Over the years, machines have replaced humans to handle predictable tasks. Tasks like collecting money at toll booths, managing customer service calls, and even vacuuming your living room floor can now be completed by a machine. But so far, machines have left humans to take on unpredictable tasks that require creativity, problem-solving, and intuition based on human experiences.

As machines have become a part of everyday life, some folks have reasoned that robots could never replace what they do. I hate to break it to you, but artificial intelligence (AI) is now creating art and tech companies are raising millions to advance AI-powered solutions for symptom analysis and patient triage.

So, is it possible for AI to replace lawyers?

Today, machines have learned to carry out tasks we could hardly imagine 20 years ago. AI uses statistical patterns in data to improve the efficiency of many different work processes. All AI needs is data, and there’s plenty of it available in the legal field.

It is possible that AI will optimize the legal field and reduce the number of errors in cases. It is possible that human lawyers take on an overseer’s role and only problem-solve when things go wrong.

I’ve thought about this possibility a lot in my work creating an AI-powered immigration law chatbot. Most lawyers I’ve talked to about this think it’s impossible for AI to fully replace human lawyers, or that that moment is impossibly far off in the future. But why are lawyers so convinced that they can’t be replaced?

Psychology plays a role in our reluctance to accept robots

It could be the uncanny valley effect, which is the uneasy feeling you get when a non-human form such as an avatar, robot, or animation is just a little too realistic and lifelike. It’s called a valley because on one hill is cartoon-ish animation that is clearly non-human, like Fred Flintstone. On the other hill is a real-life human, or an AI Instagram model so lifelike that people are fooled into thinking it’s truly human.

In the middle is the uncanny valley. A good example of the uncanny valley is the creepy baby Billy in the short, animated film, Tin Toy, produced by Pixar in 1988. Another example is the 2009 computer-animated adaptation of Charles Dickens’s A Christmas Carol which was described as creepy and off-putting by critics.

The chasm between looking “almost human” and “fully human” leaves people feeling a sense of unease, strangeness, disgust, or creepiness. It is possible that human lawyers feel a sense of revulsion thinking about an AI lawyer taking their place.

It’s also possible that AI is taking too long to successfully complete the complex tasks required to be a good lawyer. This idea, that technology was advancing but then was just not good enough for a long time, is called the trough of disillusionment.

What is the trough of disillusionment?

Here’s the story behind it. Gartner, a technology research and consulting company, created what’s called the hype cycle to track emerging technologies. The hype cycle has five key phases and shows us how a technology or application will evolve over time. At one end of the hype cycle is the Innovation Trigger where potential technology breakthroughs start. At the other end is the Plateau of Productivity where mainstream adoption starts to take off

Somewhere in the middle is the Trough of Disillusionment. Lawyers see what AI has done in other fields, but implementations in the legal field have failed to deliver. The improvement of existing tech has stalled, and new tech isn’t readily available. Simply put, lawyers are unimpressed and think that legal tech has hit its limits.

Are the disillusioned lawyers with the creepy uncanny feeling right about the future of AI in the legal field? It’s possible, but there are a couple of things to consider before we write off robot lawyers for good.

AI lacks intuition to make decisions about the unknown

Machines are designed to strictly follow rules, they simply don’t have the intuition that humans use to make logical leaps. Human lawyers rely on experience and intuition to solve unknown problems and can make decisions. This is something computers can’t do right now, but that doesn’t mean it’s impossible.

When humans have incomplete knowledge, we make decisions based on intuition. Strong intuition is an admired trait in humans, but we’re often unsettled at the idea of a machine that can use intuition or less than complete data to make a decision.

Computers process enormous amounts of data to function. Given the right kind of data, it is possible that AI can actually make highly intuitive decisions. AI has no bias so it can make decisions based on every minute detail, observation, and influence. In fact, there are people who believe that it’s possible to train AI to make more intuitive decisions, and there are researchers at Aarhus University currently trying to combine our human intuition with an AI’s ability to quickly access troves of information.

Machines have been replacing human workers for centuries

While some believe that it’s impossible for AI to gain the human-like intuition needed to be a good lawyer, others think it’s simply impossible. I hate to be the bearer of bad news, but not too long ago, people thought it’d be impossible to replace the horse.

For centuries, horses were the backbone of modern civilization. Horses stayed in the labor force through many new technologies that threatened their standing. For example, telegraphs replaced long-distance delivery men on horseback and trains replaced cross country horse-drawn carriages. And still, horses did work on farms and hauled people around cities.

That is until the internal combustion engine hit the scene, and suddenly millions of American horses were unemployed.

Machines have also replaced humans over the decades. Buttons replaced elevator operators and the internet drove travel agencies out of business. Between 1990 and 2007, approximately 400,000 US factory jobs were lost to automation.

If you’re a lawyer you might be thinking, “A machine couldn’t do my job.” Well, don’t be so certain. And that’s coming from a lawyer.

AI is the ultimate student. With help from data – good data and lots of it – AI can carefully study how behaviors and contextual inputs result in progress towards a goal. AI runs millions upon millions of simulations to get very good at a particular task, and when taught correctly, AI can get good at just about anything you tell it to get good at. Including legal work.

The future of AI-powered lawyers may be closer than we think

According to a paper written by economists at MIT and Boston University, robots could replace about two million workers by 2025. While many of these lost jobs are in manufacturing and other jobs that inherently require more repetitive, linear tasks, it’s not outside the realm of possibility that AI continues to creep into the legal profession.

For now, human intelligence and artificial intelligence have one major difference — the why. AI can easily learn how anything can be done better than humans. But AI still doesn’t have the curiosity to ask “why,” which prevents machines from doing the most meaningful legal work. That isn’t to say curiosity isn’t in AI’s future.

I have long been interested in the use of artificial intelligence in the legal field. I combined my passion for the law and AI into the creation of a bilingual AI-powered immigration law chatbot, YoTengoBot. If you want to discuss the future of AI in the legal profession, let’s connect! If you’d like to learn more about what I’m building with AI and the law, check out YoTengoBot!