By Steven Fettig. Taken at Meijer Gardens and Sculpture Park, Grand Rapids, Michigan.
By Steven Fettig. Taken at Meijer Gardens and Sculpture Park, Grand Rapids, Michigan.

In 2002, reporters asked Secretary of Defense Donal Rumsfeld a question at a U.S. Department of Defense news briefing. In answering, he set up a taxonomy that has become popular to catalogue our state of knowledge. In the Rumsfeld Taxonomy, there are things we know, things we don’t know, and things we don’t know we don’t know. In the Secretary’s words, “There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things that we know we don’t know. But there are also unknown unknowns.” The last category, as Secretary Rumsfeld noted, is the most intriguing.

Scientists Discover Tacit Knowledge

Michael Polyani was a Hungarian physical chemist. He studied in Budapest and Karlsruhe, Germany, but WW I interrupted his studies. He served as a medical officer during the war and, during a sick-leave, managed to write his PhD thesis (encouraged by Albert Einstein). He received his PhD from the University of Budapest after the war.

After teaching for years in Hungary, he emigrated to Germany and then found his way to the University of Manchester. With the turmoil in Europe, his interests had shifted from chemistry to economics. The University accommodated the shift by creating a chair for him in Social Science which he held until he retired from his distinguished career in 1958.

Years before he retired, Polanyi gave the Gifford Lectures at the University of Aberdeen. He published a revised version of his Lectures in 1958 as the book, Personal Knowledge. In the Lectures and book, Polanyi argues that all knowledge relies on personal judgments. That is, he argued, one cannot reduce knowledge to a set of rules. Polanyi’s views countered those of his friend Alan Turing and were the basis for some early critiques of work in artificial intelligence.

Polanyi extended this idea of personal judgments to a concept he called “tacit knowledge.” According to Polanyi, we experience the world both through sensory data and through other knowledge—tacit knowledge. Tacit knowledge includes things we aren’t aware we know, but which play an essential role in our lives and work.

Polanyi’s ideas have been the subject of much research. That research helped explain a problem that has bedeviled scientists for many years. As any high school student who has taken a science class knows, one of the bedrocks of science is the idea of the repeatable experiment. Scientist A conducts an experiment which yields results that meet a basic significance test. She publishes her results in a journal. Scientist B wants to extend Scientist A’s work. To get started, Scientist B tries to replicate Scientist A’s results. B runs the experiment as described in the Journal article, but gets results different from A. Were A’s results a fluke? Were B’s results a fluke? After many attempts, B and scientists C, D, and E are unable to repeat A’s results. Now what?

At first you might think such an outcome uncommon. Scientists publish in peer-reviewed journals. We assume that by the time an article makes it into print, the results it reports aren’t a fluke. Scientist A may have repeated her experiment several times before publishing to make sure her first results were not a fluke. The peer reviewers would catch any flaws in what she did. The data is public. So, absent fraud, we think A’s results are reliable. In fact, scientists still struggle with unrepeatable results. Why can’t anyone repeat them?

This is where Polanyi’s theory comes into play. Under the tacit knowledge theory, the steps in the journal article are not sufficient for other scientists to replicate the experiment. The missing element is tacit knowledge. In the case of A’s research, she has some tacit knowledge necessary to make the experiment work. Tacit knowledge goes beyond failure to create detailed instructions. It includes knowledge the person can’t articulate.

Science and the Unknown Unknowns

It is the time of the Cold War. Russian researchers led by Vladimir Braginsky at Moscow State University are working on ways to detect and measure gravitational waves. Measuring these waves is a big deal—you may recall seeing articles in 2016 describing how scientists had, for the first time, detected gravitational waves. Albert Einstein had predicted such waves 100 years ago.

The Russian researchers’ instruments used sapphire mirrors. Every little thing mattered in the search for gravitational waves, including the quality (“Q”) of the sapphire used in the mirrors.[1] The Russian researchers claimed to have measured a new, high quality level for their mirrors, something of great interest to those searching for gravitational waves. But, despite their best efforts, researchers at major universities including Caltech, Stanford, Perth, and Glasgow could not match the Russian’s results.

Since it was the Cold War, many were skeptical that the Russians had achieved what they said. As the years passed and no one could repeat the results, the skepticism grew. By 1998, the Cold War was over. Scientists from Glasgow University visited Moscow State University to learn how the Russians had managed to measure the impressively high Q.

After a week, the Glasgow scientists trusted the Russian scientists. With distrust out of the way, the Glasgow scientists focused on what the Russians were doing. It turned out, there was a lot to know beyond what the Journal article said.

Remember, the equipment is very sensitive. Construction and technique play critical roles in the measurement process. This was where the Russians had tacit knowledge. The Glasgow scientists learned how to suspend the sapphire, what to use (a certain silk thread from China worked best), the best length for the suspension thread, the most efficient way to create a vacuum for the test, and many other factors. They also learned patience. The Russian scientist doing the experiments would re-run the same experiment over many days making minute adjustments, before he would accept the results.

Some changes had explanations. But for many, the answer was akin to the famous dictum from Supreme Court Justice Potter Stewart when writing about pornography, “I know it when I see it.”[2] The Russian scientist could not articulate what he need to do, he just knew when he had to adjust the apparatus or run the experiment another time.

AI, Law, and The Tacit Knowledge Risk

As we see the earliest incremental steps of artificial intelligence creeping into law, we should ask whether tacit knowledge plays a role in the legal universe. It is easy to be dismissive and argue no (though I suspect lawyers will try to answer yes). Law is not an “exact” science like physics. The steps that physicists outside of Russia missed when trying to replicate the Q experiments were in many cases matters of omission. Had the Russians given long and detailed explanations of everything they did, the other scientists may have replicated the experiment.

If we push a bit further, the “yes” answer gains currency. Harry Collins has written extensively on tacit knowledge. In Tacit and Explicit Knowledge, the third book of his non-fiction trilogy studying knowledge “top to bottom,” he developed a “Three Phase Model” for tacit knowledge: relational, somatic, and collective. Relational addresses the “contingencies of social life,” somatic the “nature of the human body and brain,” and collective “the nature of human society.” Without delving into the Model, we can see that tacit knowledge includes more than what our senses tell us, it includes much going on around us.

In law, we moved from formalism to realism in the beginning of the 20th century (pragmatism never caught on). What lawyers and judges did involved something beyond formalism. Looking at the facts, reading cases and statutes, and applying the latter to the former was necessary, but not sufficient. The process needed an additional something, and it came from experience, both life and current. Reading the cases or statutes applicable to a set of facts did not give you all you needed to “apply the law.”

The tacit knowledge concept puts a name to what many lawyers try to articulate when they say we need lawyers. Sending a computer to law school, where it learns the theory and rules of law, is not sufficient to give us a practicing lawyer. Even having the computer read all the decisions of all the courts, study the hornbooks, and peruse law review articles falls short. The computer may learn what is in print, but it will not learn the “unknown, unknowns.” It will not learn what the lawyer or judge omitted from the papers. As important, it won’t know what it doesn’t know.

Tacit knowledge plays a role in shaping the biases and heuristics that Daniel Kahneman brought to our attention in behavioral economics. A judge deciding a case employs those biases and heuristics as she applies law to facts. To claim otherwise attempts to argue that judges are not human. But where does this knowledge take us?

Consider tacit knowledge along with artificial intelligence. AI uses machine learning. Imagine we gave AI software all of the cases ever decided involving securities law. We gave the same computer all the law review articles written, all the books published, and any other written thing we could find. The AI used machine learning to scour the materials for patterns. It found things we knew and some “patterns” we didn’t know. But are the new patterns correct? And, what about everything that wasn’t written down?

AI software stumbles when it comes to certain challenges. Law can magnify those challenges. Writing quality varies widely among judges. On a good day, judges may omit essential information from their opinions. On a bad day, they also omit logic. AI will have difficulty inferring what is missing. If 1,000 cases lack the same information, AI may find the pattern. But if only one case lacks the information, AI can’t find a pattern. Another challenge involves deciding what weight to assign each fact. The judge may list 10 facts, but not the importance of each fact to the outcome. Facts change by case, so finding a pattern is difficult.

Think of a decision involving a criminal sentence. Case law requires that Judges list the factors that played a role in sentencing. Most do, but some omit some or all of the factors they considered. The software may see a factor in the case and incorrectly think the judge considered it. The judge may have used her experience to weight recidivism risk factors when deciding what support services the defendant would get, but not listed her experience. Tacit knowledge plays a role in judicial decisions.

When we introduce AI into law, we need to ask what happens to tacit knowledge. If we think of AI as just doing a better job finding things, then we can argue it has little to no impact. AI finds cases faster than a person, but the person still reads and interprets the cases. But how does the AI know which cases to select versus the human? Would a person have selected a case, even though ambiguous, because it gave hints about new directions to pursue?

