In August 1915, Robert Frost first published arguably his most famous poem, “The Road Not Taken”. The poem has since become a ubiquitous part of the American culture, with its signature phrases – “Two roads diverged in a wood and I–I took the one less traveled by, and that has made all the difference”— being championed by countless high school valedictorians as an ode to individualism and following one’s own path. While the poem begins with a congratulatory tone as the speaker extols his decision to take the path less taken, by the end, the two roads that once diverged in that wood appear to be equally travelled, interchangeable, and circular in nature. David Orr, a poetry columnist for the New York Times Book Review, once wrote of Frost’s poem: “It is a poem about the necessity of choosing that somehow, like its author, never makes a choice itself—that instead repeatedly returns us to the same enigmatic, leaf-shadowed crossroads.” See The Paris Review: The Most Misread Poem in America (2015). The beauty of Frost’s words and their deeper juxtaposition are perhaps more relevant today than ever when considering the challenges that society, and the law, is now facing with advances in artificial intelligence (AI).
AI has had a controversial history since its early conceptions in the 1800s. Some have cast AI in a positive light: Sabine Hauert, a professor at the Bristol Robotics Laboratory at the University of Bristol, has opined that “[r]obots are not going to replace humans, they are going to make their jobs much more humane. Difficult, demeaning, demanding, dangerous, dull – these are the jobs robots will be taking.” Others have taken a bleaker view. Renowned physicist Stephen Hawking warned that “[t]he development of full artificial intelligence could spell the end of the human race.” Decades earlier, Alan Turing—widely considered the father of AI—believed “we should have to expect the machines to take control.”
Many people quickly dismiss sentiments like Turing’s. But throughout modern history, technologies that are now tightly woven into the fabric of our lives were originally seen as passing fads that would never have widespread use. Personal computers, laptops, and the internet were all dismissed by many early on. But today, those things are indispensable parts of our modern lives. AI is already being used in a wide variety of mainstream applications, including targeted advertising, digital voice assistants, online content generation, and online chat-based customer service or reception services. So, it’s not hard to imagine AI, like the internet, eventually becoming so indispensable that we forget we ever doubted it.
It’s also not hard to imagine how applications of AI could work in the divorce and child custody context, even now in straightforward matters that require relatively routine applications of the law. AI-based software could be used to process cases involving agreed, cookie-cutter orders based on some basic inputs by the parties. This could conceivably eliminate a large portion of the Attorney General’s Child Support Division, which spends the majority of its time processing basic visitation and child support orders and then enforcing them. It could also reduce or streamline the workload for overworked judges across the state who may manage dockets of literally thousands of cases. So perhaps there is some merit to Sabine Hauert’s prediction about AI reducing mundane work and making it more humane.
On the other hand, a divorce or child custody dispute highlights a key issue in the discourse surrounding AI: there’s a difference between what technology can do and what people are comfortable using it for. In a divorce or child custody dispute, clients are often experiencing a real trauma, and an attorney as “counselor” is no more apparent (or important) than it is in a contested family law case. Human contact is a critical component of the attorney-client relationship, especially in matters as sensitive and emotionally charged as a family law case. It is arguably a critical component of the role of a judge or juror as well (not to mention the constitutional implications of replacing jurors with AI—a topic for another blog post). And while case outcomes may often be predictable, the guidance and support that attorneys can provide their clients is not. This conclusion, of course, arises within a cultural milieu where AI has not had widespread use, at least in family law, for a lengthy period of time. However, what people are comfortable with technology doing is a moving target—people become more comfortable with technology doing more things the more they are exposed to it.
Thus, it seems that the law too has arrived at the same enigmatic crossroads between technology and human ingenuity. But are those truly divergent paths? Will AI eventually replace human divorce and child custody attorneys? For that matter, will AI eventually replace judges and juries? Or will it simply change the nature of the attorney-client relationship and the nature of role of judge and juror? For now, at least, the human component seems inextricably part of the attorney-client relationship. Most clients would agree that getting guidance from a human attorney—or a thoughtful ruling from a human judge—is not the same as talking with ChatGPT. But attitudes 10, 20 or 100 years from now are anyone’s guess. Time will most certainly make all the difference.