AI: Where the Rubber Hits the Road

Topics: Artificial Intelligence, Business Development & Marketing Blog Posts, Data Analytics, Efficiency, Legal Innovation, Legal Operations, Legal Project Management, Midsize Law Firms Blog Posts, Practice Innovations, Process Management

AI

A Common Refrain:

Artificial Intelligence Will Replace Lawyers in the Next Five Years!

Thought no one to themselves:

Or, maybe it will just force them to change some of what they do every day? Hmmmm…

On the other hand, it may all be hype cooked up by tech companies and consultants. Though…

It probably represents an existential threat to the profession. Yet…

I suspect many lawyers are currently using it, even though no one really understands what it is, right?

What the Heck Is AI?

I find that many conversations about artificial intelligence (AI) in a legal context begin with a string of similarly incoherent and contradictory phrases. This is all the more frustrating because these sentences do not come from incoherent people. These are intelligent, articulate people who have become cognitively discombobulated through a steady drip of marketing mumbo jumbo, academic technobabble, and management misunderstanding. They are caught between vague demands to incorporate AI in their practice and conflicting information from experts.

AI optimists claim that it is massively disruptive and it’s widely available now, while AI cynics state that it’s hype and it’s a long way off. And less scrupulous vendors speak as if it’s magic that can do just about anything you imagine. The reality, as usual, lies somewhere in the middle. AI is not as here as I would like it to be, but it’s not as far off as you think it is, and it’s not as magical as one might hope.

Despite its ominous name and its depiction in movies as being omniscient, omnipotent, and typically evil, AI is none of those things. It is simply applied technology. And it’s not fundamentally different from other technologies we use every day. It’s not actually smart. It doesn’t think on its own or make unilateral decisions. It does not form opinions, generate insights, or “care” about anything — at least no more than Microsoft Word does.


AI optimists claim that it is massively disruptive and it’s widely available now, while AI cynics state that it’s hype and it’s a long way off. The reality, as usual, lies somewhere in the middle.


In addition, artificial intelligence is not one thing. AI is a category of technologies that all seek to perform tasks generally thought to require human intelligence. Ironically, as we become more comfortable with computers doing increasingly complex tasks, we no longer think of those things as requiring human intelligence and therefore real AI always seems to be just over the horizon and just out of reach.

You can see this effect in the development of self-driving automobiles. Most people agree that a truly self-driving car, in which you enter an address and take a nap until you arrive, definitely requires true artificial intelligence. But many cars today perform several aspects of driving on their own. They can parallel park themselves, automatically stop in an emergency, and even stay in a lane at an appropriate distance from the car in front of you at the push of a button. We don’t tend to think of those as AI, because demonstrably, they don’t require human intelligence anymore.

We assume that the real AI in personal transportation lies somewhere between the self-parking car of today and a magical future self-driving automobile. But the truth is, by the time any of us buys that magical car it will cease to be “artificially intelligent” and driving will just be a thing that cars do.

Breaking It Down

The tasks that AI can currently accomplish in a legal context have much more in common with the parking, stopping, and staying-in-lane capabilities of smart cars, than they do with future door-to-door automatic transportation. And while any attempt to make artificial intelligence easy to understand naturally involves a gross oversimplification of the technology, I find it quite useful to think in terms of just three simple tasks that artificial intelligence already performs for the practice of law today:

  • Pattern Recognition
  • Prediction
  • Performing routine tasks and reasoning

Pattern Recognition is probably the most common and most commercially competitive type of AI in legal. Tools like Kira, Eigen, eBrevia, Luminance, and RAVN all enable lawyers to mine the vast collections of documents they own to get to the underlying data those documents contain. These tools can help lawyers quickly find relevant clauses, extract key terms, and classify certain types of documents. None of them actually do the legal analysis and reasoning, but they make the drudgery of due diligence, contract review, and wading through mountains of documents simpler so that lawyers can focus on the more interesting and rewarding aspects of their profession.

AI

Predictive Technologies are the product of data scientists pouring through large data sets to find the patterns that occur in data over time — so, in effect, it’s still a form of pattern recognition. The scientists then create models that allow lawyers to manually enter key data points from a new matter to make predictions about how that matter will play out given the previous outcomes of similar matters in the data set. Tools like Westlaw Edge’s Litigation Analytics, Lex Machina, Premonition, and Ravel can provide insights into the likely outcomes of litigation or how a particular judge may rule on a motion given the specific arguments used.

Some tools employ both pattern recognition and predictive technologies such as with CaseText’s Cara, Ross Intelligence, and the Quick Check feature within Westlaw Edge. Among their capabilities, you can upload a brief from a new matter, they’ll analyze it, and return other potentially relevant precedents that you may have missed. The AI product market is nascent and constantly changing, so it’s possible there are other similar solutions out there, or that new ones will be developed.

Finally, there is AI that Performs Routine Tasks and Reasoning. That’s a mouthful, but it takes into account a wide range of rules-based tools like Expert Systems and Robotic Process Automation (RPA). Both of these technologies differ from the others mentioned in that they are not trained on large data sets like Westlaw Edge. Instead, they follow very clearly defined rules to reach conclusions, produce documents, or replicate a series of steps that would otherwise have to be performed manually. Tools like Contract Express, Checkbox.ai, Autto, Neota Logic, and Bryter are low-code expert systems that walk a user through an interview process to determine whether they are compliant with new regulations, or to ensure that they are correctly and comprehensively filling out forms, or other similarly well understood processes. RPA tools like Blue Prism and Automation Anywhere are similar to macros in Word or Excel, except they work at the desktop or internet level.

These tools can automate those otherwise hard-to-automate processes that only that one secretary in accounting knows how to do. (Open this file, copy this text, paste it into this website, take the output, paste it in a new document, upload that document to this repository, and notify these five people that the new document is out there.) While it may only take one person 20 minutes to perform those steps, an RPA system can do the same in seconds, freeing up people for more productive and profitable activities.

Can I Pre-Order my Lawyer Bot on Amazon?

While artificial intelligence does not currently provide the robot lawyer of the future that so many are fearfully (or eagerly) awaiting, that does not mean that AI tools are not useful assistants that can be incorporated into legal practice today.

How often do lawyers need to recognize patterns, make assessments of likely outcomes, or perform routine repetitive tasks? These are all areas where current AI-enabled technologies may be helpful. And thinking of AI in terms of those three tasks, provides a pretty good heuristic to determine whether AI tools might be useful in your practice.

Just as the standard driver-assist technologies like parking, stopping, and stay-in-lane cruising can make you a better driver, lawyer-assist AI technologies can make lawyers more productive and more fulfilled, while making law firms more profitable.