Course intro

From: Brock Rutter
To: Digital Drafting spring 2010 class
Re: First day of class – course overview
January, 2010

How are journalism and the practice of law similar? Practitioners of each see themselves as engaged in more than just a way to make a living, but in a profession. As a profession each sees itself as adhering to a sort of code of ethics and has its own set of standards and professional norms. Each of them has also had a relatively stable business model for the last hundred years or more. For lawyers it has been personalized service based on the billable hour. For journalism it has been a model of professional journalists employed by newspapers. The newspapers have made money selling a products at a nominal cost (the cost of buying a paper) making most of their money from selling advertisements.

We shouldn’t overemphasize the similarity between journalists and lawyers. Law is by far an older presentation and has generally been the more prestigious profession. Lawyers were held in high regard by founders of this country, as was a free press. The concept of the rule of law is at the heart of the U.S. Constitution, and it has generally been assumed that it is the job of lawyers to defend the rule of law. Likewise the idea of a free press is enshrined in the Constitution. (Although the free press of the time of the Constitution quite possibly had more in common with a bunch of anonymous bloggers than the professional journalism that developed later.)

Another trait journalism and law have in common is that in their current incarnations both came to be seen as a given – an unchangeable fact of life. It seems natural to assume that the New York Times will always print all the news that’s fit to print and that Big Law will always serve the needs of big corporations, at big prices, paying big salaries, and scooping up the best graduates of the best law schools. Yet, 2008 and 2009 have not been kind to either professional journalism or traditional law as practiced in the big firms. For journalism –scores of regional papers closed, the Boston Globe almost closed, the New York Times was seen as such a bad credit risk it was forced to pay an exorbitant 14 percent interest rate to borrow money. Industry soul-searching about the uncertain future was exemplified by efforts such as “Saving the news: Toward a national strategy for journalism”, a publication dedicated to looking for viable solutions to the newspaper industry’s problems. The last two years have not been kind to the legal profession either. Many firms have either closed or have drastically curtailed hiring.

Technology lies at the heart of both the newspaper industry’s problems and those of the legal profession. The once dominant newspaper of general circulation is now threatened by niche publications and free news – all available over the internet. Meanwhile, new advertising platforms, such as Craigslist, have taken the “bread and butter” of many papers – local advertising, especially classifieds.

Of course it is tempting to speculate that the widespread economic downturn is responsible for a decline in law firm business. However, lawyers have often been able to have it both ways. In the good times the firms transactional and mergers and acquisition departments could make a profit and in the bad times the litigation and bankruptcy departments could do pretty well too. In fact it is often taken as a given that litigation increases in a downturn. However, this downturn there have been real concerns that something might really be changing – that the long dominant law firm business model might be changing.

The dominant business model in question involves two hallmarks; 1) personalized service from the ground up, and 2) big firms with pyramid structures with a few partners at the top and lots of associates at the bottom (doing much of the work). Of course these firms are usually located near the businesses they serve – that is in very expensive parts of town – and employ any of the same signaling techniques once used by banks – big fancy buildings meant to convey strength and permanence. They also incidentally cherish a carefully groomed image for conservatism and risk aversion – a sort of counterweight to the animal spirits of the business clients they help navigate through the world of legal risk. Even if this description is skewed toward big firms serving the needs of businesses, it holds true that in general, we expect lawyers to be conservative, risk averse people. We expect fully customized service, even if it is expensive. And, although exceptions may be more common among smaller firms, the trend has been toward the pyramid-shaped business model.

The unchanging nature of the legal system is so remarkable – and perhaps different from other business sectors – that is almost a truism. At a recent meeting concerning overhaul of the Vermont judiciary, an outside consultant remarked that if an 18th century lawyer was transported to a 21st century courtroom the only difference he would notice would be the lack of wigs and the changed style of dress. Likewise, Richard Susskind, one of the foremost authors on law and technology, remarked to a Canadian bar magazine that not much has changed in the practice of law since the days of Jarndyce and Jarndyce – that is the time of Chares Dickens.

Other examples abound. Property law provides an extreme case. A myriad of antiquated forms and practices survive in part because no lawyer wants to be the one to try something new and have a conveyance later be challenged because she deviated from a time-honored form. While other businesses prodigiously give rise to neologisms, the jargon of law remains steadfastly conservative. Despite the inroads of the plain language movement, when the chips are on the table, most lawyers will not stray from time-honored terms of art.

For whatever reason or reasons, law has notoriously slow to adopt new technologies. A great many still-practicing lawyers can remember when it was seen as nearly scandalous that a lawyer would have a computer on is or her desk and do his or her own typing. That has now changed, but the change came more slowly to law than to other professions. A comparison with (the comparatively upstart) profession of accounting provides an example. Somewhere between ten and twenty years ago, the penetration of technology in the two professions was roughly equal. Accounting adopted technology much more quickly. And while the two business models were previously very similar, many of the trends technology is bringing about in law are much further along in accounting than in law.

Lawyers have long had the luxury of seeing themselves as involved in an endeavor that was somehow special or different from other aspects of business. This may be partly because of law’s dual nature as both a protector of the republic and also as a more mundane transactional formality or economic necessity.

While we would perhaps not want to delegate our First Amendment jurisprudence to a computer program, or outsource a federalism question to smart foreign lawyers willing to work for cheap, a computer program can almost certainly be programmed to provide advice in a bounded area of law – say securities compliance – or prepare a document such as an nondisclosure agreement or a will. While they might not sound as glamorous as arguing the next headline Supreme Court case, relatively repetitive – commoditizable – legal services comprise the bulk – the bread and butter – of the run of the mill business of a great many legal firms.

It is in these areas that technology is making inroads into the business of law. Computer programs are being used to draft contracts, give advice, administer government benefits, and generally allow businesses and individuals to do more without the direct intervention of legal counsel, or in some cases with the preventative involvement of automated systems that alert humans to potential legal issues before they become real legal catastrophes.

Other computer programs are also helping with the massive responsibility that our reliance on computers has brought on by vastly increasing the complexity of discovery now that businesses store data by the giga- or tera-byte. In some cases – such as the 2006 amendments to the FRCP, rules had to be re-written to make sense for the digital era. In this case lawyers had to follow business into the world of computers and found that old discovery rules written for the paper world didn’t always produce desirable results for the computer world.

The use of technology and the business model changes it has made possible has also raised new ethical questions. You might have heard of some raised by metadata in your legal profession class, but others abound. For example, how are ethical obligations changed when the lawyer and the client do not interact face-to-face but over the internet? What if a lawyer sets in motion a computer program that gives legal advice that turns out to be faulty? Can a computer program practice law without a license? What about confidentiality and data security? Is it ethical to trust a system – e-mail, “the cloud” – that inevitably breaks down every once in a while?

We will examine all of these areas later in this semester. We will also undertake a few projects to engage in the types of legal automation projects that are actually going on in the real world. In teaching this class, we will also strive to emphasize the continuity that does exist between new, technologically enabled practices and the practices of the past. The analog antecedents of document automation can be seen in legal forms books or law office precedent libraries. Metaphors to the physical world abound in the world of law v and technology. Some of available metaphors get stretched at the edges; a recent opinion involving the privacy rights attached to e-mails and reasoning by analogy to old fashioned mail was complicated enough that it caused Orin Kerr, a George Washington University law professor to post a blog entry one day and then another the next day saying he had misread the opinion.

You might be intuitively skeptical about the potential of computers to perform legal functions. Perhaps without even knowing exactly why, you might doubt the ability of computers to do meaningful work in the field of law. In the late 18th century a machine called The Turk was for a while a favorite curiosity across Europe. The “machine” would play chess – often quite well – to the astonishment of spectators across Europe and the United States.

The machine was, in fact, a hoax; a collaborator hidden within the table beneath the chess board actually moved the pieces. The real ingenuity had not been programming an “automaton” (in the language of the time) chess player, but in artfully hiding a human one. But what made it such a convincing hoax was that it had been well done and people actually believed they were witnessing something they intuitively didn’t believe was possible – a machine was playing chess.

To us, computers doing law seems similarly counterintuitive. But huge developments have been made since the days of The Turk in the late 18th century. Today, computers can play chess – quite well; in 1997 a machine built by IBM, Deep Blue, defeated world grandmaster Gary Kasparov in a series of 6 games (3 draws, 2 wins for the computer, and only one for Kasparov.) Deep Blue actually accomplished this win through “brute force” computing – being able to think between 17 to 19 moves into the future in the middle of the game. There are only so many possibilities, and with enough computing power they can all be considered. Checkers, which is mathematically a much more limited (bounded) universe than chess, has already been solved. All of the 5 billion billion (5 x 1020) possible game positions have been mapped into a database. A computer with access to the database cannot lose. If two such computers played each other they would tie.

Ares of the law where computers have made the most inroads have similarities with the game of checkers. Like a checkerboard, they are bounded areas of possibilities, and like chess, they are of limited granularity – there are only so many moves or so many possible combinations of facts or clauses in a document. They are rules based as opposed to case base. These bounded areas of the law are where we will spend much of the course time. Pick any area of law, and provided it is sufficiently bounded and rules based, someone will have attempted, or will soon attempt, to increase efficiency through a computer program of some sort.