I am not pretending to answer the tacit knowledge question in this article. But I think we must ask the question as we expand our use of proto-AI and AI technologies. The question may not be what we found, but what we missed.

[1] The quality or Q factor for a material measures the rate at which its resonances decay. Think of a bell. You ring the bell and it takes time for the ringing to subside. The longer it takes, the higher the Q. The scientists wanted “high Q” sapphire and the Russians had measured a Q of 4 x 10 to the 8th.

[2] Jacobellis v. Ohio, 378 U.S. 185 (1964).

OverHypeWe know Benjamin Franklin for his many sayings. Some he created, most he borrowed and improved. One we all know. Two things are certain in life: death and taxes. Everyone has a take on the third, so I will add my voice to the fun: artificial intelligence in law is over-hyped. If the hyping AI is the most popular thing in legal industry writing , explaining how AI in law is over-hyped is the second most popular.

Collect all the AI in law articles, combine them into one big summary, and this is what you get. AI can do everything lawyers can do, but better. The future is on the horizon and the horizon is close. Retire folks. AI will do the research, write the brief, file the brief, read the brief, and decide the case. All in less time than it takes to say “unplug the darn thing.”

Law has company in suffering through hype. In fact, hyping tech has become such an art form that it has achieved consultant model status. Gartner (according to Gartner) is “the world’s leading information technology research and advisory company.” They put a name and a diagram on hype. They call it the “Gartner Hype Cycle” and it looks like this:

Gartner Consulting
Gartner Consulting

Plotting technology X on the Hype Cycle can be fun. Lawyers have avoided the Hype Cycle, because lawyers have avoided technology. But, we have joined the fray. AI is our achilles heel.

They Hype Cycle is a rearview mirror metric. It is tough to measure a technology’s place on the Cycle, but looking back you can see the peaks and troughs. It feels like we are near the top of the first incline approaching Peak of Inflated Expectations. If so, a few years should plunge us into the Trough of Disillusionment. Tighten your seat belt, please.

Since we know the cycle it seems natural to ask a simple question: can we skip all the craziness and go to the Plateau of Productivity. That is the question Eddie Copeland asked in his essay, “Busting the hype cycle: 5 questions to ask about any new technology.” In turn, my friend Peter Carayiannis asked whether Copeland’s essay ideas might apply to AI in law. I promised Peter a nuanced maybe a bit surprising answer. Let’s start with Copeland’s thoughts.

The Copeland Five Asks

Copeland identifies at least two downsides to hyped technology in the context of government initiatives. First, the government wastes taxpayer time and money as it pursues initiatives that have little or no hope of succeeding. At the same time, it sidelines initiatives that could help. Second, the disappoint that comes from realizing the hype means the anti-technologists dig in and changes becomes harder.

Copeland offers five questions we should ask if hype tempts us:

“1. What are we actually trying to do?

2. Are we over-engineering the solution?

3. Is it significantly better than what it replaces?

4. Is there a connection with those who will pay for and those who will benefit from the technology?

5. What skills and processes need to be in place for the technology to work (and are we willing to adopt them[)?]”

But that wasn’t Peter’s question. The simple answer to his question is “yes,” answering Copeland’s questions would help many firms. The interesting question is whether hype does us any good. Copeland answer the question with a “no,” but I’m going to give a quasi counter-argument.

Over-Hype Can Help

My first argument for hype in the legal industry is “the burning platform” view. Managing partners at law firms say they understand their firms need to change, and change bigly. The last report I saw put the number at 96%. But, equity partners at those firms oppose change, with 67% saying they want things to stay they way they are. The problem: in many firms the platform is peaceful.

We know the metaphor. Nothing happens until the platform starts burning. With fire comes a flurry of activity. The danger for large law firms sounds like the frog in the pot of water metaphor (and yes, I know this metaphor is wrong). The story is that if you put a frog in a pot of boiling water it will jump out. But if you put it in cold water and raise the heat it will stay until its unfortunate death. In real life, that isn’t what happens but it gives us a vivid mental image. We could compare large law firms to the frog in the story. At most firms, things are peaceful. Partners seem content to wait.

For some, waiting means seeing if they can make it to retirement without investing in serious change. For others, retirement is in the distance but the pressures of today exceed future risks. They risk being the boiled frog. They seem content to take the chance.

Hype may help. Hype creates a sense of urgency. It makes it sound as if the lawyers face rapid change. In the case of AI and law, the hype suggests that if law firms wait, the future will be dark and stormy. That hype is the burning platform triggering some firms to do something. In fact, this is what we have seen.

For two years, we have read reports of some firms sliding into AI activities. They have licensed software or started using AI-enabled services. Great fanfare, blowing of trumpets, and “huzzahs” have accompanied their moves. These firms get it! The grand transition to AI has begun. So, even with all the downsides, hype may cause some movement. In the legal industry, movement is tough to achieve, so hype may have some value.

My second argument is that hype my spur some change below the AI level. As firms look at the products and services available, they may realize that they should stay in a pre-AI state. But, some things below the AI level — some of the questions Copeland suggests — may be worth asking. Looking at an all-electric car and you may decide you should stay in a pre-electric car state. So you move to a hybrid, however, because it will help.

As Copeland’s first question implies, ask what you are trying to do and you may find better solutions than hyped tech. AI may sound like a cool way to do something. Process improvement coupled with automation may get you to a solved problem faster and at lower cost. Process improvement and simple tools may bring higher rewards than AI can bring in a narrow area of expertise. Getting scared by AI may cause you to ask the questions you should have asked.

My third argument is tech awareness. Most lawyers are to tech savvy as Neanderthal Man is to Elon Musk. AI hype may cause some lawyers to realize that tech ignorance lacks the cachet it once had among the client elite. If an outpouring of social media venom can humble the CEO of a major company within hours. If new tech products can obsolete businesses within a decade. And, if some of the most respected scientists of our time think tech has the power to transform and extinguish our society. Perhaps it is time to check out this tech thing.

Bad Things Can Lead To Good Things

General counsel face a strange battle within corporations. The way to avoid some of the most significant legal costs a corporation may face is to engage in preventive law. To succeed with preventive law, one must appreciate the risks of failure. Corporate leaders who have avoided the costs and pain of major, existence-threatening lawsuits, may lack respect for failure. They underestimate the risk. That inhibits them from supporting spending on preventive law. Many general counsel have wished in their heads for a devastating lawsuit. Nothing like a burning platform to get the message across.

The legal industry faces a similar challenge. We may see climate change re-shaping the world. We may hear all the experts telling us that unless we act, we will lose the opportunity to act in the future. Lawyers have resisted. It was easier to throw labor at a problem than to move to tech. My vegetable garden does fine and in fact does a bit better as temperatures warm in my zone. I can let fixing the climate (or legal industry) be some other person’s problem.

Hype has many downsides, but it has some upsides. Getting those lawyers who firmly believe tech is a fad engaged in the future could be a big upside. If over-hype means a few lawyers get scared into asking the right questions, I can live with the over-hype.

BespokeThurgood Marshall was not just a justice on the U.S. Supreme Court, he was the lawyer that led many of the civil rights cases striking down de jure racism in the United States. While he successfully prosecuted many cases, he is perhaps best known for the Brown v. Board of Education landmark decision in 1954. The Supreme Court unanimously ruled that “separate educational facilities are inherently unequal” and, therefore, racial segregation of public schools violated the equal protection clause of the 14th Amendment.

Justice Marshall’s strategy to bring down racist practices in the United States is still studied today as a model for how to run an activist program through the court system. But today’s court system is far more complex, and the contexts in which activists bring their challenges far more convoluted, than at any time in the past. If a few years from now you were faced with a challenge of the type that Justice Marshall faced, could you run a similarly successful campaign?

The Maker Era

We have entered the maker era. You can buy a 3D printer, put it in your basement, and within minutes you can make or replicate replacement parts, toys, or your own inventions. Connect your computer and software, and now you are in the business of designing whatever you want. Once you are satisfied with the design, you can send the file to a commercial manufacturer or license it over the web. You are in competition with the global world of manufacturing.

What if you are a nascent clothing designer? New nanotechnologies allow you to spray on clothing. You can add colors, create texture and design combinations, and even build in special features (anti-perspirant?) into the new shirt. If you don’t like what you created, simply peel it off, dissolve it, and start over again.

If you aren’t interested in manufacturing or clothing, perhaps you would like to use software to mine behavior? Go to the internet and you can download software that lets you analyze text, apply machine learning, and find secrets hidden in the words. You can analyze sentiments, build behavioral paradigms, and test your ideas without leaving the comfort of your office.

