Zapp2Photo/Shutterstock.com.

 

There's no shortage of technology providers offering machine-learning solutions to age-old business problems. Now, as organizations in various industries continue hiring for tech-focused roles, 2018 may be the year machine learning truly takes off.

Among the roles increasingly sought out by businesses is that of the data scientist. As reported by Forbes in spring 2017, an IBM study found that by 2020, there will be nearly 62,000 jobs for data scientists and similar positions in the U.S. And according to Andy Abbott, chief technology officer and co-founder of contract review technology company Heretik, these scientists will undertake more research in 2018 so that companies remain competitive.

“This past year we have seen a significant influx in attention on machine learning and data science overall,” Abbott told Legaltech News, leading to increased focus among companies, many of which, he says, are now hiring their first data scientists.

“They typically do not know the goals of building out a data science team, but fear they will be at a competitive disadvantage if they do not begin to research them,” Abbott said. “I think the data science industry is still very much at its infancy and will continue to be throughout 2018. However, we will begin to see its real impact as its deployment continually increases.”

IBM's report found that the demand for data and analytics roles depends on the industry, with the greatest demand being in finance and insurance and professional services. In the past several years, the legal industry has seen a steady increase in data science roles, with law firms such as Drinker Biddle & Reath and Littler Mendelson staffing data professionals.

For modern legal service delivery, you now not only need lawyers, but also data scientists, UnitedLex vice president of legal solutions Peter Krakaur told LTN. He also noted that his company will be looking at how artificial intelligence “augments delivery of legal services” and how that will change as the industry “shifts towards more automation and smarter systems.”

Abbott also said that international government and industry regulatory changes may also drive a spike in machine learning.

“Our clients are quickly trying to prepare for the impact resulting from what's happening in Europe with Brexit and GDPR [General Data Protection Regulation], and also upcoming accounting practice changes such as ASC 606/IFRS 15 and IFRS 16,” Abbott said.

Also, there is an increasing “burden on legal departments, transactional and corporate practices, attorneys and legal operational staff” to implement machine-learning solutions.

“The amount of material that requires diligent review has surpassed the capacity of the manual workflows reviewing it and [is] similar to where e-discovery was 10 years ago,” Abbott said.

In addition to implementing these methods internally, businesses are requiring external counsel to use machine learning during review.

“Corporations understand the significant benefits of using machine learning, including the reduction of risk resulting from human errors and the increase of review coverage when time is a constraint, as it often is,” Abbott said. “Corporations are also increasingly unwilling to pay steep fees for repetitive simple operational tasks like contract abstraction. They are, on the other hand, comfortable incurring higher costs for more strategically impactful work, something machine learning is empowering attorneys to focus on.”

Yet machine learning implementation isn't always met with open arms. Abbott said that a considerable challenge to machine learning adoption is “trust,” noting that part of product development at Heretik, which leverages machine learning, is geared toward addressing this point. This includes having “an open conversation” with users and providing “additional functionality” such as “checks and balances in the user interface” and “detailed audit logging of analysis” to ensure trust.

Still, Abbott pointed out that industries are “very early in the life cycle of AI and machine learning.”

“During this beginning phase we'll see many attempts to mimic and simply automate existing processes as they exist today. However, those processes were developed and perfected with the tools at hand. As we introduce new tools and techniques, facilitated by machine learning, processes will evolve. Maturity will have reached the industry when we can look at a process and see no resemblance to how it was done it the past,” he said.