James Williams, a commentator on the article “Meet ROSS, the bankruptcy robo-lawyer employed by some of the world’s largest law firms”, laments the lack of reference to supporting research, or consultation with the “40 year old research community that has long focused on the use of technology in a legal setting” in Toronto. My SLAW column “The Long Tale of 2 Systems” highlighted the role both Toronto and Sydney played in the development of legal technology.
So while IBM’s Watson was mentioned a few times at the 10th annual Sinch Legal Tech Sydney Conference (#SLTS16), the stars of the show were the pioneers of legal AI, and other smartening tools and methods for lawyers. The good news is that the legal insights, algorithms and methods explored over decades can finally be put to practical use now that market acceptance, adequate hardware processing power, communications and data sets can be added to the knowledge mix.
Once again, the legal community was treated to a special, and challenging day with over 30 speakers at SLTS16. Here are some comments from just a few:
Mother and son team of Pamela and Xenone Gray and developers of the superexpert system shell eGanges pointed out that big data analysis uses ordinary language deconstructed algorithms, not legal language and interpretation deconstructed algorithms. They raised the issue of the need to police systems like Watson. Experts capable of recognising ‘red flags’ and signs that an error may have occurred – will be as important as ever. Why not go one step further and use an expert system to keep deep learning systems in check? Their paper, “eGanges: Policing Watson” can be found HERE.
Nicks Brestoff supported my view that maybe a better place for AI is to help prevent litigation. There is a potential huge payback for society, and corporations investing in early warning systems. Nicks research quantifies this view.
Richard Granat reported in that the US legal marketplace is “stuck like a clogged pipe”. He wen on to say:
“Only a law firm is allowed to provide legal services. The ethical rules prevent outside investment or ownership in a law firm. Law firms operate inefficiently and lack the necessary incentives or means to innovate. This inefficiency hinders the flow of legal services from law firms to clients. Legal services become overpriced and difficult to access. Demand in the marketplace suffers. Law firms struggle to grow and new lawyers can’t find work. And the cycle continues…”
Rhondda Nicholas was speaking with experience when pointing out the handcuffs on innovation in Legal Services. She suggested a business structure to operationise that great legal idea, i.e. As lawyers are very regulated, set up a company to deliver services to a law firm. Some jurisdictions may have made improvements but are really in the 19th Century as they have so many restraints on innovation. In previous years she tried to modernise her jurisdiction, but is now suffering from the “payback factor”. The last hurrah of the old guard. But also, she asked, why are old ways still enshrined in the legislation? For example, it can be a challenge to get some form of online initial payment. How that initial payment is characterised is important as you may be dealing with trust money. It is so “last century” in that your regulator is determined by where your office is. It prompted one previous speaker to set up his business as a company, not as a law firm as his Law Society thought his initial payment would have been trust money. Unfortunately, a lot of legislation is premised on the fact that lawyers are actually meeting clients face to face, which is now rare with online firms.
Despite the ongoing challenges, Greg San Miguel noted that this year, the legal tech landscape has all changed. For the first time in 15 years of providing systems, he can now prove value, and derive metrics that demonstrate the success of his initiatives. So it is a great time for legal tech in his broad use of the term.
His area is where law is touching business users. The trend is against lawyers providing the traditional monopoly services. So, how can lawyering drive a bottom line benefit?: Embed law into a much larger process. What is the opportunity for law and lawyers in this environment? He gave an example of a General Counsel putting their sales team onto “front foot strategies” by training them to go into an area where traditionally it was viewed as too vexed to even consider. Equipping their sales people with the skills to deal with potential problems became a perfect marriage of legal and business.
When Philip Argy was a Trade Practices specialist with the largest firm in Australia in the 1980’s, too much of his time was dealing with “quick questions” from colleagues which did not help his billing targets. His solution was to set up a PC outside his office that could answer a lot of the questions he was regularly asked. If the expert system he developed couldn’t provide the answer, he would then deal with it personally.
Being an engineer, as well as a lawyer, meant he was certainly not shy about evangelising IT. Buoyed by the success of his own expert system, he worked with some of the partners to capture their knowhow. The biggest fans of the systems were the experts themselves: on just about every matter, the system reminded them of something they had forgotten to address. It turned out that the experts whose knowledge built the systems, were the greatest advocates of these pioneering smart machines.
Professor Tania Sourdin, dealt with the research question …”Why do some civil cases end up in a full hearing?” The raw data (after culling the sample cases) suggests that about 6.5% of civil cases end up with a final decision in the NSW Supreme Court and that in Victoria, about 4% of civil Supreme cases end up with a final decision being made. The percentage of final civil cases heard by a judge is much smaller and approaches 1% when some case types are excluded. The data mining techniques used were IBM Watson Analytics.
In the higher civil courts, judges often note that the disputes that end up being finally litigated will often have high levels of task complexity (that is they may involve large amounts of information, complex transactions and multiple participants).
In addition, litigated cases may involve high levels of behavioural complexity, that is, those that are involved in the dispute may be less likely to behave in a rational manner and be more likely to engage in excessively adversarial behaviour.
Marc Lauritsen pointed out that while the virtues of databases, document assembly, and expert system are well known, none of these are particularly good for helping people make choices. Choices involve competing values and perspectives of the people involved. They’re not algorithmic. You can’t look up or compute the answer. His introduction to a tool he called Choice Boxer was compelling. While not a new idea from Marc, Choice Boxer certainly seems to be maturing.
Dominic Woolrych introduced us to a “RoboLawyer” – an expert system that collects data to build a profile on a client from the content of their documents. Based on this profile, the system can then react and suggest actions based on predetermined rules and external factors. Due to cost, clients see a lawyer as a reactive step when something goes wrong. The RoboLawyer will sit in the background and proactively foresee risks to your business and then advise accordingly.
- it “was created to explore how legal services might be delivered in the future”,
- it is not a lawyer (they’re human), and
- that it was in Beta.
It then walks the tightrope of pointing out that while they have gone to great lengths to ensure that the information provided is accurate, it is not legal advice.
Michael Curtotti reported on research that found that as the more diverse the vocabulary of a sentence, the more readable it is, hence due to the need for consistent terminology, this makes legal documents less readable than other text.
So this year’s Conference was a nice balance of the theoretical and historical meeting the practical as we start the real journey for smartened lawyers. Seat belts will be needed over the next few years, particularly as the legal tech revolution seems to be driving with a mind of its own.