Lawyers often struggle to comprehend how fast the world is moving around them. 3D printing, nanotechnology, and artificial intelligence are just three of the many areas rapidly scaling from the laboratory to real life. Yet most lawyers are still in the quill and paper era, trying to master Word and occasionally venturing into Excel or PowerPoint.

Most law and technology posts focus on simple automation or perhaps elementary artificial intelligence applications, such as finding all bankruptcy cases with certain elements. The “scary” futurists speculate about robotic lawyers sitting next to human lawyers as they compete for the next case that comes over the transom.

In this essay, I’m going to take you on a bit of a journey into the future. Let’s imagine how a clever software engineer with knowledge of the law and a lawyer with a bent for social justice could start a movement in the 21st century.

The Designer’s Lawsuit

Thurgood Marshall faced a significant challenge in the 1930s, 1940s and 1950s when he was attacking racism as a lawyer for the NAACP. He had to maneuver key issues through federal district courts and appellate courts, often hostile, to get those issues before the U.S. Supreme Court. To do so, he relied on research, experience, intuition, and luck. Could he do it differently in the 21st century?

Imagine using computational linguistics and machine learning to look at all the reported opinions relevant to the issues you wanted to get before the U.S. Supreme Court. The software “reads” the opinions and the briefs, finding those obscure connections you couldn’t find even you if you had the time to read the hundreds of thousands of pages.

To make the analysis more interesting, you also look at databases built from the biographies of the judges who may sit on the case at each federal court level. The databases include all the information about the judge’s experience (undergraduate school and major, law school, etc.), training (law firms, government positions, etc.), and personal characteristics (age, gender, etc.). It also includes everything the judge has written or said that is public, outside of opinions. Speeches, law review articles, and op-ed pieces sit in the database.

While it may seem like you have the relevant information you need, you don’t stop there. You pull together information on the community where each judge lives. What is the political sentiment within the community? Is it affluent? What religions predominate? You dig deep into the community to understand how it may affect the judge’s thinking.

You also look at the national climate. Where have the trends been going—are the people in the United States moving in your favor or against you on the issues? Are there similar issues from which you can gain guidance? What about legislative movements at the local or national level?

With this massive database, you turn loose your machine learning software again on a small subset, the “training” database. The algorithms (and you use many, stacked to mimic the way the human brain processes information—as best we can tell) run through the information over and over again. The algorithms are learning, attempting to “understand” the data. You ask questions and the algorithms respond.

As the algorithms respond you check the results. You reject most responses and inform the software of its hits and misses. Over time, the hits grow and the misses shrink. The algorithms seem to be organizing information the way an expert might. You reach the point where you think the algorithms are ready for the big time.

You turn them loose on your large data set and wait to see what happens. As the results start to come in, you see that the algorithms have identified some court possibilities you would have picked, but there are some unexpected choices as well. In fact, it turns out the unexpected choices rank higher on the probability of success than your expert choices.

That was the easy part. The next step is to ask the software to design the lawsuit. What arguments will work best in each court? Which arguments should be emphasized and which added as “just in case.” Again, you find some of the picks familiar, but some are unexpected. In fact, there are a few creative uses of arguments that came out of cases decided long ago, but that seem to fit with the times.

With knowledge of the arguments, you turn the work over to the next program: the legal argument drafting program. For many years now, software has been writing corporate earnings articles and recaps of sports games without human intervention. The software takes the financial results or the game record and, using some training on writing styles (a dash of Hemingway mixed with a bit of Lardner), turns out articles that people can’t distinguish from articles written by journalists.

That software has now been trained on legal writing styles. Using opinions by Holmes, Jackson, Hand, and Posner, you have trained the software to write like a judge (or, perhaps, like some judges wish they could write). With the legal arguments preferred by the machine learning algorithms, the legal argument drafting program turns out a passable first cut at a motion for summary judgment. You can use that to back into a complaint. Once you know the district judge and she issues her opinion, you can use the software again to churn out the first draft of your appellate brief. You have designed your first lawsuit. (And before you yell at me about all the professional responsibility issues involved, please remember the spirit in which this essay is intended – to spark interest in what types of analyses will be possible, not to suggest ways to improperly create and file claims.)

Don’t Wait, Dive In

The designer lawsuit is, of course, a thing of the future. The software we have today can do bits and pieces of what I have described, but we still have a way to go before all of those pieces can knit together a new case tweaked for what may work best in each court. Still, we are closer than what many lawyers think.

The point of this exercise is not to scare lawyers into thinking software will replace them soon. It also isn’t to add on to the fatalist pile the thought that lawyers will soon become extinct, or nothing more than the handmaidens of computers.

I do hope the story has piqued your interest in staying current with what software can do. The amount of data available to lawyers is vast and far beyond what we can reasonably consume and use to help our clients. It grows much larger each day. As tools come online that can help us digest that mass, to not use them approaches the irresponsible. That information contains judicially recognizable information that may tip the balance in an argument. It puts judicial decisions in the context of what is happening in society (and if you still think judicial decisions aren’t political or are limited to extending law on the books, then I apologize for rudely dragging you into the real world).

We have entered the augmented age, where humans plus computers can take us further, faster, than humans or computers can go alone. If you believe that is fantasy, think about what you hold in your hand: a smartphone you can talk to that reaches out to all the data on the internet to answer your question. In seconds, Siri or Cortana translates your spoken request to digital commands, processes them, and come back with an answer or at least relevant web sites. Your mind has been augmented, through the smartphone, by the internet. That connection grows closer and stronger each day

When will we have designer lawsuits? Five years? Ten years? Longer? We don’t know and the correct answer isn’t relevant. The designer lawsuit will not be like falling off a cliff: one moment you do all the work and the next the computer does it for you. We will creep closer a step at a time, with the steps coming quickly at some points. Whenever the time comes, you will be much better off for having kept pace with the changes, than trying to quickly run to catch up. No matter how fast you are, you will not succeed.


“I actually think that law enforcement should be difficult,” Marlinspike says, looking calmly out at the crowd. “And I think it should actually be possible to break the law.”

Moxie Marlinspike

Moxie made that statement while sitting on a distinguished panel of cybersecurity experts at RSA, the main conference on computer security. Meanwhile, the legal industry was pre-occupied with how much to bill or pay for legal services and whether knowing almost nothing about computers is okay for lawyers. This juxtaposition plays out every day as the legal profession ignores the issues that concern society and focuses more and more on a shrinking set of issues that concern only lawyers.

Cybersecurity is Big Time

This past week, one of the big news stories was the hack of the Democratic National Committee’s email system. This was the latest in a now long string of hacks that have raised cybersecurity to one of the biggest issues on many CEOs agendas. The United States government believes Russia was behind the DNC hack, but regardless of the perpetrator, that fact that it happened and affected one core aspect of a democratic nation—the presidential election process—brought cybersecurity into full view, again.

Moxie Marlinspike is one of the good guys fighting the cybersecurity battle. Our vision of what a good guy looks like is closer to Moxie’s peers. We think of military officers wearing uniforms filled with medals, serious looking men and women in dignified corporate uniforms, and academic types who lean towards the corporate style. Then there is Moxie. At 6’2” and rail thin, he stands out (literally) among the crowd. Top that off with a head full of dreadlocks, and you have a guy who does not give off at first glance the good guy vibe.

Moxie is on the panel where he made his “break the law” statement because he created Signal, a free encrypted messaging and voice-calling app which most consider the easiest and most secure to use. In fact, it is so easy and so good, former law enforcement officials recommend it and computer scientists drool over it (literally).

Moxie made his statement because he is highlighting a deeper question: do we really want a world where law enforcement can know everything? This question has become one of the hotly debated topics of our time. If you don’t think so, just recall Edward Snowden, Apple’s battle with the FBI, and closer to home the FBI’s warning to all large law firms that they are being or have been hacked.

In a world where we are trying to connect everything, we can be certain the bad guys are trying to invade. To what extent will we give up our privacy to fight the bad guys? Each of us has an answer. All of us should agree the question is a serious one deserving plenty of attention. In fact, “The World Economic Forum…listed cybersecurity as one of the greatest threats to businesses globally.” But, the troubling question is why lawyers are more focused on their own issues than client issues?

Climb Into the Lawyer Foxhole

The cybersecurity talk is interesting, but you may ask what use is it to lawyers to know a dreadlocked cybersecurity expert thinks it should be possible to break the law? The answer lies in the role you think lawyers should play in the evolution of our digital society. It appears that most lawyers think turning a blind eye to the issues and problems is the best approach. But if CEOs think cybersecurity is a major issue, shouldn’t their lawyers take a more active interest in it?

