legal technology

Understanding the New Language of Legal Technology

The terminology surrounding today’s legal technology, particularly with respect to artificial intelligence, can be confusing. On a daily basis and in a variety of contexts we now routinely encounter terms like augmented intelligence, cognitive computing, analytics, natural language processing, machine learning, and more. The terms aren’t used consistently from one resource to the next, and sometimes they seem to be used interchangeably.

How important is it to keep all these terms straight? Once you remove the marketing hype, what actual problems do these technologies solve, and to what extent are they being used successfully in the legal profession today?

 Artificial Intelligence

“Artificial intelligence” (AI) is a broad term that describes a group of technologies designed to perform tasks that once required human intelligence, interaction, or decision-making. Among the more prominent applications of AI technology are speech and image recognition, self-driving cars, computer-aided interpretation of medical imagery, algorithmic stock trading, and robotics. That may seem like a diverse set of use cases—and it is—but this list represents only a tiny sliver of the full range of AI applications in use today.

In the context of the legal profession, there are already dozens of technology applications using AI to assist with tasks like legal research, e-discovery, document review, billing, managing and analyzing contracts, and mining litigation data for strategic insights. That said, there are two very important points to keep in mind:

  1. While AI technology is being deployed to help legal professionals achieve better or faster results in the work they do, it is not replacing those professionals, nor will it in the foreseeable future.
  2. Although all of these applications involve processing and manipulating very large amounts of data very quickly (an area where humans can no longer compete with machines), each is focused on a specific computational task, or a narrow range of tasks.

Today, no AI application possesses the uniquely human ability to synthesize multiple skill sets and make reasoned judgments the way lawyers do.  When they are collaborating with colleagues to assess a legal matter, negotiating with adversaries, developing case strategies, building arguments, and presenting a persuasive case, attorneys are leveraging uniquely human skills. AI applications cannot articulate a thought process, nor can they explain why a particular solution or decision is the correct one.

That said, attorneys can now leverage AI-based legal tools that can effectively mine large data sets, to identify drafting errors, predict likely motion and legislative outcomes, and improve the quality of research by better understanding user query intent. These AI-infused solutions are in the market today in many of the tools you rely on day-in and day-out.  Make no mistake, these solutions are changing the practice of law in meaningful ways.

Augmented Intelligence and Cognitive Computing

The excitement currently surrounding AI, and the prospect of “robot lawyers” rendering their human counterparts obsolete, probably explains the more recent emergence and growing popularity of the term “augmented intelligence.” Augmented intelligence refers to the same set of technologies that comprise AI, but the term is preferred by some parties because it steers us away from the misconception that AI somehow replicates or substitutes for human consciousness. I am strongly in this camp. For the foreseeable future, AI systems only help humans do certain kinds of complex, data-intensive work much more quickly and efficiently.

“Cognitive computing” is another term that is sometimes used interchangeably with AI and augmented intelligence. While there is no widely agreed-upon definition of cognitive computing, the underlying technologies are often the same as those used in AI. The difference seems to be a matter of emphasis. Proponents of cognitive computing like to stress the importance of the interaction between humans and machines. Artificial intelligence, from their perspective, is about machines doing powerful things with data to come up with answers and insights. Cognitive computing, by contrast, involves humans in the computer-learning loop. As with augmented intelligence, the emergence of the term cognitive computing probably has more to do with explaining the human-led computer learning process than with clear-cut differences in technology.

Analytics

Is analytics synonymous with AI? The short answer is no, although there is plenty of overlap when you look at the actual application of analytics. Analytics can be defined as information—or insight—derived from the analysis of data.

Advanced analytics applications can take massive volumes of data, structure it, eliminate irrelevant or redundant information, and make it easily searchable. Analytics allows users to find very specific kinds of information or insights in a few minutes that could take humans weeks or months or more to compile. Some analytics applications are based solely on statistical analysis and don’t deploy AI technologies at all. But advanced analytics may involve a variety of sophisticated tools or techniques, some of which are also deployed in AI applications.

For example, a legal analytics technology in broad use today uses machine learning and natural language processing to clean, tag, and structure raw federal litigation docket data so it can be easily searched and reveal new insights. “Machine learning” is deployed to comb through very large amounts of data in order to identify patterns and appropriate action based on those patterns. It uses algorithms that can “learn” from data in an iterative fashion and make predictions by building a model from sample inputs. “Natural language processing” helps IT systems process human speech as it is written or spoken, without requiring users to change their syntax to conform to system requirements. It helps software discern the intent of a user’s question, and it identifies key phrases and information within unstructured data that are critical for better search results.

In the practice of law, analytics is helping lawyers develop litigation strategies based on historical outcomes of cases with similar fact patterns. For instance, lawyers can use legal analytics tools to get facts-based insight on the behaviors and tendencies of specific judges and opposing lawyers and parties in legal matters similar to the one they’re currently litigating. In the business of law, analytics can help legal departments compile a list of firms with the best track record in litigating specific kinds of cases in specific legal venues, or help firms identify litigation trends to inform the establishment of business development and marketing priorities.

From a functional perspective, it may be useful to think of analytics in the legal domain as being primarily about helping lawyers make better, data-driven decisions, where AI applications are essentially about automating language analysis and complex tasks in ways that we once thought only humans were capable of. Distinguishing between technological categories like AI and analytics is probably a lot less important for today’s lawyers than educating ourselves about specific tools, regardless of the underlying technology, and their uses in our daily workflow—as well as their current limitations.

Putting It All Together

In summary, artificial intelligence is a broad term that describes a group of technologies that automate complex tasks we once thought were the exclusive domain of human intelligence. Augmented intelligence and cognitive computing describe similar groups of technologies, but these terms have emerged to emphasize different aspects of AI that tend to get overlooked.

Analytics is the insight gained from the mining of vast amounts of data to uncover trends, patterns and behaviors. Many of today’s most useful analytic applications deploy technologies that are also used in AI applications. Examples of such technologies include machine learning (applications that identify patterns in large amounts of data and “learn” and improve results without being explicitly programmed) and natural language processing (software designed to process and understand the intent of words as they are naturally written or spoken). Machine learning and natural language processing are typically associated with artificial intelligence applications, but they also are key components of advanced analytics.

Neither AI nor analytics, nor their any of their many permutations, are close to replacing the skills, judgement, and experience that are the foundation of a lawyer’s competence and effectiveness. But these emerging technologies are coming together in exciting ways that are already transforming the profession by helping legal teams provide more value to clients more efficiently and at lower cost.

The best expressions of these technology will be invisible to the lawyer; said another way, when done right, the technology simply delivers better results and better legal insights.

About Jeff Pfeifer

Jeff Pfeifer
Jeff Pfeifer is vice president of Product Management for LexisNexis. Over a 28-year career in legal technology, he has worked to introduce a series of cutting-edge solutions for lawyers and other legal professionals. He is responsible for the product development strategy for LexisNexis in North America and is leveraging artificial intelligence to deliver better outcomes for his customers. Follow him on Twitter @JeffPfeifer.

Check Also

Legal Technology

Visualizing the Future of Legal Technology

Visual analytics and other legal technology are currently shaping the future landscape of the legal industry as we know it.