By Mark Homer
Last month, I discussed how we have used technology at GNGF to provide more enjoyable, efficient work for our team, and thus more value for our clients. I went so far as to say:
“There will constantly be the opportunity to utilize machines to provide more efficiency to my team, and in turn, more value to my clients. Technology, when implemented well, will allow our team to increasingly spend more and more time on high-level strategy and proactive communication.”
One interesting technology and strategy that I have been keeping a close eye on the past couple of years is Cognitive Computing often referred to as Artificial Intelligence.
When IBM’s Deep Blue beat the world chess champion, Gary Kaparov, in 1996, I speculated about what complex, strategic game computers would beat next. Then 15 years later, in 2011, IBM’s Watson beat Ken Jennings at Jeopardy, using natural language processing and semantic algorithms to create an answer with a confidence score. In 2016, 20 years after IBM’s Deep Blue won a game of chess, Google’s DeepMind AlphaGo computer beat the champion player of Go, which is considered the most complex board game.
While it is fun to watch these games, and they are certainly great PR for Google and IBM, the benefit for everyone is the progress and utilization of cognitive computing to help knowledgeable workers solve complex problems. By combining a computer’s computational speed with significant advances in cognitive computing algorithms and technology, computers can work on complex problems that we previously reserved for well-educated, white-collar workers. Business analysts, medical professionals, government advisors, and others are finding that much of their knowledge can now be provided by a computer. Already, Watson’s technology has moved from interacting with Alex Trebec to helping doctors and nurses diagnose patients. In 2013, IBM Watson’s business chief Manoj Saxena indicated that 90% of nurses in the field, who use Watson, were following its guidance.
As a marketer, I stay in touch with the marketing technologies that are utilizing cognitive computing. These technologies can do a task, identify patterns, make an educated guess, test the guess, and adjust all faster than a person could do the initial task.
Two promising applications of these technologies I hope to use for our clients, which are already being utilized by the largest brands, are Predictive Content Curation and Ad Targeting. We produce a lot of content for our clients. Tools that can help predict content that is on the rise in discussion can help us produce content that provides better results. If we can get ahead of potential long-tail searches, we can take those suggestions and utilize our workflow tools to produce content relevant to hot, trending topics and get clients ranking faster on terms that no marketer has even thought about yet.
On the ad targeting side, machine learning used for optimizing bids (and even adjusting some of the ad copy) to achieve the best cost per acquisition is showing promise that, under the watchful eye of an experienced SEM strategist, it can provide a better return on ad spend in less time.
Once you tackle some of the basic practice management technologies we discussed last month, then you too can be on the lookout for how cognitive computing / AI is showing up in the legal world.
Right now there is a fast growing upstart named ROSS Intelligence. It is using IBM Watson cognitive computing to provide smarter, efficient legal research. Instead of just searching keywords and bringing back a lot of information, much of which is irrelevant to you, ROSS learns over time and senses patterns that match the type of query a lawyer is entering and what should be returned. It then keeps track of the query and will provide updates during a case if it finds anything new. It won’t do the high-value thinking and discussions with your clients, but this could be the next tool that saves you a lot of time and allows you to provide value to more clients.
Taking things to the extreme, the “robot lawyer” chatbot, DoNotPay, created by 19-year-old Stanford University student, Joshua Browder, helps users contest parking tickets in an easy to use, chat-like interface. As of June, this new program had helped people appeal over four million dollars in parking tickets in just the initial launch in New York and London.
In an interview in The American Lawyer, Robert Weber, general counsel for IBM, and a former litigation partner at Jones Day, said that while he doesn’t expect Watson to replace the judgment of a senior law firm partner it could handle tasks of senior associates. He sees it researching, writing memos, and suggesting the most persuasive arguments and precedents. It could also quickly review stacks of contracts, looking for differences in indemnification clauses. He went so far as to say that “I think Watson could pass a multistate bar exam without a second thought”.
Look at areas where we already see forms of high quality natural language processing, or some machine learning, like legal research, e-discovery, compliance, contract analysis, and case prediction. Just a couple of years ago many of those tasks required someone with a law degree and limited experience. Now they are starting to be done competently by machines. Machines can not only provide lawyers answers to queries that often took a day of research, but can free lawyers to spend their time doing higher value and often more interesting work.
Leave a Reply