You may say I do real estate / benefits / tax / antitrust / environmental / [fill in the blank] law, and that is an issue for cybersecurity lawyers, so leave me alone I don’t have time to climb out of my foxhole and look around. In the most narrow, “I only do one thing” sense, you may slide by with that response. But in the broader sense of lawyers serving clients’ interests, it is hogwash.

Lawyers Need to get in the Tech Game

We can go back to Moxie and Signal. Start with Model Rule of Professional Responsibility 1.1 “…a lawyer should keep abreast of changes in the law and its practice, including the benefits and risks associated with technology…” Then look at Rule 1.6(c) “A lawyer shall make reasonable efforts to prevent the inadvertent or unauthorized disclosure of, or unauthorized access to, information relating to the representation of a client.”

You may be one of the lawyers complaining about these requirements. You have to stay current on the law and now you have to stay current on technology? On top of that, you have to market, bill, and manage, so where do you get the time to become a technology guru? If that is what you said, then you have company in Virginia. The Supreme Court of Virginia recently considered adding the Model Rule requirements to the state’s rules, and received comments in opposition voicing those same thoughts. The Court, wisely, added the ABA provisions to the state’s rules.

Before I go further, let’s debunk the straw man arguments. The Rules do not require lawyers to become IT gurus, just that they get in the game. The lawyer who thinks he can operate off the technology grid has a fool for a client. The CEO of a major company does not know how every aspect of her business works. But, she also better not answer an analyst’s question about a key area of the business with “I never did mind about the little things.” (Extra points if you know the movie from which I took the quote, without using Google.) There is a noticeable difference between tech guru and tech familiar.

Now let’s look at Signal. Signal is distributed by Open Whisper Systems. As the name implies, it is an open source product. For those not into the lingo, open source means the developer makes the computer code freely available. In most cases, they go further and ask the community to contribute improvements to the code. (Stop me if this sounds like the antithesis of lawyering.) Open source isn’t a fringe area. Linux is one of the more commonly used computer platforms (don’t be surprised if your law firm uses it) and it is open sourced.

You install Signal on your smartphone and it works on both iPhone and Android operating systems. It works with calls and text messages (but you must have an internet connection). Signal says it provides the following benefits:

  • “Send high-quality group, text, picture, and video messages, all without SMS and MMS fees.”
  • “Use your existing phone number and address book.”
  • “We cannot read your messages, and no one else can either.”
  • “Pay nothing.”
  • “Make crystal-clear phone calls to people who live across town, or across the ocean, with no long-distance charges.”

Signal seems to have a fairly strong value proposition, putting aside for the moment whether Edward Snowden and Laura Poitras endorsing it is a plus or minus for you. When we compare what Signal can do to what Rules 1.1 and 1.6(c) say, you should at least ask whether it is “reasonable” to not use such free encryption software. We can do a quick comparison of the reasonable test in the Comments to Rule 1.6 and what Signal offers:

I am not pushing Signal, I am merely using it as a convenient example (and yes, I recognize Signal has some limitations). Before you run out and install Signal, make sure you ask your IT department whether it will work with any systems they have installed or connected to your phone. But, recognize that the National Lawyers Guild uses it and Facebook has installed the Signal protocol in its Messenger system. As they say, what is good for over 1 billion people can’t be bad for everyone. Or, as Marlinspike says, “The big win for us is when a billion people are using WhatsApp and they don’t even know it’s encrypted.”

So on the one hand we have Signal, a freely available product being built into the most widely used social media tool, and on the other hand we have lawyers who should be taking reasonable efforts to protect client confidentiality. Most of those lawyers use their smartphones to communicate with clients, the lawyers at large law firms know they are targets of hackers, and yet we can surmise that most lawyers don’t think about whether their communications with clients are protected. Notice that the Rules don’t focus on whether the lawyer practices real estate / benefits / tax / antitrust / environmental / or [fill in the blank] law.

Mind the Big Things

If the goal of this story was simply to chastise lawyers for not being more careful with confidential information, we could stop here. But I have bigger goals in mind. Let’s start with lawyers’ obsession with revenue and profits. I use the word “obsession” because the focus goes far beyond building healthy businesses. For lawyers, it is the beginning, middle, and end of each day (I have actually heard lawyers greet each other in the hallway with “how are the billings?”). Run a Google search on “lawyer and revenue” and you get 36,900,000 hits in an impressive 0.38 seconds and for “lawyer and profit” you get 73,600,000 hits in 0.53 seconds.

That obsession is pushing out time for anything else. Moxie created a system that we can easily imagine bad guys using to help them break the law. Moxie then sat on a panel and said he thinks being able to break the law is a good thing. I think Moxie is raising a point we should discuss and it is healthy to debate privacy versus law enforcement access. But there wasn’t a lawyer on the panel to engage in that debate. It was technologists debating among themselves, and that is not healthy.

Lawyers are ceding their role in society by building a wall around themselves. That wall has a sign on it that says “don’t come talk to us unless you bring money.” Sure, lawyers need to make a living. But part of making that living is serving a broader purpose, just as corporations have learned that part of operating in society involves more than focusing solely on profits.

Clients want to do business with engaged lawyers, just as customers want to do business with engaged corporations. It wasn’t Moxie’s fault that no lawyer was there to respond. It wasn’t even the fault of the RSA conference organizers. Lawyers have created the perception that they are uninterested and irrelevant by training others to believe they don’t care about technology or really anything except the economics of the practice.

Today, Moxie is an industry insider, though he describes himself as an anarchist living on the outside for a long time. He led Twitter’s security team, has the deal with Facebook to use Signal in Messenger, and has created the “largest end-to-end encrypted communications network in history.” He also isn’t a radical, at least in the sense that he isn’t the wild-eyed one we all fear. His comment about breaking the law ties to his bigger theme that “privacy allows people to experiment with lawbreaking as a precursor for social progress.” The debate about encryption software isn’t about software, it is about larger societal values and where breaking the law fits in, something all lawyers should find interesting.

You don’t need to be a cybersecurity lawyer to ask whether you are doing what is reasonable to protect client confidentiality. You don’t need to be a computer expert to care about the issues that are top-of-mind for your clients. You do need to be a lawyer concerned about more than the next six minutes of your time. Lawyers need to do a lot to climb out of the deep foxhole they are digging for themselves. The first step in job security is having a skill that others need. Clients don’t seem to need people whose primary skill is counting the money they get paid.

GoldRushOn January 24, 1848, John Wilson Marshall was building a water-powered sawmill for John Sutter. The mill was close to Coloma, California, near the base of the Sierra Nevada mountains. Marshall was a carpenter who had emigrated from New Jersey. Although Marshall wasn’t looking for gold, he later claimed that he knew immediately upon seeing the flakes what he had found.

At the time Marshall found the flakes, Mexico and the United States were still at war over the California territory. The population of the territory was mostly Native Americans (around 150,000) with some 6,500 Californos (people of Mexican or Spanish descent) and about 700 others (mostly from the United States).  A few days after the discovery (and not because of it) the United States and Mexico signed the Treaty of Guadalupe Hidalgo which ended the Mexican-American War and left California to the United States.

Sutter and Marshall tried to keep the discovery confidential, but obviously did not succeed. Within two months a newspaper was reporting on the discovery. By mid-June, three-quarters of the male population of San Francisco had left to goldmine, and by August the number of miners in the Sutter Mill area had reached 4,000.

The next year was written into U.S. History as The Gold Rush of ’49. The non-native population of California grew from 700 before the discovery to 100,000 at the end of 1849. By September 1850, California had become a state. The Rush peaked in 1852 when approximately $81 million (in 1852 dollars) of gold was pulled from the ground. By 1857, the annual production had dropped to around $45 million where it stayed for many years. The Rush was one of the most important events in re-shaping the face of the United States.

The 19th century in the U.S. was known for gold, but the 20th century was marked by hydrocarbons. While some believe the 21st century’s gold, especially later in the century, may be water, the current rush focuses on data.

The Strange Stories of Law

You have heard the statistic: each day more data is generated and stored than the amount of data that existed in all of history prior to the computer age. Large companies that entered the world as retailers, search engines, or social media companies found the real value of their businesses was in data. In Silicon Valley, it almost became irrelevant what your business could do, the focus was on the data set it could build.

And then we come to the legal industry. We can tell two versions of the legal industry story. The first story goes like this:

Recognizing the threat cybersecurity breaches present to their clients, law firms decide to thwart the attackers using an unusual approach. They accepted the futility of keeping hackers out of their systems. Instead of following the norm of keeping information as accessible data, which can be indexed, accessed, and manipulated, law firms keep their information somewhat like teenage boys keep their rooms. As one law firm leader said, “We decided that if our data was a mess and even we, who know it best, have difficulty finding and doing anything with it, hackers would have more trouble and simply give up.”