Document assembly can take an area of law where a limited set of documents – or permutations of possible combinations – is used again and again. Based on a custom made interview set and a custom made set of possible clauses or document components, a program can determine what elements to include in a document and assemble them in the appropriate order. You will be designing document assembly projects to deal with a limited area of law – for example designing a system to build a nondisclosure agreement. As long as all the possible permutations are considered – as long as they are all in your database – the program will work. There might always be outliers, such as a new contract that calls for a clause or group of clauses not considered when the program was designed, but that is akin to changing the size of the board or the number of possible moves. Even the new moves can eventually be added to the database.

Other expert systems are used as advice systems. Once again these function in bounded areas of the law. For example, an advice system could function to administer a public benefits system or help decide if a proposed securities offering is legal. Such systems are now used by some of the biggest law firms in the world – with clients staking serious money on the strength of the advice. They have been used both as internal productivity tools and also as outward-facing subscription services where the client never interacts with a lawyer. You will probably also notice that these share more similarities with document assembly systems than differences. The basic architecture is the same but the output is advice or instructions rather than a document such as a contract.

Connectivity is becoming the default mode of the internet, machines, and people in general. Like other aspects of life, people would like the law to be interoperable. Making a computer that can read natural language (the way you and I speak and write every day) remains an elusive goal of artificial intelligence and computer programming that no one has yet gotten quite right. Despite this, computer “tagging” – XML and its variants – is making various it possible for computers to act as if they knew what various contract clauses and other legal bits of information mean. Like the database that enabled a computer program to “solve” chess, each bit of information that is to be tagged will first require human attention. However, a key benefit of a system of tagging is that it can be rapidly expanded and added to as needs arise. However, like other systems involving digital capture of analog information, sacrifices are made in terms of granularity in order to gain the benefits of computer efficiency.

Computers’ capabilities certainly do have their limits. This can be seen when very large or very fine grained systems are in play or vague language is used. We certainly have intuitive notions of subject matter that is not amenable to computer solutions. Sometimes it is a matter of the problem set being too large or the possible elements being too numerous. For example, chess is reckoned to have 1040 possible game positions. A computer that could solve checkers in a nanosecond would reportedly require 3,000 years to solve chess.

But still theoretically, chess could be solved. However, there might be things that computers can’t do at all. Reasoning by analogy may be chief among them. At various times, enthusiasts for computers and law or artificial intelligence and law have made perhaps unrealistic claims about the ability of computers to perform legal functions. Certainly many of them have not come to pass. To a certain extent, the term AI is laden with baggage picked up through association with the over-optimistic hopes and claims of its enthusiasts. Even if we imagine a world where a computer can solve for the 1040 game position possibilities of a game of chess – and that still can’t be done, but possibly will be in your lifetimes – this is still a far cry from the infinite possibilities of real life. We favor the more pragmatic modern view of computers’ potential in the legal field. Until computers can intuit and reason by analogy – and there’s no evidence they will be able to any time soon – there will always a significant area of the law where they are simply not useful. Drawing this line at, or slightly before, analogical reasoning, we are skeptical of the ability of computers to engage in case-based reasoning, and that will not be part of this course.

Slightly after the halfway point in the semester we will move beyond areas where computers have been doing a fairly good job of performing limited legal functions. Our next area will focus on an area where lawyers have had to follow the techies – they’ve had to learn new tricks to practice law in an era when almost all information in any organization of any size is stored in computers. For two or three weeks when we discuss e-discovery this class will proceed more like a regular law school class; we’ll read cases and discuss them. It is safe to say that e-discovery will remain a major area of the law for a significant part of your careers. In fact, it will probably be the main type of discovery you experience in the practice of law.

The advent of e-discovery and questions about how the traditional rules of discovery would apply in the electronic world led to serious confusion and ample opportunities for shenanigans in recent years. Now it seems that the broad outlines of the world of e-discovery are sketched out, although the details continue to develop. E-discovery has also lead to a vast increase in the importance of discovery relative to the matter at hand. It’s possibly limitless scope – it’s potential to run up costs that eclipse the amount in controversy of the underlying litigation – has led to the development of new concepts of proportionality, and rules for shifting the cost of document production in certain cases, and sometimes a new emphasis on cooperation. E-discovery has also helped spawn a new class of “litigation support” professionals straddling the boundary between law and technology. Lawyers who can speak the language of technology have come into increased demand as lawyers and IT must work together in the course of litigation and also before litigation starts to devise document management and retention policies.

Finally, we will look at the ethical implications of brought about by the use of technology in law. These are legion, and can range from whether a computer program can commit unlicensed practice of law, unbundling issues (whether a lawyer can represent a client in a limited portion of an issue without taking on responsibility for all aspects of thee matter), and issues of security, when is it safe to send sensitive information over the internet or similar networks. We’ll also look at some of the secondary effects of developing technology, such as a perceived power shift away from outside counsel to in-house counsel, changing business models, whether the rules familiar rules of professional conduct need to be changed in the name of economic efficiency or social justice.