Zack Halper

It’s often a running joke that law has a difficult time adapting to changes in society; where the medical profession implements robots, lawyers are stuck behind less than state-of-the-art search engines. But perhaps the greatest fault of the legal profession, is not utilizing the growing field of artificial intelligence (AI).

Coming from a cognitive science background myself, I see massive potential for the merging of the two fields that could very well make a positive ripple in the lives of lawyers, and their clients alike. Amongst the many developing options, machine learning is one such area that has caught the interest of legal revolutionaries. Without getting too technical (to avoid boredom for those reading this), machine learning employs a system much like our own brains to provide pattern recognition and data analytic abilities beyond a normal human’s. In a process that is very similar to the way our neuronal networks operate, machine learning finds ways of categorizing or predicting variables by running simulations or by being given data sets, whereby guesses are made by the system for the right answer, and each time it is correct, the functions the system used to achieve the correct guess are reinforced. This means its ability to guess correctly increases, and the speed and efficiency of how it guesses also does. Fairly simple right?

Well, it gets more complicated- but for the sake of brevity, I’ll give some quick examples. AI can be given a supervised, or pre-loaded data set, to allow it to understand where corrections need to be made. This can apply to contract review or criticism, whereby the most complicated aspects of the black letter of the law can be pre-loaded and subsequently commented on, as well as the simplest aspects of grammar that need a touch up (which I think we could all use).[1] Many technologies startup companies such as Lawgeex and Klarity have begun to develop this kind of AI.[2] So before long, we may not have to think before we sign a contract – so long as we have the newest gadget.

Alternatively, unsupervised machine learning can look for patterns or trends in a random assortment of data- think litigation, where charisma has just as much a place as accuracy. However, those human elements might be a thing of the past if Blue J Legal has anything to say about it.[3] The Toronto based team is currently working on an AI system that predicts litigation outcomes in tax law, which can allow clients to decide whether they want to pursue the difficult (and expensive) process of litigation.[4] With a 90% accuracy rate, this new technology might force lawyers to brush up on their mediation skills.[5]

With these growing innovations in machine learning, who’s to say how the law will meld with technology in the near future. I personally predict that machine learning will allow us to create AI that will inevitably make the leap to performing a judge’s duties one day. We’ve already seen the forerunner of this- in the form of an experiment in Estonia that tested the limits of an AI judge to settle contract disputes.[6] By reviewing the documents of the involved parties, the robot judge decided whether the parties deserved the honor of being able to go before a real human judge in small claims court.[7] A system like this could help with the difficult backlog that many courts face today, and the considerable access to justice issues that arise from the cost of litigating disputes with minimal damages to be awarded.[8] This kind of scenario may seem like a dream (or nightmare) steeped in science-fiction, but it’s entirely possible that Judge Judy may have an AI competitor for most well-known TV judge sooner rather than later.

On that note, it is becoming more and more probable that one day, AI judges may operate in tandem with human judges. In theory, while an AI judge handles the logical, axiom-based connections of legal doctrines and ratios that correspond to a party’s argument, a human judge could then judge the ethical, pragmatic and “societal” based considerations, to create two halves of a strange whole. While we are a long way from systems like this developing (leading to the conclusion that I may need to write a book of fiction on the topic), it is likely that we could be closer than ever to a legal paradise of accurate laws and expedient help for wronged individuals with our trusty AI companions. However, I’m sure that depending on who you ask, that future looks more like a dystopia where our robot overlords rule over us with iron gavels. Only time will tell.

Sources:

[1] Rob Toews, “AI Will Transform the Field of Law”, Forbes (19 Dec 2019), online: https://www.forbes.com/sites/robtoews/2019/12/19/ai-will-transform-the-field-of-law/?sh=1c35e1e97f01.

[2] Ibid.

[3] Ibid.

[4] Ibid.

[5] Ibid.

[6] Tara Vasdani, “Estonia set to introduce ‘AI judge’ in small claims court to clear court backlog”, The Lawyer’s Daily (10 Apr 2019), online: https://www.thelawyersdaily.ca/articles/11582/estonia-set-to-introduce-ai-judge-in-small-claims-court-to-clear-court-backlog-.

[7] Ibid.

[8] Ibid.

*Author’s note: this post relates specifically to the “Designing Expert Legal Systems,” class taught at the TRU Faculty of Law by Professor Katie Sykes. Some references may be obscure to the passing reader, but the overall message remains*

As a law student enrolled in a legal app-making class, you might wonder, as I sometimes do, if you are contributing to the demise of your chosen profession. We recognize the need for access to justice, but cringe as our LOC balances climb.

But is this fear rational? Can access to justice, well-paid professionals and artificial intelligence co-exist? Can we have our app-selected cake and eat it too? Or does the transfer of Jake’s beautiful mind into a question flow indicate that the human touch is no longer relevant?

One of the best (though anecdotal) arguments I have heard to the contra comes from the fundamental human experience: no matter how tight machines reason, they are missing the human component. Would you feel comfortable, for example, with an app deciding the fate of a murderer – particularly considering the high bar for conviction? That is, if someone’s liberty were at stake, would you trust a machine to make the relevant considerations, both factual and moral, to determine that person’s sentence?

To me, these questions infringe on issues of conscience and emotion. It is easy to program a machine on the basis of pure logic (the if-then’s and yes/no’s we’re familiar with as Neota magicians). But the forces driving these rules are harder to pin down. Why is it logical to have a specific legal outcome, in the context of human ethics and morals? At some point, rationality loops back to human emotion. Ultimately, our sense of justice (sociopaths aside), comes not from logic, but from our feelings. It is unlikely that we will project these feelings onto apps, as we don’t fully understand them ourselves. As thinkers like Sam Harris point out, consciousness very much remains a mystery, and might always lie beyond human explanation.

Existential threats aside, there are reasons to believe we will still be gainfully employed. As Tara Vasdani pointed out, there is an explosion of entrepreneurial options for those (such as ourselves) entering the workforce during the human-AI transition. And there is reason (amidst clouds) for hope, even beyond these “first-finder” opportunities.

Richard and Daniel Susskind, a father and son team from England, detail their predictions for the impending techno-age. The bad news is that they expect a fundamental dismantling of the professions as we know them. The need for specialists with particular knowledge/skillsets will inevitably collapse as information is widely disseminated and applied by increasingly smarter machines. Expert knowledge will cease to be monopolized by scholars undergoing years of focused training. The upside – beyond the obvious boon for democracy and access to justice – is that this new technology will require engineers (including knowledge technicians), implementers, managers, and maintenance personnel of all stripes. And, if you are reading this because you are part of the DLES class at TRU Law, you can rest easy knowing you are already a part of vanguard (*pats back*)!

Jokes aside, Daniel Susskind offers some practical advice for young workers: “learn to be good at the sorts of things these systems…cannot do or…try to build the machines” yourself. It is a matter of staying ahead of the curve, rather than being underneath it. More hearteningly, he seems to think that conventionally trained professionals are up to this task. When asked about aspiring doctors and lawyers, he recommends (with caveats) that they take a “traditional” path:

     “But be far more agnostic and open minded along the way about opportunities that come up. Because what’s interesting is that if you look at these technologies – take the systems developed by DeepMind or the ones that recognise melanomas by Sebastian Thrun’s team at Stanford – these teams contain lots of domain experts, such as trained doctors.”

While the message is complicated, it is clear that young professionals still hold an advantage in the job market. The difference, for law students, is that our advantage now stems from exposure to new ideas and thinkers (and networking opportunities), rather than formal education, as it did in the past. The opportunities that lie ahead for us are hard to predict, but will undoubtedly lie outside of our traditional conceptions of what lawyers “do.” In the long-run, Universal Basic Income, and machine replacement of labour, might make these considerations moot. Until then, be kind to the machines, and pray that they decide to erase your debt.

Sources:

Harris, Sam. “The Mystery of Consciousness,” (October 11, 2001) Samharris.org.

<https://samharris.org/the-mystery-of-consciousness/>.

Susskind, D. and Susskind, R. The Future of the Professions: How Technology Will

Transform the Work of Humans and Experts (Oxford: Oxford University Press, 2015).

Tucker, Ian “The Observer: Daniel Susskind: ‘Automation of jobs is one of the greatest

questions of our time,’” (January 28, 2020) The Guardian.

<https://www.theguardian.com/technology/2020/jan/18/automation-jobs-universal-basic-income-daniel-susskind-interview>.