A team of red hat associates was tasked with hacking the system to find a group of documents to use as templates when drafting for a client. The blue hat defense team’s strategy was simple. “We pretended we were partners and randomly withheld helpful information from the red hat team.” The red hat team gave up after many hours and decided to draft from scratch.

The second story goes like this:

While clients and the world around them screamed about data, lawyers continued their quest to be oblivious. Lawyers in firms, corporations, and other service organizations knew that if they hadn’t enjoyed knowledge management, they would enjoy data management even less. Again adopting the “to do nothing is to do something” approach, lawyers have ignored pleas to treat their documents as data gold.

When asked about this strategy, a lawyer responded “The world around us has been changing for decades and yet here we sit today, almost unchanged. To respond to this ‘data fad’ by doing something would go against our strongly held belief that all tasks should be done by lawyers and not other service providers, even computers. Indeed, we are considering asking bar associations to file actions against all computer companies for the unauthorized practice of law.”

Choose which version you prefer, but the reality is that lawyers in firms, departments, and other legal service provider organizations are in the same boat. Legal data is not created and stored as the precious commodity it is.

The Stories Data Could Tell

The work that lawyers do tells stories of risks and responses. What gave rise to the lawsuits? What did the parties do? What steps were taken in the ligation? How long did they take? What were the responses? We can explore similar questions in any area of law, and it is those questions and responses which are embedded in what law firms store on their servers.

The challenge with most data sets, unlike those in law, is not getting to them. Finding data sets can be easy. The challenge is getting them in shape to use. Data scientists call this step data wrangling or data munging and it eats up 80% of their time.

Think about a data set in your firm that you actually keep as a data set: your customer information. Your firm or law department has a system for keeping track of customer (in the case of law departments, law firm) information. If you check the system, you will find out of date entries, missing information, duplicate entries, and incorrect entries. Imagine how long it would take if you froze the system today and had someone focus on cleaning up the database. Of course, as soon as they finished and you resumed using the database, you would find it out of date.

Now apply those problems to your real data. All of the documents sitting on your servers form your database. You may have a knowledge management system, and still your data is not ready for use. At best, you have a collection of documents with some perfunctory information filled in by field. Your knowledge management system uses a not-too-sophisticated search process to locate documents responsive to your request. When you find them, you can’t do much with them except copy and use them as templates. Definitely not state-of-the-art.

When I talk about data, I mean the ability to access specific information from those files, combine it with other data, and produce information that will help solve client problems. For example, what if you could combine data from all the employment lawsuits you have handled with data from government and court data sets. Could you construct a model that gives specific information about each type of employment lawsuit?

You may think of this as fantasy, but it isn’t. Today, startups have breached the barrier and are applying this type of analytics and more as they find and use data sets. One small but growing area is computational linguistics. Put very simply, CL applies statistical tools to text. Through machine learning, computers can use the CL tools to understand text far beyond “supreme w/5 court.” Tools using CL in law are in the early stages, but they all face the same challenge: getting access to clean data sets.

This is where lawyers enter the picture. By recognizing today that the information built into the data sets is the gold that will help law firms and law departments protect clients, lawyers take the first step. The second, is to start transforming what already is in the sets into data, and the third is to store whatever new items are created as data.

If You Make It The Bad Guys Will Come

At this point, a good question to ask is what about the cybersecurity threat? As they say, there are two types of companies: those that have been hacked and those that don’t think they have been hacked. The experts with whom I have talked agree that law firms are and will continue being hacked. The firms just do not have the sophistication to prevent the hacks. That is not a slam against law firms. It is hard to find any organization, and so far no one has named one, that is immune to hacks.

So if the hacks will happen, why should lawyers turn what they have into data? My first scenario above was written in jest, but lawyers do ask if it isn’t better to have the hackers find the messy teenager’s room than a nice, neat library?

The response to hacking isn’t to abandon the quest for data, just like the response to computers isn’t to become a modern-day luddite. All firms and corporations should take reasonable steps (and today more are going well past reasonable) to protect against hacks. Assume there will be hacks and focus on the data. Just because a hacker can get into a system doesn’t mean the hacker can get access to, un-encrypt, and assemble all the data in a way that will help them. You have a security alarm on your house, but you don’t leave all of your valuables lying on the kitchen counter. Thieves still take gold, but we still mine it. Cybersecurity is a challenge, not a bar, to keeping and accessing data.

Data Is Becoming Essential

Data is going to be more than a way to use and manipulate what you create and store. It will become an essential part of the modern law practice. Let’s look at one last example: blockchains. I won’t go into a detailed description of blockchains, I’ll keep it simple. A blockchain is a database that is distributed, not centralized. Each record in the chain may hold data, a program, or both. The records are hardened against tampering through strong encryption and distribution. Blockchains reduce and sometimes eliminate the need for intermediaries.

The terms of a smart contract are built into the code embedded in the blockchain. If condition X happens, then Y occurs. No ambiguity, no equity (at least that is the theory). Once the contract is formed and built into the blockchain, no on can alter the blockchain (more precisely, an altered blockchain becomes an instantly visible anomaly rejected by blockchain holders).

Lawyers who do not understand blockchain, code, computers, or how the system should work will be at a severe disadvantage. Yet big banks and other large players are actively looking into using blockchain or similar technologies as part of their systems. Since the contract is in the code, we can treat the contract as data and start combining and manipulating it.

Mine the Data Now

Lawyers have believed for centuries that they need to study the law, but they can pick up everything else quickly so that they can apply the law to it. Litigators are famous for believing they can litigate an employment case in the pharmaceutical industry this week and an antitrust case in the retail industry next week. Large firms have moved beyond this by making everyone specialize (and sub-specialize), but the feeling still exists. So, lawyers wait and watch. When they think something has become so well established that the world can’t possible go back, lawyers make their move.

While lawyers may believe they can wait until everyone is deep into data and then put their toes in the water, it doesn’t work that way. I mentioned at the outset competitors in the retail, search engine, and social media industries. They have built data sets so large and deep that it is unlikely anyone can catch them. In fact, recognizing that the prize is data and not tools, Silicon Valley has embraced a new trend. These companies are posting on the Internet for anyone to use many of the most sophisticated tools they have developed.

Why would they open source the tools? Because these companies know that the tools are useful and by open sourcing them they may get interesting insights from others who use them. Making the tools available allows the scientists who developed them to showcase their work, an important part of attracting and keeping talent. But these companies also know that without their incredible data sets, others will not be able to use the tools to replicate what these companies do. The tools help, but the data sets are essential.

Law firms and law departments have yet to realize that tools are becoming widely available. The firms and departments will need help, from academia, consultants, and others, to understand and employ the tools. But, the tools will not be the chokepoint. The real value is in the data. Each firm and each department has value in its proprietary data. To realize that value, they must start treating it as gold and not as dirt. Welcome to the 21st century.

ModularityLawyers think about things in somewhat discrete units called matters. No one has a formal definition of matter we all must use, but we all know them when we see them. A lawsuit is a matter, a contract is a matter, a policy is a matter.

We have, of course, managed to confuse this neat world by using “matter” indiscriminately at times. If you do a lot of over-the-phone counseling for a client, your system may show “Grady Counseling” as a matter. We also collect things, like all the work necessary for company A to buy company B, into a matter we call “A Acquisition of B.” This matter includes many contracts and other documents, counseling activities, and other tasks.

When we talk about matters which include multiple tasks—that is, when we refer to matters as collections of tasks rather than discrete tasks—we create the possibility and usually the reality of a networked system. “A Acquisition of B” includes contracts and other documents, some of which are based on templates taken from “X Acquisition of Y.” These matters are now linked through a common document template, the asset purchase agreement template.

Active law practices have lots of networking, or linking, because templates get used often. I may have a standard form or motion I use in litigation, a basic form of will, or a standard ERISA 401k plan document. The more often I use it, the more networking affects my system even if I modify it for each particular use.

We can think of each template as a module that gets used and re-used in our network. That module addresses one thing. But, we can plug the module into many situations which need the one thing. Using the module stops us from re-inventing the wheel each time we need whatever the module covers. Using the example above, you would not want to draft an asset purchase agreement from scratch each time you had a client making an asset purchase. You use a template (the module) which allows you to customize the basic template.

The Modular Office

Although programmers have used modules for a long time, certain of the key programs most lawyers use are not modularized when viewed from the user’s perspective. The most famous programs comprise the Microsoft Office Suite—Word, Excel, PowerPoint, Outlook, Access, and OneNote. Office (in one of its many forms) has about 1.2 billion users so anything involving Office could have a massive impact.

Microsoft has explained many times that the organization’s focus is moving toward the cloud and mobile computing. Neither is a surprise as these are overwhelming trends in the tech industry. The interesting note here is that Microsoft also plans to modularize Office. What does this mean?

