Getting a head start in the new AI economy
Photo: The CBA LC AI Panel: Noah Waisberg, Dera J. Nevin, Omar Ha-Redeye (moderator)
Like many advances in technology before it, new developments in artificial intelligence are driving businesses across a number of sectors, from finance to the insurance industry, to gain a tactical advantage on the competition.
So why should the practice of law be any different? That’s the question for a growing number of legal innovators.
Indeed, today’s machine-learning algorithms and techniques can be used to do much of what lawyers routinely do at work, namely reviewing, analyzing and synthesizing vast amounts of data. They can also perform risk and outcome analysis. The more sophisticated tools – think IBM’s Watson – can even reproduce human decisions and make correct predictions based on new, unseen data.
These developments may be unsettling to some law practitioners, but not to a growing group of “early adopters” who are embracing what they view as a new source of competitive advantage over traditional firms.
“If you could give your clients man advantage in court and increase their odds of winning, wouldn’t you?” Dera J. Nevin asked a crowd of lawyers sitting in on a panel discussion on AI at last month’s CBA Legal Conference in Calgary. “If it was ethical, of course you would.”
Nevin, a director of eDiscovery Services at Proskauer Rose LLP in New York, welcomes the advent of machine learning as a deliverance of sorts from the nightmare of drowning in data, adding that she isn’t worried in the least about machines replacing her job (Read her recent opinion piece for National here). “Machine learning has replaced a lot of things that I used to do, but my age has also replaced a lot of things I used to do,” she says. “I don’t want to do the same thing as a 20-year practitioner as I did as a first year practitioner. “
More importantly, she sees enormous potential in how AI could revolutionize legal practice. “There are some really interesting opportunities,” she says. “If we’re creative enough to think through some of the current blind spots in our service delivery, there’s a lot of potential there.”
Kyra Systems’ CEO Noah Waisberg, also a panellist in Calgary, sees opportunity for the more experimental providers of legal services. “You need to use AI right now while you can still get a competitive edge,” he told attendees, adding that in 15 years time, those who had initially been behind the curve will have caught up to those who were a step ahead. “The key is to take advantage of the spread. Right now is the time when you can see the opportunity to use these tools that can tell you information that not everybody else knows.”
Some firms are already poised to take advantage of the most recent advances in legal technology. The global law firm Dentons made headlines this summer when it announced it was teaming up with ROSS Intelligence, a start-up developing a legal adviser app powered by IBM Watson, the famous robot that appeared and won on Jeopardy. ROSS is an application designed to answer legal questions thanks to machine learning.
“Obviously Watson wasn’t built to win a few hundred thousand dollars on Jeopardy,” said Ian Kerr, holder of the Canada Research Chair in Ethics, Law and Technology at the University of Ottawa. “This is seen by IBM as a potentially huge business,” he told the crowd at the panel session. “What we’re seeing is that law firms like Dentons are starting to buy into this, literally, by trying to fund ROSS.”
How long, then, before the bulk of routine tasks performed by lawyers is taken over by robots? It’s still early days and it won’t happen overnight, said Nevin. “A lot of the AI right now is not perfect. It’s not going to do what a human does. It’ll do some pieces of what a human does better but it’s very much about working in partnership with the AI and building interfaces that allow you to forge these partnerships. “
Thinking about AI in terms of machines replacing humans misses the point, according to Waisberg. It isn’t about machines or humans working along, but the two working together effectively. “The better measure is what happens to a person using our software reviewing contracts versus a person reviewing contracts without it,” said Waisberg discussing his company’s flagship product Diligence Engine, which does automated contract review.
That said, even if it does take time before machines develop a strong track record at carrying out tasks without the intervention of a human being, that will inevitably change, said Kerr recalling headlines two years ago claiming that IBM’s Watson, applied in the medical field was better at diagnosing cancer than human oncologists.
It isn’t too early to start thinking about the day that AI will be able to mirror human judgment, and the implications of delegating certain tasks to machines.
“The thing that’s important to understand is that Watson already works in such a way that its programmers can’t always predict the outcomes that it generates,” Kerr says. “And so as we have more and more unforeseen, unpredicted outcomes —whether it’s a driverless vehicle, a Google car, whether it’s a Watson-type device, whether it’s a ROSS-type device. Once we move down that road all of a sudden we have a lot of new challenges.”
Kerr argues that AI is certainly appropriate for certain tasks like contract review, and some aspects around litigation documents. “But do I want a machine to be the sole decision maker, for example, in determining whether I am a contractor or an employee, or whether I am someone who was wrongfully dismissed? And certain people are starting to propose we try to use things like ROSS to do it down the road. Once we get into the realm of those kinds of decisions, where ultimately what we’re talking about is judgment, we then have to ask the question: Well, what exactly is it that the machine system is now doing? And is that the right thing to delegate to that machine system?"