By Elizabeth C. Tippett and Charlotte Alexander 5 minutes Read
Imagine what a lawyer does on a given day: researching files, writing briefs, advising clients. As technology nibbles the edges in the legal profession for some time, it is difficult to imagine that these complex tasks are performed by a robot.
And it is these complicated and personalized tasks that have led technologists to include lawyers in a broader category of jobs considered to be fairly secure in a future of advanced robotics and artificial intelligence.
But, as we discovered in a recent research collaboration to analyze legal cases using a branch of artificial intelligence known as machine learning, the work of lawyers is much less secure than previously thought. Turns out, you don’t have to completely automate a job to fundamentally change it. All you have to do is automate some of it.
While this may be bad news for tomorrow’s lawyers, it could be great for their future clients, especially those who are struggling to afford legal assistance.
Technology can be unpredictable
Our research project—In which we collaborated with computer scientists and linguists at MITER, a federally funded nonprofit research and development organization, was not meant to focus on automation. Like right professors, we tried to identify the textual characteristics of successful and unsuccessful legal briefs.
We have assembled a small cache of legal briefs and opinions of judges and processed the text for analysis.
One of the first things we learned is that it can be difficult to predict which tasks are easily automated. For example, quotes in a memoir, such as “Brown v. Board of Education 347 US 483 (1954) ”, are very easy to locate and separate from the rest of the text. This is not the case with machine learning software, which was triggered by the punctuation blizzard inside and outside the quote.
It was like these CAPTCHA boxes you are asked to fill in on websites to prove you are not a robot – a human can easily spot a telephone pole, but a robot will be confused by all the background noise in the image.
A technological shortcut
Once we figured out how to identify citations, we inadvertently stumbled upon a methodology to automate one of the most difficult and time-consuming aspects of legal practice: legal research.
MITER scientists used a methodology called “graphical analysis” to create visual networks of legal citations. Graphical analysis allowed us to predict whether a thesis would “win”, based on the performance of other theses when they included a particular citation.
Later, however, we realized that the process could be reversed. If you were a lawyer responding to the other party’s brief, you would normally have to laboriously search for the right cases to cite using an expensive database. But our research suggested that we could create a database with software that would simply point lawyers to the best cases to cite. All you have to do is put the other party’s file in the machine.
Now, we haven’t actually built our hotkey search engine. It would take a mountain of legal briefs and legal opinions to do anything useful. And researchers like us do not have free access to data of this type, even to the database managed by the government known as PACER fresh per page.
But it shows how technology can turn any extremely time-consuming task for humans into one where the heavy lifting can be done with the click of a button.
A story of partial automation
Automating the difficult parts of a job can make a big difference both to those doing the work and to consumers on the other side of the transaction.
Take, for example, a hydraulic crane or a power forklift. While today people regard the use of a crane as manual labor, these motorized machines were seen as labor saving devices when they were first introduced. times, because they supplanted human power involved in moving heavy objects.
Forklifts and cranes, of course, weren’t replacing people. But as if to automate the work of legal research, electric machines multiplied the amount of work that a person could accomplish in a unit of time.
Partial automation of sewing machines in the early 20th century offers another example. In the 1910s, women working in textile factories were no longer responsible for sewing on one machine – as is the case today, on a home sewing machine – but fighting over an industrial-grade machine. with 12 needles sewing 4000 stitches per minute. These machines could automatically do all the hard hemming, sewing and even sewing work of the “white underwear embroidery trim. “Like an airline pilot flying on autopilot, they didn’t sew as long as they watched the machine for problems.
Has the transition been bad for workers? Maybe a little, but it was a godsend for consumers. In 1912, women browsing the Sears mail order catalog had the choice of “drawers” with hand embroidered cut and a much cheaper machine-embroidered option.
Likewise, automation could help reduce the cost of legal services, making them more accessible to many individuals who cannot afford a lawyer.
Indeed, in other sectors of the economy, technological developments in recent decades have allowed companies to shift work from paid workers to customers.
Touch screen technology, for example, allowed airlines to install check-in kiosks. Similar kiosks are almost everywhere, in parking lots, gas stations, grocery stores and even fast food restaurants.
At one level, these kiosks replace the paid labor of employees with the unpaid labor of consumers. But this argument assumes that anyone could access the product or service when it was performed by an employee.
In the context of legal services, the many consumers who cannot afford a lawyer are already give up their day in court or dealing with legal claims themselves, often with bad results. If partial automation means that a busy legal aid lawyer now has time to handle more client files, or clients can now afford to hire a lawyer, everything the world will be better off for it.
Additionally, technology-based legal services can help consumers better represent themselves. For example, the Missouri Federal District Court now offers a platform to help people declaring bankruptcy to prepare their forms, either by themselves or with a free 30-minute meeting with a lawyer. Because the platform offers a head start, the lawyer and the consumer can better utilize the 30-minute time slot.
More help for consumers may be on the way — there is a exceptional harvest of tech startups scrambling to automate various types of legal work. So while our hotkey search engine wasn’t built, powerful tools like this might not be far off.