Start with how you work today. If you want to write a document, you open Word. If you want to crunch numbers, you open Excel. And if you want to tell a story, you open PowerPoint. Behind the scenes, these and the other Office applications share data and processing modules, but to the user they are three separate programs. With the evolution of mobile computing, things get a bit more complicated because some of these applications are easier to use in the desktop world and some work just fine on a smartphone.

Microsoft’s vision involves looking at the “problem” from the user’s perspective and re-defining how the applications work based on what the user wants to do. One example given by the executive vice president of Microsoft’s Applications and Services Group is the post-meeting distribution task. After a meeting, you need to circulate the notes and PowerPoint deck from the meeting to the attendees. Doing this task today would involve using parts of different applications. Tomorrow, in Microsoft’s view, you could simply ask (orally) your computer to do the task. It would know the meeting, the PowerPoint, and the notes and send an email (via Outlook) of the relevant documents to the meeting participants.

Modularity and the Augmented Lawyer

Microsoft’s focus on the modular Office product takes us in the direction of what I call the “augmented lawyer.” This lawyer combines human skills and computer capabilities to deliver solutions to client problems. Augmented lawyers look for ways to combine the best of what computers can do and the best of what humans can do to find higher quality, lower cost, and more timely solutions to client problems.

The augmented lawyer could use the new form of Office to accomplish many tasks faster and with better quality. Assembling a motion for summary judgment might happen when the lawyer asks the computer to assemble the various parts into an e-filing document. Quality increases, because each time we take the human side down and bring the computer side up we have an opportunity to reduce mistakes (when we don’t get better quality, we typically had a process design issue not a computer problem).

Modularity is another way of talking about disaggregation. At the macro level, we can disaggregate projects into tasks and operations. At the next level, we can disaggregate tasks and operations into components done by humans and ones done by computers. As we disaggregate and automate (again, putting aside our mistakes in re-designing the process), we make improvements. Each improvement may seem small, but over the course of days, weeks, and months these small improvements can mean the difference between a viable practice and one that is too inefficient to survive.

Modularity is Coming to Legal Services

Lawyers who have not already done so need to think about modularity in their practices. Having lawyers in a firm or law department continuously repeating what others have done does not add value. When several lawyers, each sitting in his or her own office, review and revise contract terms that that have been beaten to death by generations of lawyers, clients get poorer and lawyers get richer but value is not created. Lawyers’ desire for autonomy needs to become subservient to clients’ desire for for improvement.

One of the early hallmarks of becoming an effective augmented lawyer will be adopting the modularity concept. Lawyers, firms, and departments that do so will see significant efficiencies and quality improvements, and most likely many other benefits. For those who move first, it will give them many opportunities and a significant lead over their competitors. The danger of being the first-mover and choosing Betamax over VHS exists, but only for those who act by tying themselves to an inflexible structure. Another hallmark of the successful augmented lawyer will be avoiding the urge to become inflexible.

Microsoft’s vision is one example out of many about where software is headed. Law firms and departments have tended for many years to prefer enterprise systems or network systems that, once installed, are difficult to adapt to rapidly changing worlds (and expensive). While the future always is murky, focusing on modularity in both computer systems and legal practice design will enable lawyers, firms, and departments to move quickly and focus on client needs rather than face the titanic task of changing course every year (or less) in the new, competitive, legal world.

LawGameIn part I of this two-part series, I covered the history of computers versus humans playing perfect information board games. In part 2, I talk about what lawyers should take from Lee Sedol’s recent loss to AlphaGo in a five-game Go match. Given the length of the series, I have cross-posted the entire piece on “The Algorithmic Society.”

When Thinking About Automation, Think Tasks Not Jobs

While no one is suggesting that board games are one step away from practicing law, AlphaGo’s significant step forward from Deep Blue and Watson suggests why lawyers must embrace a different future. I call this future the “augmented lawyer.” Lawyers leverage the growing power of computers by using them to handle “power and volume” tasks while lawyers contribute skills computers have not mastered.

Artificial intelligence is not an “either or” situation, often phrased as either the computer can do the lawyer’s job or it can’t. Lawyers’ jobs are composed of thousands of tasks. Some tasks are extremely complex, but many are very simple. Most are mixed. Legal service processes can be disaggregated, for example, during the lean thinking exercise of process mapping. Once a process is disaggregated, the question is whether computers can automate a task or even part of a task. AlphaGo’s win shows there are many tasks lawyer do that computers can learn.

We already have seen computers tackle legal tasks that require significant power or involve lots of data. Consider Shepardizing a case. Shepardizing is a term that dates back to before the modern computer era and comes from a series of books published as Shepard’s Citations (named after the original author, Frank Shepard). Shepard’s books listed all of the published case decisions by federal and state courts. They were arranged by citation. Underneath each case citation heading, Shepard’s listed subsequent cases, that is, cases referencing the case cited in the heading. Shepard’s used symbols to indicate why each subsequent case cited the heading case (overturned, reaffirmed, questioned, cited, etc.).

Lawyers want to know whether the cases they present to judges in briefs are still good law. To Shepardize using books, the lawyer would look up the main citation in the most recent bound Shepard’s volume. Then, she would move to unbound volumes covering each month since the last bound volume. Then, she would move to small booklets covering the weeks since the last monthly volume. She had to repeat this process for each case cited in her brief.

Once she knew all the cases citing a case in her brief, she had to decide which ones to review. If Shepard’s listed five cases citing her case, she might not need to read all five. But if Shepard’s indicated any of the subsequent cases had questioned, limited, or overturned her main case she had to read the subsequent cases. That meant going to the library, tracking down the book with the published case and either reading it right there or photocopying it for later reading.  If there was a significant lag between finishing the Shepardizing process and filing the brief, she had to check each case against the latest small booklets. Junior associates had the pleasure of Shepardizing and it took quite a lot of time and client money.

The Shepardizing process is quite different today. Software automatically checks each citation in a brief against databases of published decisions. Within seconds, the lawyer has a list of all cases citing any of the cases in her brief. By clicking on a link, she can go directly to a citing case and the place in the case where the citation occurs. The tasks of finding the citing cases and tracking down the published decisions have been automated, significantly reducing the time spent on those tasks. If there is a lag between Shepardizing and filing, updating means clicking the button again and checking any new cases. Lawyers still must read the citing decisions and draw conclusions about whether those cases have any impact on the case they cited. Large Parts of Legal Service Processes Already Can Be Automated

Most legal service processes involve this combination of tasks amenable to automation and tasks that lawyers still must do. The question is not whether the computer can replace the lawyer, but which tasks the computer can do more quickly, less costly, and with higher quality than the lawyer. The number of tasks where the computer excels is growing, but lawyers resist change and are holding on to those tasks. That resistance increases costs to clients while also reducing quality.

The false “either or” dichotomy masks a continuum. By using process mapping, continuous improvement, and workflow automation reviews, lawyers can construct a pipeline where processes have waste removed and tasks move to automation when the timing and cost make sense. Instead of fancy, complicated software systems lawyers can use simple, low cost and easy to modify automation tools.

A recent article by Michael Chui, James Manyika and Mehdi Miremadi in the Harvard Business Review made estimates about the percentages of daily tasks knowledge workers perform that could be automated. Thirty percent of the tasks performed by knowledge workers can be replace with current automation. Add to that another 20% reduction in tasks from taking waste out of processes (waste reduction estimates range from 20% to 95%, so using 20% is conservative). That means lawyers should be able to get a 50% reduction in the tasks done by attorneys be eliminating or automating them without significant spending or overwhelming time investment.

AlphaGo’s Victory May Be A Tipping Point

AlphaGo’s victory does not mean a defeat for lawyers. But in three ways it is a warning.

First, given that only a year ago experts were predicting it would take a computer another decade to beat a human at Go, it shows the pace of machine learning development is faster than anticipated. The more lawyers resist automation, the farther behind they fall. The faster computers move up the machine learning curve, the greater the gap between what computers can do and how lawyers use them. This will accentuate the difficulty of becoming an augmented lawyer.

Second, lawyers base many of their arguments against using computers in legal services delivery on the distinction between rote tasks (find all documents with the word “discriminate”) and legal thinking. AlphaGo’s use of something akin to intuition shows that, at a minimum, computers are capable of more in law than their critics had claimed. In truth, this already was the case but the AlphaGo win makes it more apparent.

Third, lawyers argue that the cost of training a computer system outweighs the benefits received from the trained system when it comes to the legal tasks that system could do. Hassabis’ statement that AlphaGo will be able to train itself how to play Go and defeat a human player in less than a year shows that the upfront burden of bringing computers into legal services will soon be, or already is, dropping.

