with all of the internal and external information related to your daily tasks, whether you’re preparing for litigation, protecting intellectual property, writing contracts or negotiating an acquisition. Pose a question and, in milliseconds, Deep QA can analyze hundreds of millions of pages of content and mine them for facts and conclusions—in about the time it takes to answer a question on a quiz show.
Deep QA presents exciting possibilities for all sorts of legal activities. But contract drafting? I think not: contract drafting isn’t about crunching data.
A lot of energy has been devoted to analyzing publicly available deal documents. It’s something you can do the old-fashioned way, by reading through stacks of contracts and tallying your findings by hand. The ABA Section of Business Law’s “Deal Points” surveys represent a good example of this approach. Or you can bring to bear gee-whiz technology, the best example being KIIAC. (Click here for my Q&A with Kingsley Martin of KIIAC.)
But if you’re interested in learning more than just how people draft contracts—if you’re looking to draft contracts expertly—the great scavenger hunt that is the SEC’s EDGAR system presents you with an insurmountable garbage-in, garbage-out problem. Crunching more data faster and more finely isn’t going to change that.
So when it comes to contract drafting, Deep QA won’t provide any meaningful help. It might play some sort of supporting role, but I’m not holding my breath.
The future of contract drafting lies elsewhere—in having a cadre of legal knowledge engineers create document-assembly systems for use by practitioners, who as a result are able to devote more of less of their time to what should be a commodity task.
We’ll soon get a sense of how that might play out.
3 thoughts on “Why IBM’s Watson Doesn’t Have Me Quaking in My Boots”
Your thesis appears to be that humans can learn from “reading through stacks of contracts” but a computer can’t: or put another way: only humans can filter garbage. Watson used the unfiltered resouce of the Internet to come up with a high rate of accuracy. The same can be done with unfiltered repositories, such as EDGAR.
Kingsley: That’s actually not my point. Instead, my point is that reading through stacks of contracts plays no part in what I do. For example, yesterday I spent a couple of hours rethinking my approach to severability provisions. For that, I looked at one treatise and zero contracts. I think it’s unpromising to expect to achieve quality by distilling the output—good, bad, and indifferent—of the transactional world. Instead, it’s editorial control that’s essential. Ken