Kevin L. Miller

I had the pleasure of speaking recently with Kevin L. Miller, the CEO of LegalSifter. He had a lot of useful insights regards the relationship between lawyers, technology, and the future of the practice of law.

Tell me a little about the impetus behind starting LegalSifter.

LegalSifter was born of Carnegie Mellon University in Pittsburgh, Pennsylvania, a city at the heart of the artificial intelligence evolution. The company knew a few things when it got started:

1) Natural language processing and machine learning, forms of artificial intelligence, had arrived, and the technologies were going to automate routine cognitive tasks.

2) The legal profession overflows with routine cognitive tasks.

3) Lawyers have not had technology that gets into the heart of their workflow because they think, read, and write for a living. Historically, legal tech software was not smart enough to think, read, or write.

4) A.I. can do a little thinking, reading, and writing.

5) Contracts are ubiquitous and the most important document in global commerce. They are also a pain to read, negotiate, and keep track of once they are signed.

6) The world needs faster, more affordable legal services. The legal profession largely serves the complex, high-dollar transactions and lawsuits. It does not optimize for the everyday business or consumer needs. 

We combined those observations with our mission to bring affordable legal services to the world by empowering people with artificial intelligence, and built the company as it exists today.

How do you see legal technology fitting into the larger legal industry ecosystem?

Most legal technology operates on the periphery of an attorney's daily activities. They can get by with their wits and own work product, built largely with Microsoft Office tools. As one of our attorneys says, "legal tech operates like a doughnut around the life of an attorney." It helps with workflow, billing, the occasional research, etc. It doesn't do much of an attorney's job.  


This reality contrasts dramatically with the rest of the world. Engineers, accountants, doctors, manufacturers, advertisers, product managers, human resources, etc. ... every other profession and function in the enterprise have tools and process that get into the heart of their workflow and help them scale. As a result, self-service, and partial-service products abound everywhere in our professional and personal lives. We can do our taxes without talking to an accountant, we can treat ourselves medically without talking to a doctor, and we can fix our car, house, and clothes without bringing in a service professional.

If we have a legal need, we have to talk to a lawyer. It's generally full-service all the time, LegalZoom, Rocket Lawyer, and the like aside. Those companies are off to a terrific start but note that they represent a micro-fraction of the $800B legal service market. 

Full-service lawyering is fine for sticky-wicket, strategic transactions and litigation. It's too slow and too expensive for the everyday needs of businesses and consumers. 

A.I. will change that. As more entrepreneurs inside and outside the legal profession figure out that A.I. does a little of what a lawyer can do, new self-service and partial-service products will emerge in the marketplace. People and businesses that had no options before because of costly, slow full-service lawyers, will gravitate to attorneys who offer proxies of themselves through technology. The world knows how to buy and use do-it-yourself solutions. The legal profession has not had reason to adopt that strategy en masse to date. The lawyers who want to change the profession--and there are a lot of them--will make the connection soon. 

In the past, legal tech served the attorney only. Tomorrow it will serve both the attorney and clients directly. Lawyers will willingly put A.I. in between themselves and their client to grow their business and deliver faster, more affordable, more complete legal support. 

The lawyers and law firms who figure this out early are going to win big, really big. They will serve their current clients more deeply, and they will attract buckets of new clients longing for something more suited to their needs.

For a proxy of how this will unfold, see what's happened in higher education over the past 20 years with the explosion of the Internet. Universities have been around with the same full-service model for over 1,160 years (University of al-Qarawiyyin in Morocco, founded in 859 A.D.). Today, every major university on the planet has an online education strategy and revenue, and that's all happened mostly in the last decade. The ones that do not are scrambling to catch up. The result has been more students, more revenues, and more education for everybody. Online education is not perfect, but it improves every year. Over time, it will lower the cost and democratize access to education globally. 

The legal profession will go the same way as higher education, only it will likely move faster.  

What is your take on the lawyers should learn to code debate?