Today, anyone with a Macbook, a basic knowledge of Python programming, and access to a data set also can turn loose machine learning on a problem. Many machine learning programs are readily available. Google is even making some of its work available (see TensorFlow, an open source Machine Learning system, now expanded by Cloud Machine Learning “a framework for building and training custom models to be used in intelligent applications”). While good data sets are hard to come by in the law, they are becoming more accessible every day. Soon, we will have all reported U.S. caselaw available to scholars and not long afterward, to everyone. With greater dissemination of the tools and more access to data sets, the power of computers to automate legal tasks will grow and at a faster rate than it has grown before.

No one should minimize the difficulty of jumping from what AlphaGo did to understanding and manipulating language and higher level legal thinking. But assuming some barrier exists is foolhardy. The question is not if, but when. Lawyers must take heed and fully engage in understanding how to work with computers more effectively as part of an augmented lawyer practice. It’s time to go.

LawGameThis is part I of a two-part series talking about what lawyers should take from Lee Sedol’s recent loss to AlphaGo in a five-game Go match. Given the length of the series, I have cross-posted the entire piece on “The Algorithmic Society.”

A shudder of excitement went through the tech world recently and its epicenter was Seoul, South Korea. There, a computer named AlphaGo played five games of Go against Lee Sedol, a South Korean master of the game ranked fourth in the world. AlphaGo won four out of the five games. Long considered a difficult and perhaps impossible task, a computer winning at Go suggests that computers are moving closer to taking over some human tasks much sooner than we imagined. It also was a strong volley by Google to be the company whose algorithms drive the “thinking” behind the takeover.

Yet, of the approximately 1.25 million lawyers in the United States, it is safe to say that few read about and understood the significance of the victory, some saw the headlines but skipped the stories, and many did not even know the match took place. AlphaGo’s win over Sedol will be one of those moments lawyers will look back on and see as another tipping point they missed. The event that shook the technology world caused barely a tremor in the legal world.

What is Go

To understand the significance of AlphaGo’s win, you must understand something about Go. The game originated in China and dates back at least 2500 years. It was considered one of the four “essential arts” of a cultured gentlemen (the other three were calligraphy, painting, and qin, a stringed instrument). Two players do battle on a board which has a grid of 19 x 19 lines. Each player strategically places his pieces, following several rules, to block territory (space on the board). The winner protects the most territory.

There are several measures of game complexity, including game tree size, decision complexity, game tree complexity, computational complexity, and state-space complexity. Journalists often use state-space complexity to describe the relative complexity of games, because it is fairly easy to grasp: “the number of legal game positions readable from the initial position of the game.” The state-space complexity for Go has been estimated at 10 to the 174, which is more than the total number of atoms in the universe. By contrast, the state-space complexity for chess is 10 to the 120. The difference is not trivial. A computer can play chess using brute force. It can calculate the possible move combinations after each play and select the next move out of the universe of possibilities, taking into consideration various strategies. Because the number of possibilities in Go is so high, a computer cannot use brute force. Instead, it must do something to approximate human intuition. It has been described as the “pinnacle of perfect information games.”

Round 1: Checkers

Computers have been beating humans at games for more than 20 years. In perfect information games, players move alternately and each knows all of the other player’s prior moves. In 1994, a computer program named “Chinook” developed by Jonathan Schaeffer at the University of Alberta, was declared the winner in a match against the world’s top checkers player in the Man-Machine World Championship. While the victory was impressive, there was a hanging question about the computer’s abilities. It was declared the victor after drawing six times. Marion Tinsley, its human opponent, then withdrew from the match due to problems with his pancreatic cancer. Chinook never actually won a game against Tinsley.

In 1995, Chinook played Don Lafferty in a 20-game match. It won one game, lost one game, and drew 18 times. Schaefer retired Chinook from competition after that match, but he and his team continued working on the checkers problem (a program that a human could not beat). In 2007, they announced that the best any human player could achieve in a game against the updated Chinook was a draw.

Round 2: Chess

The next human loss to computers in a perfect information game happened just a few years after the Chinook defeat. Chess had long been viewed as a game that challenged the smartest humans. A computer beating a human would make quite a statement about the state of computer “intelligence.”

In 1996, IBM’s Deep Blue played Garry Kasparov in a six game match. Kasparov won 4–2. But in 1997 they played a rematch which Deep Blue won 3 1/2–2 1/2. That was the first time a computer had beaten a Grand Champion chess player in a match following tournament regulations. Deep Blue’s victory was significant, though the victory represented brute force more than elegant play. At the time, some believed that Kasparov did not bring his best to all the games in the second match and could have won had he done so. Some believed if Kasparov had played with more human intuition he would have beaten Deep Blue.

After Deep Blue defeated Kasparov, IBM wanted another challenge to show off its software. Jeopardy presented that challenge. Jeopardy is more complex for a computer than chess. First, there is the format. The host gives the answer and the contestant must respond with the correct question. Second, Jeopardy involves language interpretation. As The New York Times described it, Jeopardy is “a game that requires not only encyclopedic recall, but also the ability to untangle convoluted and often opaque statements, a modicum of luck, and quick, strategic button pressing.

In February 2011, IBM’s latest masterpiece, Watson, played Jeopardy against Ken Jennings and Brad Rutter, the two leading human contestants. After three matches, the results were a clear win for Watson: $77,147 to Jennings’ $24,000 and Rutter’s $21,600.

As with Deep Blue’s win against Kasparov, Watson’s Jeopardy win against the human contestants was impressive, but it also showed that Watson was not perfect. Computers still had a long way to go when it came to matching wits with humans.

The Final Round: Go

With the chess and Jeopardy matches under its belt, the computer world wanted another win. Go was seen as the ultimate perfect information game challenge. A win against a human would show that  computers had moved beyond brute force and were taking on human “intuition.” The computer could not simply crunch numbers, it would have to do something else to beat a Go grandmaster. Two competitors took on the challenge, Google and Facebook, and Google got there first with AlphaGo.

In October 2015, AlphaGo played a match against Fan Hui, the European Go champion. AlphaGo won 5–0. While a significant victory for AlphaGo, its next match against Lee Sedol was an even bigger challenge. Sedol watched the games between AlphaGo and Hui and was able to evaluate AlphaGo’s strategies and weaknesses. Sedol predicted that, while the computer was good, he was still better.

As it turned out, Sedol was (mostly) wrong. In fact, most experts were wrong. In 2015 before the match between AlphaGo and Hui, most experts were predicting it would be another decade before a computer could beat a Go grandmaster. But in the few months between the match against Hui and the match against Sedol, AlphaGo continued improving. Unlike a person, AlphaGo could play games continuously and at a furious pace, learning all the time. What it learned gave it the edge.

How did AlphaGo learn to play Go so well? According to researchers at Google’s DeepMind:

AlphaGo was programmed to sift through a database of expert Go moves, and then play against itself millions of times to improve its performance. Researchers called that part of the program the “policy network.” Another part of the program runs through Monte Carlo simulations to evaluate board positions.

Today, Demis Hassabis, who founded DeepMind and still leads it after the Google acquisition, says his team believes AlphaGo could learn to play entirely through self-learning. As Hassabis says:

Actually, the AlphaGo algorithm, this is something we’re going to try in the next few months — we think we could get rid of the supervised learning starting point and just do it completely from self-play, literally starting from nothing. It’d take longer, because the trial and error when you’re playing randomly would take longer to train, maybe a few months. But we think it’s possible to ground it all the way to pure learning.

Because Go is so complex, during training AlphaGo had to learn how to use some measure of computer “intuition.” What do we mean when we say AlphaGo has “intuition”? Geoffrey Hinton, called the “godfather of neural networks” and a member of the AlphaGo team describes it this way:

The really skilled players just sort of see where a good place to put a stone would be. They do a lot of reasoning as well, which they call reading, but they also have very good intuition about where a good place to go would be, and that’s the kind of thing that people just thought compute[r]s couldn’t do. But with these neural networks, computers can do that too. They can think about all the possible moves and think that one particular move seems a bit better than the others, just intuitively. That’s what the feed point neural network is doing: it’s giving the system intuitions about what might be a good move. It then goes off and tries all sorts of alternatives. The neural networks provides you with good intuitions, and that’s what the other programs were lacking, and that’s what people didn’t really understand computers could do.

Does AlphaGo’s victory and use of “intuition” at some level mean computers are getting close to human abilities? According to Hinton, not for more than five years (he refuses to predict anything that he thinks is farther out than five years):

My belief is that we’re not going to get human-level abilities until we have systems that have the same number of parameters in them as the brain. So in the brain, you have connections between the neurons called synapses, and they can change. All your knowledge is stored in those synapses. You have about 1,000-trillion synapses—10 to the 15, it’s a very big number. So that’s quite unlike the neural networks we have right now. They’re far, far smaller, the biggest ones we have right now have about a billion synapses. That’s about a million times smaller than the brain.

