Mar 3
Not All Tech is Created Equal — Why I Left Self-Driving Cars for Legal Tech
As I mentioned in my previous blog, I quit my job in self-driving cars and started a new gig of my own. This new gig is legal tech, which seems entirely unrelated to self-driving cars.
You may think I’m crazy. How could you leave such a sought-after industry for something like legal tech?
Well, let me first clarify what legal tech means:
Legal technology, or legal tech for short, typically refers to technology that helps lawyers, law firms, or large organizations with their legal-related tasks, such as document storage, case management, legal research, etc. It’s not a well-known industry, largely because most legal tech applications are intended for lawyers, not clients.
Not all tech is created equal
Unlike the self-driving car industry, legal tech does not receive much funding or attention. In fact, these two industries represent opposite ends on a broad spectrum of tech popularity. To give you some numbers, the top self-driving car companies have received $8 billion in investment in 2020. Legal tech, meanwhile, received only $510 million in the same year, just 6% of what was given for self-driving cars.
Legal tech is not appealing to the top tech talent, either. If you ask software engineers which industry they want to work in, their answer is rarely legal tech. First of all, legal tech doesn’t sound as “fancy” as its tech counterparts, such as crypto and self-driving cars. Legal-related projects just sound very old-school. Second, the average pay in legal tech is lower than in other tech industries. In San Francisco, a software engineer earns an average of $148k per year, whereas a legal tech engineer earns only $117k, a full 21% lower than the average.
Lacking attractiveness to talent, legal tech struggles with delivering high-quality products. These products are supposed to ensure efficiency and cut costs for lawyers and their clients. Clearly, the value chain in the legal industry is broken.
Is it a bias that lawyers are so slow and expensive?
The bias towards lawyers is not coming out of thin air. To explain this, let’s first discuss efficiency. You would be astonished to learn how out-of-date current legal processes are due to lack of technology adoption. This is not because lawyers don’t want to work with software engineers to build better tools, but rather because lawyers who can afford technology R&D are not incentivized to innovate in the first place, given that they already have plenty of Fortune 500 clients paying them millions of dollars per year. By contrast, the lawyers who are willing to innovate simply do not have the resources to do so.
Partly due to low efficiency, legal costs remain consistently high. A typical lawyer charges $200–$3,000 per hour (Yup, most lawyers today still charge by the hour instead of by deliverables). A typical personal injury trial in New York can cost anywhere from $15,000 to $150,000 in total. That could be an American’s annual salary! You can’t protect yourself and your family without breaking the bank.
Despite that the U.S. is undoubtedly a world leader in many aspects, one disappointing fact remains: The U.S. has the most costly legal system in the world, according to a study published by the U.S. Chamber Institute for Legal Reform and NERA Economic Consulting.
Consequently, this creates a vicious cycle. No funding -> low salary -> lack of talent -> subpar products -> low efficiency -> high cost for the general public.
The good news is, it’s about to change
Why are we not seeing enough changes to legal tech? Let’s look at the self-driving car industry as a reference.
The self-driving car industry became widely popular among tech talent around 2014. This was when several self-driving startups raised sufficient funding and various large corporations began investing in their self-driving car divisions. One significant contributor to the rise of this industry concerns the advancement of machine learning and the sheer volume of machine learning talent that has become available.
Eight years later, the self-driving problem is still difficult to solve, but practitioners remain passionate about its growth potential. Even now, we are already seeing self-driving cars testing on the roads of San Francisco and many other cities around the world.
Self-driving development couldn’t have happened 20 years ago. People were happy with their traditional cars and didn’t even know what machine learning meant. Despite their cars causing more than thirty thousand deaths and generating more than seven hundred million metric tons of carbon dioxide in the U.S. each year, it still took until recent years for the general public to finally say no to traditional cars.
I see a similar trend coming for legal tech. Legal service has to be made accessible and affordable to everyone. Radical changes will be happening soon, much like what’s happening to the car industry. The following breaks down the four reasons why legal tech is about to change and why I’d urge you all to join forces for the change.
Reason 1: Legal AI is ready to bring about fundamental changes
Much like self-driving cars need computer vision, legal tech needs NLP (natural language processing), a concept that has been around for some time now. In 2011, IBM released Watson, a Q&A software that won the Jeopardy! contest, defeating the best human players for the first time. In 2014, Amazon released Alexa, a virtual assistant that understands basic human languages. In 2018, Google introduced BERT (Bidirectional Encoder Representations from Transformers), one of the first large language models with 100 million parameters, followed by the introduction of GPT (Generative Pre-Training).
