For lawyers, what needs to stay analogue, i.e. physical, impromptu, unstructured and face to face, and what should be moved into the digital realm? That is the key question. As, after you have answered that, you can then ask the second key question: of those things that are digital, what should be automated?
For Artificial Lawyer that really is the crux of everything happening in legal tech and more broadly across the legal market as a whole. This is because digitisation, and then taking that a step further into automation, drives real change in how legal services are made and delivered, ranging from an economic benefit, to the client experience, to how we define what ‘a lawyer’ actually is expected to do and what they are not.
This site has focused a lot of attention on considering digitisation and automation, for example, moving at least some document review work to an automated phase using NLP and machine learning software. But, of course, you can’t get into that unstructured data (i.e. text in documents) until it’s been digitised and made into a machine readable format.
I.e. deciding what can be digital is the first step and that raises a fundamental question about the law: what absolutely must remain analogue, with the digital realm at best acting in support, and what in the name of common sense and good economic practice should be moved into the digital realm, and then after that, where it’s ethical and economically beneficial, what needs to be automated with as little human (analogue) input as possible in order to drive up that automation’s efficiency?
Keeping It Human
People are often surprised to learn that despite this site’s name, the core philosophy of Artificial Lawyer revolves around the importance of what humans can do best, and helping that activity to increase.
Technology is there to extract inefficiency from economic systems, to drain it away from productive activities and processes. The goal of Artificial Lawyer, and many others in this space, is to help drive the movement of lower value tasks, which are a waste of human labour, money and time, i.e. a sink for economic productivity in general, first into the digital realm, then automate them so as little human effort is wasted as possible, allowing people to do what they do best, i.e. tap their very human skill sets.
Or, in this sector, this means allowing lawyers to reach their full potential as people who provide vital legal services. To not be low value paper shufflers and fleshy versions of what software can do 100 times faster.
But, there is also a flip-side to all of this. Lawyers are there to serve. They serve clients, naturally, but more broadly they are there to serve a need that society has as a whole. So, for anything that we move to digital, then automate, the question needs to be asked: is this helping society?
What Society Wants
Does human society want greater efficiency, a reduction in wasted labour, and legal services that meet critical needs in a timely and cost-effective way? Yes.
If tech, first via digitisation, then via automation, can help with this then that is ‘a good thing’ and we should push forward with it.
But, like all things to do with our species, it’s not that simple. This site does not believe that all things in the law should be digitised, or automated after that. Some things need to stay as human as possible for the good of society as a whole, regardless of the short-term economic benefits.
What is meant by that? Let’s take a high value example: judges.
Could we at some point develop machine learning systems to a point where even complex trials could be adjudicated by software? Possibly. But, that would be a very bad idea. As this site has explored, decision-making software is simply a crystallisation of biases, that we approve of or not.
A human judge is also biased, as all humans are. But, at least there is someone there, a person, who can be held accountable. Someone who really put their heart and soul, and reason and logic, and knowledge of the law, and a dozen other more abstruse elements into play, to come to a decision.
Automated systems passing judgment are fixed systems. They are also part of a ‘corporate’ entity, and as such are hard to be held accountable. Where does the company that made the software end in terms of the decisions that have been made? Who do they really benefit? How do you argue against a piece of software? How do you appeal a decision made this way?
Nope. That is not a good road to go down. (And, it’s worth adding that the same applies for the broad use of algorithms to make key decisions about people in general, from providing mortgages, to tagging people via automated surveillance systems in public places).
Good Digitisation and Bad Digitisation
As explored above, all digitisation, and then automation, is not necessarily good, or for that matter bad. There is a dividing line and it’s going to be one of the great battles of this century deciding where that dividing line is.
Digitisation is not neutral. It has an effect. It changes the relationship between people.
Let’s go back to the legal world again. Digital courts are a great petri dish for this debate. Technically we could make all courts digital, i.e. all video-based, all evidence digitised, and so on.
Some of that digitisation looks good and drives efficiency, such as digitising documents involved in a case, that allows in turn automated review systems, helps to link to automated research platforms and so on. It all helps, and the ethical risks are small.
Ending in-person court hearings is a different matter. This site is of the view that in-person court systems are a triumph of civilisation, and should be kept wherever possible. People should be in the same room when others give evidence for and against them. Juries should be a few feet away from the person they are judging – after all, this is not a TV reality show for their amusement, this is a real person in the dock. And judges should be physically there too, for any serious matter.
Now, here comes the dividing line again. Some judicial matters clearly don’t need a full, in-person court hearing, which is complex, cumbersome, slow and costly. For example, adjudicating on a parking fine appeal. But, where do we draw that line? Do we put a financial limit on it? That will have to be decided.
For the next few months the debate about digital courts is going all one way, because of Coronavirus, and it is a ‘no-brainer’ that if you cannot meet in person and want justice to keep going then remote/digital court hearings need to happen. But, that is not going to be carte blanche forever.
The same goes for giving advice via automated systems. Again, it may surprise some readers, but Artificial Lawyer is not…..repeat: not in favour of moving legal advice into a series of automated workflows.
Expert systems and Q&A dialogue systems are great for helping a client or lawyer get some very basic and early help with an issue, mostly on a very mundane factual basis. But, beyond that, let’s bring in those human lawyers and let them do what they are best at: listen, understand and provide bespoke advice fitting to a very specific and particular case involving a person.
And, while it’s fine to download a basic digital template from the multiplying doc stores out there, let’s not pretend that complex negotiations for a major deal can be entirely replaced with a cookie-cutter document. Again, let’s get the expert lawyers in the room to sort that all out. And again, it’s the dividing line.
Reducing Legal Work Costs So That Human Lawyers Can Help More
The point then is this: if we want more lawyers to do what lawyers are good at, in a way that is affordable, fast and efficient, then we need more digitisation, and then more automation where it’s ethical and makes sense.
The best legal services sector is one where everything that can wisely be kept analogue stays that way, and everything else goes the technological route.
At present the debate around these points is opaque and sometimes confused. We need to focus on what society needs. We need to digitise and automate where we should (and we’ve only just started on that really), and we absolutely need to keep human and analogue an enormous amount of activity too.
In fact, Artificial Lawyer’s hope is that as more basic tasks are digitised and then automated, human lawyers will be able to do more and more, face to face, with their clients in person, because that is where their real value is.
And, allowing people to show their true value is how societies prosper.
By Richard Tromans, Founder, Artificial Lawyer – April 2020.