There will not be another perfect information game challenge that surpasses Go. While there are other strategy games, they involve human language and interactions and other dimensions which still are well beyond computer capabilities.

Oren Etzioni, CEO of the Allen Institute for Artificial Intelligence, described AlphaGo’s wins as representing “an outstanding technical achievement, … demonstrat[ing] that when the goal is crystal clear, and the rules of the game are simple … computers will dominate.” Etzioni contrasted that situation with the ones that lawyers confront: “when the problem is ‘ill-defined,’ as in understanding a sentence, writing an article, or even comforting a friend – this is still way beyond our [AI’s] abilities.” Lawyers should not assume Etzioni’s comments mean machine learning computers are not ready for legal services. It turns out there are many ways computers can augment what lawyers do.

Part II of this two-part series will be posted on April 7, 2016. 

LumpOfLawyersImagine the world of law in the United States as a fixed amount of legal work that needs doing. That fixed amount includes all the legal services lawyers do for large corporations, individuals, criminals, and others. One recent estimate put the market for legal services by lawyers in the United States at about $275 billion.

In the United States there are about 1.2 – 1.3 million individuals who call themselves “lawyers.” Using some basic math, we can say that each lawyer has a claim to $275 billion divided by 1.25 million or roughly $220,000 of legal work. Some lawyers are very good and they get more than their share, other lawyers do not practice law so they forfeit their share, and many are employees so they don’t get a share they get a salary. In other words, do not go looking for your $220,000, that is just the average per lawyer.

In 2014, roughly 44,000 students graduated from law school. Of those, 71% or about 32,000 got jobs practicing law (more precisely, they found jobs either requiring a law degree or in a position where a law degree was preferred). Add those 32,000 to our 1,250,000 and we are at 1,282,000 lawyers. Of course, some lawyers retired, so we will drop the number of lawyers to 1,275,000. Redo the math and each lawyer now has a claim to about $216,000 of legal work. Makes sense, right? The more lawyers given a fixed amount of legal work, the less legal work per lawyer.

Now assume that we can start replacing lawyers with computers. The Harvard Business Review recently ran an article suggesting that 30% of tasks done by knowledge workers can be done by existing computer programs. Assume that number holds true for legal work (there is nothing suggesting that it wouldn’t, since the tasks are basic ones that apply across all types of knowledge worker jobs) and we could say that of the $275 billion of legal work, $83 billion could be done by computers leaving $192 billion for those 1.275 million lawyers. That would reduce the legal work claim by each lawyer from $216,000 to $151,000. Ouch!

Economists do not believe a world exists where the amount of work available to laborers is fixed and so they call the imaginary world I describe above the “lump of labor” fallacy. They argue that in the real world, the amount of work increases as technology enters the market. This is the broader argument by many technologists: artificial intelligence and increasing computer power may displace workers from some jobs they hold today, but those jobs will be replaced by other jobs many of which we cannot imagine right now. People may shift jobs, but we will not have mass unemployment in the foreseeable future. Relax.

The Lump of Lawyers Argument

A good lawyer always has another argument and when it comes to technology replacing lawyers, lawyers are ready. Many lawyers argue that law is different than other knowledge work. It is not a profession that is amenable to computerization like, say, medicine or finance. It takes a certain amount of lawyers to process a given amount of legal work. The amount of lawyers may vary a bit depending on a lawyer’s experience, expertise and (not to be forgotten) pedigree, but within a narrow band the amount of lawyers to do the work is fixed. Therefore, lawyers argue, computers will not greatly affect the legal profession in the foreseeable future because computers cannot replace lawyers doing legal work. It takes a certain lump of lawyers to do the work. This is the lump of lawyers fallacy.

Unlike the lump of labor argument, which is in disrepute, the lump of lawyers argument has tremendous support in the legal profession. Lawyers trot out the argument almost any time they are in a room talking about technology. Some of what they say comes from ignorance, some from the challenge of believing that a computer could replace any part of what took them three years of post-graduate education and 20 years of practicing to learn, and some from economic self-interest. Mix all of that together and lawyers dismiss the “computers replace lawyers” idea and head back to their offices to review the latest draft for typos. No fallacy to see here, keep moving.

The Amount of Legal Work is Growing

While a few lawyers may challenge the lump of labor argument, it seems most believe that the amount of legal work is growing. We have seen increases in compliance work, regulatory work, and laws going on the books in countries that are modeling their legal systems after the United States. Any lawyer at a global corporation knows that his or her workload is not shrinking. Add to that the legal complexity coming from new technologies (e.g., 3D printing, artificial intelligence, nanotechnology) and you have plenty to do if you are in the corporate legal world.

If we go further and include that universe of middle income and low income people who need legal services, the amount of legal work certainly is increasing. Each new layer of regulations and laws means individuals have more areas where they are at risk for not consulting a lawyer. Despite some feeble efforts to stop the tsunami, the laws on the books of the states and federal governments seem to increase each year.

Looking at the labor to do certain legal tasks, we can point to some obvious areas where computers have made progress. EDiscovery is one, but we can add in Shepardizing, due diligence, and contract automation as a few others. Even ancillary tasks, such as recording billable hours, have seen advances. Some advancements are just starting to penetrate the thick membrane lawyers use to keep out anything new. These include ways to research caselaw using something other than boolean searches and programs that can construct basic texts from facts. Keep in mind that automation replaces tasks, not people. When it comes to automation, the legal industry is only beginning to see the tip of the point at the end of the spear.

Legal Education Needs to Drop the Lump of Lawyers Fallacy

A recent report from the Christensen Institute (yes that Christensen, the one who wrote The Innovator’s Dilemma) suggests that law schools need to dramatically restructure the legal education process if they want to survive much longer. The authors recommend moving from the time-based approach used today to a competency approach using modules, online learning, and other techniques that will make legal education more relevant and less costly. As you might imagine, many legal educators will not warmly embrace these suggestions since they would mean change in a system that has been used for 150 years and would undermine the lump of lawyers argument. The recommendations focus on students learning through modules. This approach fits with automation — as tasks are automated educators can re-focus modules on new tasks students need to learn.

Watch Out for the Black Swan

As any experienced driver knows, the greatest danger often comes not from where you are looking but from where you are not looking. In more modern terms, we call this the Black Swan. Black Swans are unexpected and for most lawyers, computers taking over much of what they do will be unexpected. But there is an even greater surprise in store for lawyers who adhere to the lump of lawyers argument, and that is the end-around.

Many programs I see in development today are not aimed at lawyers. These programs are aimed at clients who will use the programs to reduce their need for lawyers. We have seen LegalZoom and Rocket Lawyer do this in the retail market for legal services. Now, software developers are targeting small and medium-sized businesses for the next round of programs. These programs are aimed at clients and can draft documents, review documents, and provide suggested alternative language.

To put it a bit more bluntly, outside lawyers have said they will not change unless clients force them to. In-house lawyers have started looking at ways to improve and have reduced their need for outside lawyers. But clients are impatient and, seeing what can be done with computers in other areas, naturally ask the question “why can’t computers reduce or eliminate the need for lawyers”? That question is music to a software developer’s ears. The market is starting to respond and in-house lawyers will feel the impact as well as outside lawyers.

The Black Swan for lawyers may be the software packages that start eliminating lawyers as the “man-in-the-middle.” Whether lawyers like it or think it appropriate, clients will move like water finding the path of least resistance to get their work done. Lawyers who build dams will find their clients going around, over, or under those dams and that may mean using software that does some or all of the work lawyers have done. No more lump of lawyers.

By definition, it is hard to imagine the Black Swan. For lawyers, it is hard to imagine clients going around them. That is the lawyer’s problem, not the client’s. The circumvention approach is not widespread yet and the impact on law firms and law departments is very small. It will grow. But lawyers can control its growth by recognizing that client needs outweigh lawyer protectionism. Law schools need to embrace new ways to educate lawyers, and lawyers need to embrace using computers to handle client matters.

Contrary to what many lawyers believe, now is a time for optimism about the legal profession. The need for legal services is strong. For the first time in history, we have the ability to shed the boring, tedious, repetitive tasks that bog down the day and add stress to every legal practice. We have the ability to leverage a tool that can free us to focus on what we love to do instead of spending our time on exchanging labor for dollars. The legal issues societies face have never been more interesting and challenging and the promise is that there are more of those issues to come, if only lawyers structure their practices in ways that clients will pay them to tackle those issues. Time to enter the real world, lawyers, embrace computers, drop the fallacies, and focus on the issues that matter to clients.