The Difference Between General AI and Narrow AI

By: Jonathan Nessler

Artificial intelligence is changing the way law firms operate.  Recent breakthroughs in a type of artificial intelligence, called "narrow" or "weak" AI, make it possible for law firms to increase efficiency while lowering costs.  Attorneys need to embrace artificial intelligence and incorporate AI based technology into their firm’s daily operations.  Firms that refuse to learn about and integrate the benefits of artificial intelligence technology risk falling behind competitors who are already using this technology to augment company performance.  If an attorney uses artificial intelligence, he has an obligation to understand the technology. 

There are two types of artificial intelligence: (1) general artificial intelligence, sometimes called “strong” or “hard” artificial intelligence; and (2) narrow artificial intelligence, sometimes called “weak” or “soft” artificial intelligence.  While “general” and “narrow” artificial intelligence are often referred to as “types” of artificial intelligence, these concepts are more accurately conceptualized as different approaches to solving the problem of artificial intelligence.  Each solution has its own set of significant problems to overcome.  However, development of reliable narrow AI has experienced rapid growth in the past several years, and is very promising as a present and near-term solution for law firms seeking to implement reliable and beneficial artificial intelligence based technologies.  

The problem presented by artificial intelligence is how to create a computer system that is capable of behaving intelligently.  The ultimate goal is to create a computer system that is capable of rationally solving complex problems under a variety of real-world circumstances.   Borrowing from legal phraseology: The issue is, "how to create a computer that will think and act like a reasonably intelligent person under the same or similar circumstances as those presented in the real world."

General artificial intelligence attempts to solve this problem using a traditional vision of artificial intelligence.  This solution is akin to the vision often depicted in fictional books and movies.  Hard artificial intelligence solves the problem by creating a computer that thinks like a human.  This solution is to create artificial intelligence that thinks and acts like a reasonable person under the circumstance by creating a machine with a general intelligence that is broadly applicable to whatever problem might be presented.  Conceptually speaking, this is identical to how human intelligence works.

Narrow artificial intelligence, on the other hand, solves the problem of artificial intelligence by focusing on creating machines that are able to do specific tasks that are, historically speaking, tasks that only a human can perform.  This solution does not require the machine to think like a human, nor does it require the machine to exhibit broadly applicable intelligence.  Instead, this solution is to create a machine that is able to learn and perform a narrow task, like searching through thousands of pictures and separating out the pictures that contain a certain image.  

Narrow artificial intelligence is the type of artificial intelligence that will benefit businesses, including law firms, in the present and near future.  Narrow AI is undergoing rapid growth in development, and scientists are witnessing breakthrough advances in development of narrow AI.  To the contrary, development of hard AI is proving to be much more difficult than most people anticipated. 

Law firms are already integrating narrow AI into daily operations.  By integrating narrow AI, law firms are generally lowering costs and creating work product more efficiently.  In the law firm setting, narrow AI is capable of performing a narrow task at a rate much faster than a human is capable of performing the same task.  For instance, narrow AI is capable of learning the relationships between words and legal concepts.  This allows the AI to search through thousands of pages of documents and separate out the relevant information.  As another example, by using predictive analytics, narrow AI is helping firms predict likely outcomes of oral arguments, trials, appeals, and other litigation. Law firms are also using narrow AI to increase the efficiency and effectiveness of firm marketing.  As research and development of narrow AI moves forward, the number of areas law firms might benefit by using AI will continue to increase.

Law firms are businesses. Like all businesses, law firms must cut costs where costs might be cut.  Artificial intelligence presents an unprecedented opportunity for law firms to increase efficiency and productivity, while decreasing costs.  Law firms will continue to incorporate artificial intelligence into firm operations, because the potential cost savings and increase in efficiency are proving to be a competitive business necessity.