Still, it is only recently that NLP has been commercially adopted for legal use cases, such as drafting contracts and finding relevant case-related content. Today, two problems remain concerning AI applications in the legal space.
The first problem lies in the target users of legal AI. The current legal AI applications focus on improving the efficiency for lawyers’ existing clients rather than potential clients. Part of this is due to the lack of R&D funding from law firms struggling to find clients. They would rather spend their budget on marketing than on technological advancement. In fact, today’s solo practitioners and small firms spend almost ⅓ of their time identifying and developing new clients. Surprisingly, an ideal solution has yet to be found to solve this pain point. The market potential for this product area is enormous.
The second problem concerns the data for model training. To function properly, good AI must rely on good data. Today, many legal AI applications use text data from contracts to train models, but contract language is rarely understood by the general public. To counter this, effective AI should also use data generated by the public. In this case, this would mean training models using plain English, not just legalese from lawyers, so that AI can also help non-lawyers process legal tasks. Very much like how we annotate real-world driving data to train self-driving cars, we also need real human language to train a legal AI.
Reason 2: Legal tech shifts the focus to clients, not just lawyers
In order for technology to thrive, it must be made available to the general public rather than just a specific group. However, legal tech traditionally focuses on products designed for lawyers, not clients, and it often doesn’t allow easy integration for clients. For example, case management systems may not come with client-facing interfaces.
One law firm I worked with last year asked me to print out an agreement to sign and send them fees by writing a paper check instead of using e-sign or e-payment, which is already a common practice for many other industries. According to a survey by the American Bar Association, 51% of law firms with 100+ lawyers use a client portal, but this figure only reaches 9% among solo practitioners.
The good news is that the industry is finally fixing this issue, with more investors joining forces. Though legal tech does not receive as much funding as self-driving cars, it is catching up at an incredible speed. In 2021, for instance, legal tech startups received a massive 548% bump to their average round size compared to the previous year.
Reason 3: The demand for legal help has been exploding
Many investors told us that they are not investing in legal tech because it isn’t a market with strong demand. That is simply not true. The fact that they themselves do not struggle with getting legal services does not mean that others don’t.
On the one hand, people with sufficient access to lawyers are often from the upper class, so they rarely complain. On the other hand, people suffering from a lack of legal support often don’t have a way to complain. According to the World Justice Project, the U.S. ranks #126 out of 139 countries in accessible and affordable civil justice. Take California as an example. A Californian needs serious legal support twice per year on average, and that number has been skyrocketing since COVID started. However, 70% of them receive no legal assistance at all.
Two trends went unnoticed:
1. Due to COVID, a large number of small businesses had to go through significant changes such as firing and rehiring employees, closing and reopening businesses, etc. They all need specialized legal assistance at a reasonable price.
2. The Great Resignation during COVID led to a new generation of startup founders. Nearly 4 million Americans on average quit their jobs each month last year. In particular, the tech industry had a 4.5% increase in quit rate. Many of them went starting their own gigs. The number of newly registered businesses almost doubled in 2021. However, legal procedures can be intimidating for founders, such as reviewing an investor side letter and splitting equity with team members. All require a good yet affordable lawyer.
Monthly quits in the U.S.
Reason 4: The young generation of lawyers is looking for opportunities outside of Big Law
Similar to what’s happening to other industries, the younger legal generation is looking for nonconventional career alternatives that would allow for flexible hours and work styles, which is also an outcome of the Great Resignation mentioned above. As a result, an estimated 63% of lawyers in the U.S. are either solo practitioners or work in small rather than big law firms. To this end, having access to legal technology is crucial to their success. Especially in the post-COVID era, most legal services can be delivered virtually. As more tools become available for solo practitioners, such as for scheduling meetings with clients or for building websites without coding, more lawyers will feel empowered to start their own gig.
Working on something cool while making the world a better place
Everyone in tech wants to work on something cool while making the world a better place, myself included. Legal tech certainly makes the world a better place, but it will be on every one of us to also make it a cool industry to work in. It would be wonderful to see more tech talent joining from “shiny” industries such as self-driving cars and bringing excitement to legal tech.
Disrupting an industry is never easy, but I believe that legal tech is the next target for change. Stay tuned to the legal tech product I’m building! If you want to exchange ideas about legal tech, don’t hesitate to find me on LinkedIn at https://www.linkedin.com/in/charlottetao/.
