Kill the Lawyers!

There’s a scene in Shakespeare’s Henry IV where a band of thugs plans to take over the kingdom.

Their leader, Cade, shares a grand scheme – a new utopia where money will no longer exist, and where everyone will have access to a life of luxury. His accomplice, Dick, then jumps up, and delivers one of Shakespeare’s most epic lines:

“The first thing we do, let’s kill all the lawyers!”

Regardless the political system, lawyers become the de facto guardians of its status quo. Anyone up for a power grab, pushing for radical change, or simply out for some fun chaos, has a strong incentive to seriously disrupt the system they speak for. Technology, of course, loves to do all three.

In recent years, venture capital has increasingly flowed into a new generation of LegalTech apps designed to automate the legal profession, making humans expendable wherever possible. Bots can now already scrub illegal content on YouTube, give predictions on the outcome of insurance payout cases, review mounds of documents in complex lawsuits, even review simple contracts.

They’re all pursuing the dream of a cheaper, faster, and more objective legal system, where claims are initiated by the click and justice is doled out in a suave, Siri-like voice.

And who’s to say there aren’t massive potential benefits there? The law is complex and expensive. People make mistakes. Judges have different subjective values, and often decide accordingly.

To top it off, lawyers are often depicted as mercurial, money-grubbing technocrats, whose chief skill lies in hiding their motives behind self-serving jargon and arcane procedures. The entire profession seems designed to profit from its own inefficiencies. The idea of Uber-fying it with technology is inherently exciting.

Of course, the “let’s kill the lawyers” utopia is just a small subset of the larger “let’s kill the humans” push of an increasingly aggressive automation agenda targeting the labour force – where McKinsey estimates that over 30% of all jobs will be replaced by automation just over a decade from now.

But there’s something particularly ironic about the sudden excitement at lawyers being replaced by robots.

Isn’t the intuitive critique of lawyers that they should be more human and less robotic? That their blind elevation of the rules should be balanced with genuine concern for real people, or with the pursuit of a broader sense of equity and justice? Are the real problems with law of the kind that computer engineers are most qualified to fix in the first place?

In his massive, deeply enlightening study The Future of Employment, Oxford machine learning professor Michael Osborne uncovers clues pointing in this somewhat counter-intuitive direction.

Osborne and his colleagues project that, in the next couple of decades, occupational demand will actually increase for skills and tasks having a strong human element. Those that technology simply can’t easily replicate.

Skills like interpersonal qualities, advanced language and social science knowledge, broad systemic thinking, creativity, and even caring and counseling are all projected to see a rise in workforce share.

Osborne’s findings imply that, while humans may be bad robots, robots can be even worse humans. By sucking away mechanical tasks and performing them better than humans, AI may end up forcing us to place more focus on the things that only humans can do.

Insofar as legal systems involve the rut application of clear, black and white rules, we may indeed be better served by outsourcing these to unbiased, hyper-efficient bots. The cold, skilled technician with trivia knowledge and limited creative or people skills will quickly become a relic of the past. Lawyers who have subscribed to this popular professional identity may have already dug their own graves.

This topic also opens up larger questions about the nature of law. Is it really just a left-hemisphere, purely rational discipline, where outcomes are generated via clear, objective steps? While many lawyers and judges love to project this “purely scientific” image of law, in truth, it doesn’t stand much scrutiny (as any number of 5-4 Supreme Court decisions will attest).

The elephant in the courtroom is that law uses fuzzy, deeply subjective concepts as its leverage points – words like “reasonable”, “equity”, “just”, and many others, whose definition is – and always will be – up for grabs.

The meaning of those words vary according to personal, political, and philosphical views. They also evolve along with broader social, cultural, and demographic change. We shape them, and they, in turn, shape us. They are vehicles through which we formulate, define and transform our social identities.

Should we outsource this critical function of democratic life to bots with no skin in the game? Even if we could define these terms definitively for our best bots to use, is this something we should do? Answering yes could lead to a wicked kind of irony: under the banner of progress, we’ll have eliminated our very ability to define progress for ourselves. To the benefit of hyper aggressive, venture-backed, monopoly-seeking corporations no less.

Already, there are legitimate concerns that AI apps are susceptible to creating new forms of prejudice.

The EU recently implemented a “human intervention” rule as part of its GDPR: automated decisions can’t be taken without human oversight where a decision could damage human interests. We may need more such mechanisms to be implemented in an age of powerful, independent, singularity-seeking AI apps.

At the same time, there’s little doubt that AI and other technologies can improve the legal system in exciting ways. We especially need them in the face of a legal system that keeps getting more expensive and complex.

But technology agendas are not neutral. They’re almost always fueled by the need to justify massive investments with even more massive returns. They create new social issues and externalities – hidden costs – that society at large has to pay for. They often end up in monopolies. Do we really want to unleash the full power of this on law – the very end-product of democracy?

Lawyers will have to re-invent the profession along lines they haven’t always fostered a comfort level with to be the custodians of a truly human system of governance.

So sure, let’s kill the lawyers who act like bad robots.

But, as they’ve all too often forgotten, lawyers can be human, too.