A recent Wall Street Journal headline caught my eye. The gist of the article, entitled CEOs Start Saying the Quiet Part Out Loud: AI Will Wipe Out Jobs, is what it sounds. It quotes a number of CEOs to the effect that AI is going to massively impact white collar jobs. The article further points out that these leaders have until recently sugarcoated the impact AI will have, particularly on white collar jobs. That view is now evolving as AI becomes increasingly capable. (These realities were recently brought home to me as I watched two very fine and knowledgeable people at Microsoft get laid off.)
Lawyers aren’t immune and are mentioned by one industry leader in the article as being ripe for downsizing. Suffice it to say that at best, AI is going to significantly disrupt white collar work either by reducing the workforce or by requiring that the nature of white collar work change.
A More Optimistic View
A second article I saw, one by Akshay Verma, is more optimistic, at least for legal. (The article appeared in the Beacon, a Newsletter of Spotdraft.) Verma says that automation and AI will free up time historically spent on administrative burdens. In turn, this will enable freed up lawyers to deliver greater strategic value. He cites how legal ops teams had a similar impact as a result of the Great Recession. Lawyers will become “strategic advisors, risk architects and business enablers,” according to Verma.
Verma’s view is the one traditionally heralded in the legal community: AI will not replace lawyers, it will replace lawyers that don’t use it. If correct, this opportunity combined with the increased workload that AI is bringing to legal as I recently discussed, could open a new world so the theory goes.
But Let’s Not Get Ahead of Ourselves
But before we go too far down that optimistic path, let’s consider some inconvenient and troubling realities. First, it’s true that at least for the time being, AI is resulting in more, not less, work. This new work comes from matters that before AI could simply not be done with sufficient efficiency to make them worthwhile doing. But that presently untapped work will at some point be exhausted, however. The question will then be whether there is even still more untapped work out there to replace it. Or will there be entirely new categories of work that AI still can’t do well?
Which raises a second point. As we move forward, AI will be able to take and do increasingly more difficult and complex tasks. If true, AI could erode some of the human tasks that today us humans must do because AI can’t. So, the new work that AI generates may one day be cannibalized by AI. While today, AI is creating more work, tomorrow it may do that work.
Which brings me to a third potential reality. Yes, AI will free up lawyers to do more strategic and visionary things. The kinds of things Verma envisions them doing. But two things. First, the standard reasoning is no doubt true: AI will no doubt give all of us more time to do the visionary thing. It already has.
But We Aren’t All Good at the Visionary Thing
But I practiced for a long time. And I’ll be honest: the reality is that there are not that many lawyers who are good at overall strategy and vision. It’s not something they teach in law school. And not every lawyer, even with more time, will grow into that role and develop those skills. The second reality is AI is gunning for this work too. AI is already not bad at strategy and vision and it’s only going to get better. (See my recent article on how medical diagnostic AI models could be applied to legal.)
The result? Less need for the high-end lawyer/thinker. That’s reality. Fewer lawyers with the skill Verma envisions, less need for those skills. So even if we free up more time for strategic thinking, that doesn’t mean that all of us are going be spending more time doing it. A sobering prospect, to say the least.
Here’s one other real problem that we are going to have to face. The business model of most law firms has for years been the billable hour. It’s a simple model: bill more hours, make more money. And because one person can only bill so many hours a day/week/year, you can only increase what you make by leveraging matters and having as many people work on a matter as possible.
The Real Business Model Problem
The result: in the age of AI, there may simply be too many lawyers. We built a model based on maximizing time spent. As a result, the profession is over built and is particularly susceptible to disruption by a technology that has the potential to so substantially reduce time spent on matters. And when you overbuild, you hire and promote people who aren’t necessarily good at being the “valued advisor” but who are good at generating billable hours. It certainly doesn’t mean that the leverage model protects us.
What Now?
Where does all this go? It remains to be seen. But it’s just these kinds of questions that leaders in the profession need to be asking. Maybe the future will be more of the same as the past.
But if CEOs are now saying the quiet part out loud, it’s time for legal to do the same.
Stephen Embry is a lawyer, speaker, blogger and writer. He publishes TechLaw Crossroads, a blog devoted to the examination of the tension between technology, the law, and the practice of law.