Attorneys Take Note: Artificial Intelligence is About to Shake up the Legal Landscape

by | Oct 18, 2018 | Legal Tech Software, Software Reviews, Useful Tech Tips | 0 comments

Artificial intelligence has arrived.  It is up to the legal industry to gradually evolve in a high-tech manner.  Commonly referred to with the acronym of AI, artificial intelligence is based on a single type of technology known as deep learning.  AI is available in several forms, some of which are proving quite helpful to attorneys and others in the legal industry.



Rule-based Systems AI


Rules are as dry as it gets yet they are essential to making machines think similar to humans.  All sorts of businesses processes hinge on rules.  Rules are essential to countless applications across every level.  This AI form has the potential to become quite complex when rules are applied in an interactive manner.  One rule’s output can be the input to another rule and so on.  If something goes awry, it is easier to pinpoint the rule that caused the issue.  Certain AI technology requires a rule-based system to function.  As an example, some services provided by digital assistants are the result of rule-based tech.



Deep Learning


This unique form of machine learning functions with a recurrent neural network comprised of algorithms that can nearly match that of humans.  In fact, some such algorithms have proven superior to those created by humans.  The only rub to deep learning is that considerable amounts of information are required for it to work.  Therefore, deep learning has limited applicability to law firms and other businesses.  However, those who understand deep learning’s functionality will find it really can help a legal case.



Supervised Machine Learning


Supervised machine learning is sometimes contrasted with deep learning.  The supervised form does not require as much training information, making it that much more easy to use.  Furthermore, supervised machine learning does not require as much thought for preparation compared to classic machine learning models.  In most cases, a few dozen examples are required to create a model.  Machine learning is contingent on feature extraction based on language.  It can be challenging to pinpoint the training document that caused the outcome when using vector machines.  Additional training input is necessary for a machine learning model to be properly analyzed and corrected.


Supervised machine learning, deep learning and rule-based systems AI are all here to stay.  Attorneys who take the time and invest the effort necessary to understand these technologies will benefit in the long run.  AI will become that much more intertwined with life and work as we move toward an increasingly tech-focused future.






Submit a Comment

Your email address will not be published.

%d bloggers like this: