Beneath the streets of Westminster, Winston Churchill and his team led the British war effort. That workspace, known as the Cabinet War Room, has been preserved just as it existed when the government closed it down in 1945. Not only is this an incredible artifact from World War II, but the exhibit also shows how people worked 75 years ago. Paper, physical maps, and telephones were the tools of the day.
Today, legal professionals, of course, spend much of their day interacting with computers. A desktop or laptop computer is the hearth of our workspace, where we do simple tasks like email, as well as complex tasks like using sophisticated systems to analyze data collections. We use our mobile phones, tablets, and ever-present digital assistants like Siri, Alexa, and more powerful tools like IBM’s Watson. Computers now assist us at trial—and they may soon drive us to the courthouse.
Of course, some of this technology, like email, has been around for decades. But the newest forms of technology, especially those casually being dubbed “AI,” now encroaches on the home territory of what legal professionals do: think.
So what will legal professionals do in 10, 40, or 100 years?
Some of this terrain is well-considered by Richard Susskind in the End of Lawyers, Tomorrow’s Lawyers, and The Future of the Professions. The role of algorithms and the integration of human and machine thinking is beautifully explored by Yuval Noah Harari in Homo Deus and 21 Lessons for the 21st Century.
In addition to these ideas, I offer five ways professionals will—and must—work with technology; that is if we want the best outcomes for our clients and the future legal system.
Decide the best division of labor.
We should use technology only when it outperforms humans. For example, a computer can process certain large data sets more effectively than human reviewers, at least for particular purposes. On the other hand, a skilled personal interview is the best way to learn from a witness, and therefore may be the best way to learn about and locate the “smoking gun” document. As machines get better, the comparative advantage will shift toward computers and the skills gap will narrow. But the best legal professionals will still know when to use technology to solve a problem, and when not to.
Learn how to work with machines.
UC Berkeley roboticist Ken Goldberg uses the term “multiplicity” to describe humans and machines working together. This way of working is happening now with navigation systems, driver assist, and collaborative robotics. In the world of chess, skilled human players initially trounced computers at the game. Now computers readily beat chess masters. But the best chess players are an integrated team of human and machine. As Gary Kasparov admits, “I lost (chess) but I survived, and I thought if you can’t beat them, join them. From now on we have no choice but to work with machines and make the best algorithms.” For legal professionals, we have entered such a period when it comes to technology-assisted review (TAR) that the best work is being done by humans working closely together with a computer. This will be an increasingly important skill as more legal work is done in tandem with intelligent machines.
Ensure that algorithms reflect our values.
Computer algorithms sometimes fail to reflect important human values, such as racial equality. Algorithmic bias happens when errors in a computer system create unfair outcomes, such as privileging one group of users over others. Not only must we be vigilant in spotting this form of bias, but we will need to work harder to identify when algorithms are making decisions in the first instance, especially as algorithmic decision making becomes more common, better, and less visible.
Frame problems using human intuition.
As legal professionals, we need to be aware of the larger context of problems and make sure our technology analyzes the critical evidence needed to arrive at solutions. Following Bayes’s Theorem, our degree of belief—expressed as a probability—should rationally change to account for the availability of related evidence. If we blindly plug data into a “black box,” we ignore a priori information critical to arriving at the right solution. In the Laws of Medicine, Siddhartha Mukherjee uses the example of a man flipping a coin at a roadside fair. When the coin lands “heads” after 12 straight tosses, the statistics student (or computer algorithm) predicts a 50 percent probability it will come up heads again. Yet a child knows the correct answer: the coin is rigged. So sure, a computer may be able to quickly solve the problem, but we must ensure it has the needed data—and know whether it’s even the right problem.
Focus on human skills.
At our best, we excel at empathy, interpersonal communication, curiosity, ethical and spiritual thinking, determining essential meaning, leading, and inspiring others. Since most legal problems are fundamentally human problems, every dispute is an opportunity to use our imagination and subtle understanding of the world.
How do we ensure machines keep serving humanity rather than the other way around?
First, as legal professionals, we must claim certain work as belonging to the human domain. If we do not, our role in shaping legal work and the future of our profession may be lost forever.
Second, we cannot allow our human talents to wither. We should spend less time trying to improve at computer-like thinking, and perhaps less time in front of screens. In its place, we must devote more time and resources to developing our human abilities. In short, to lead in the age of machines, we should become better humans.