It's a complete waste of time. Lawyers should not learn to code. They have enough to do. Leave the coding to the engineers. Do what universities do - partner with software companies. Lawyers should pick software companies that have competencies in A.I. Most legal tech companies do not today, but over time, they will all get there.  

The best engineers are not going to work for law firms. They will work for companies like LegalSifter, who can offer them a culture and support that works for them. 

Lawyers should understand what the current capabilities are of these technologies, including both the current limits and future possibilities. Lawyers should focus on how to insert these technologies into their practices or general counsel's office to help their unscalable full-service models deliver faster, more efficient support to their clients. They should learn the best practices for adopting technologies and driving change with software. These are well-worn paths by every other function in the enterprise, so there is no shortage of talent to show attorneys how to think about this stuff. 

Moreover, the legal tech firms will not have nearly the impact they want unless they weave in expertise into their solutions. Lawyers bring that knowledge. 

At LegalSifter, we often say that an attorney plus an algorithm is stronger than either by itself. I do not not trust the lawyer to get it right by herself, and I absolutely do not trust the algorithm to get it right all the time. The combination of the two is powerful. 

In short, leave the coding to the coders. Lawyers should stick with what they know and focus on learning the change management and product management skills that they currently do not possess.  

What do you think about legal tech companies having or not having lawyers on their team/in leadership roles?

They should have them on the team.  I'm not sure how they get it right otherwise. At LegalSifter, we have software engineers, data scientists, lawyers, and business professionals from all functions. It takes the points of view of everyone to pull off something that's going to stick. You have to have the voice of an attorney in the room, just like you have the voice of market, sales, service, software engineering, and data science.  

For a lawyer wanting to get up to speed on legal tech developments, how would you advise them? 

1)  Review thought leader writings. 

Read your blog. 

Subscribe to Richard Tromans' Artificial Lawyer site - https://www.artificiallawyer.com/

Check out stuff from Stanford's CodeX -https://law.stanford.edu/codex-the-stanford-center-for-legal-informatics/

Follow the Future Law Innovation Programme in Singapore - https://www.flip.org.sg/

Read Richard Susskind's books: https://www.amazon.com/Tomorrows-Lawyers-Introduction-Your-Future/dp/019966806X

2)  Do something with an A.I. firm-any of them-now. Stop dithering. 

Lawyers, old and young, tell me every day how they wish things were different. They tell me they want to come up to speed on A.I. 

Then they dither over whether to try new things. It's systemic in the profession, and completely opposite to the fundamentals of making real change. 

Stop. Just do something. Anything. The only way you are going to learn about A.I. or any other new legal technology is to put a shoulder into a project. If innovation is a core value at your firm, this is a simple thing to do. If innovation is a strategic initiative, then it's likely harder. 

And stop asking for stuff for free all the time. A.I. gets smarter over time, and these small companies can really do magical things for your general counsel's office and your law firm. You have to pay your suppliers if you want them to really help you.

I heard a senior partner at a successful mid-sized law firm speak publicly last month during a panel discussion on driving innovation in the legal profession. He was asked how his firm successfully drives innovation. He proudly explained how the firm had an innovation committee that evaluates solutions ... and then makes recommendations to an executive committee. The executive committee then takes up the recommendations and decides whether to put a task force in place to evaluate the solutions that were recommended. 

The natural conclusion of the innovation committee was a task force. The firm takes months to make decisions that should take minutes. 

As a former Chief Operating Officer of a 1,000-person global technology and service company, I was dismayed for that firm, its members, and its clients. They clearly have not been trained on how to build a successful decision-making enterprise that drives perpetual innovation. I do not blame them. It's a gap in knowledge. 

Much of the rest of the world--and the internal and external clients of lawyers--focus on speed, decentralized authority, and driving change. Give team members autonomy, mastery and purpose (see Daniel Pink's book Drive: The Surprising Truth About What Motivates Us). Make decisions quickly, try, learn, adjust, and try again. Lawyers should just do what everyone else is doing. They are clever enough to figure this stuff out. They just need to take a breath and stop crawling through a single decision in the time the rest of the world makes ten.