Obiter Dicta | The Blog of Ehren J. Brav

Not much. But how much scrivener work will computers replace? Almost all.

It turns out we’re doing more and more shadow work as time goes on: we scan our own groceries, we locate books at Barnes & Nobel by punching them into a computer, we pump our own gas, we do much of our own personal banking and investing without ever talking to a human, we do our own check-in at the airport. Routine services all around us are being replaced by technology. Maybe it’s a generational thing, but I feel more comfortable interacting directly with the system than having a person intermediate. And I’d rather just log into a web page to handle my business directly than wait on hold trying to connect to a customer representative – I get frustrated if I need to call someone to resolve an issue and usually find it faster to figure it out myself.

It was the same way for me practicing law. I rarely utilized my assistant – most of the time I did my own time sheets, filed my own expense reports, edited my own documents and booked my own conference rooms. Our system had made these tasks so simple that it was actually faster for me to handle these things myself than to outsource them. These administrative tasks historically were performed by people but have largely been replaced by computers.

Technology will continue to move higher up the value chain, making workers more and more productive (whether or not you feel this is a good thing – and it’s certainly debatable – the trend isn’t reversing itself anytime soon).

In the law, the next frontier is what I call the scrivener work – locating, categorizing and comparing language (both case law and contractual text). This is largely hunt-and-seek work: finding precedents, summarizing contracts and comparing language. Because the technology to handle this hasn’t previously existed, up to now it has been done largely by lawyers themselves. But it has started to take an oppressive amount of time to handle – it’s a problem both for lawyers who get bogged down by scrivener work and for clients who pay high-value workers to handle lower-value tasks.

Scrivener work is distinct from legal work. Legal work is what you went to law school to learn – advising clients on risks, opining on the legality of a proposed action, distilling experience and precedent into actionable recommendations and of course drafting and negotiating documents. But the legal work is increasingly crowded out by the scrivener work.

This phenomenon has already played out in discovery – in any sizable case, it is no longer remotely realistic to have high-priced attorneys manually reviewing each document. Discovery materials are now segmented into different categories for different levels of review: (i) the critical information, which must be reviewed by experienced attorneys, (ii) relevant but not crucial information, which can be reviewed and summarized by more inexperienced attorneys and (iii) everything else. eDiscovery software makes this segmentation possible.

Similarly, the scrivener work in corporate practice will soon be replaced by computers. It will be much easier to find precedents, assemble standard agreements and pinpoint missing or unacceptable provisions in a set of complex documents. Lawyers should view this as a good thing – it allows them to focus on the higher-value tasks (which also happen to be more professionally fulfilling) and provide their services more efficiently.

So the real question is what the adoption curve will look like. Law firms are notorious for still using Windows XP and Office 2003. But ultimately client demands drive adoption – not only in the form of alternative fee arrangements and push-back on billable rates, but also in terms of business development: the firms that can best demonstrate technological efficiency to their clients will gain marketshare. And firms will face increasing pressure from legal outsourcing, both from South Asia and low-cost jurisdictions in the US. The adoption of this technology will mirror that of eDiscovery – the pain becomes so acute, it forms a completely new category of technology (and its own neologism) in just a few years.

For younger attorneys, this is a big opportunity. Being able to understand and leverage technology is becoming a real way of differentiating yourself in an increasingly competitive job market. When I started at my firm, nobody asked me whether I knew how to use Windows, Microsoft Word and Westlaw – it was assumed. In my day, attorneys who were adept with Excel and PowerPoint were ahead of the curve. In the future, using next generation tools to perform more efficiently and produce higher quality work will be a big advantage. For those attorneys that embrace these technologies, it will also make practicing law more fun.

Why is Google’s famous slogan “organizing the world’s information” rather than managing the world’s information? If managing something means controlling it, knowledge management is really about controlling information.  But isn’t that like saying we need to control the internet? The producers of information will always win – in the long run, there’s just no way you can hope to control all the data that people create.

An early knowledge manager grapples with data from the Venerable Bede 

In the early days of the internet, several companies tried to categorize all the web pages out there into a giant index. If you wanted to buy a car, you might click on the “Automobile” category, then the “Consumer Reports” or “Dealerships” categories and browse through the various listings, which were of widely varying quality. But this task of indexing and categorizing the internet quickly proved entirely unworkable – there were simply too many pages out there, there were more going live every day, and a mere index could not tell you which of those millions of pages were actually worth looking at.

Companies are now facing this same challenge in microcosm.  Information is exploding – in emails, documents, archives and even voice mails. We’re become shockingly adept both at producing new information and recording it.

So why should companies take the same approach as those early internet portals?

The only feasible alternative is “Knowledge Organization” rather than “Knowledge Management” – companies need an internal Google to allow the consumers of information to consume it in the way they want, rather than being forced into a particular framework that works for one point in time and for one type of user.

This problem comes up in spades with Contract Management Software (an unfortunate name, usually abbreviated as CMS). Every company beyond a certain size needs to come up with a process for tracking its agreements. This is a huge problem – without some way of handling this, deadlines are missed, cost-savings opportunities are lost, and even worse, agreements might be unwittingly breached. The current solution is to use software that imposes such a process. But producers of legal text, consumers of it and managers of the system are forced into a framework that the software designer thought would work. And it does work, so long as you have enough people willing to manage all this contractual data and so long as those people buy into the software’s way of doing things.

But in the long run, it won’t scale. For one thing, producers of information have a limited tolerance for entering meta-data. This drove me crazy in my law practice, where it seemed you couldn’t open or save a document without spending 10 minutes filling out a form. Patience for doing this just won’t last. I’ve heard of situations where a law firm knowledge manager would actually bribe attorneys with Starbucks cards to get them to fill out the forms. The costs of managing all this information increases rapidly with its volume. Worse, as we continue to get better at producing and recording information (think social media), this problem grows in a non-linear (read: scary-fast) way.

Hard day at work today? Actually, work was fine – it’s the meta-work that’s killing me! (Photo by Fred Fehl)

We need a Google for contracts – something that organizes the information without trying to control it. Imagine typing in a query like, “Leases needing to be renewed in the next six months,” and having the software give you a list of hits, with the relevant language from each agreement at your fingertips.

The cost savings such a service provides would be staggering. It would dramatically reduce the need for manual review of these agreements (especially if delegated to outside counsel) and neither the producers nor consumers of the information would need to slow down their work to deal with meta-data.

If all this sounds futuristic, it’s actually quite easy to envision. Technology to parse and recognize text has advanced dramatically over the last ten years, thanks in large part to eDiscovery. When it comes to legal language, particularly form contract legal language, the challenge of recognizing and extracting the desired provisions becomes quite manageable.

So what’s needed is a computer that can understand a Google-like query to find the relevant provisions among a company’s contracts. Not necessarily a computer to actually understand the language, but simply to locate it (Google doesn’t understand the hits it generates, it just finds them for you). That’s the next generation of legal tech – on the horizon now but rapidly coming into view.