Asia Career Conversations: Interview with Jerrold Soh

Updated: Aug 7, 2020

Conducted by: Joshua Sung

The Asia Career Conversations series gathers leading industry professionals hailing from all across Asia to bring insight into various career prospects and their thoughts on current developments in the world today.


Jerrold is an Assistant Professor of Law at the Singapore Management University (SMU) School of Law. His research focuses on the application of empirical and computational methods in law, particularly with respect to artificial intelligence, legal technology, and tort law. As a co-founder of the legal analytics firm Lex Quanta, Jerrold has led commercial legal data mining and machine learning projects. Jerrold graduated with First Class Honours in both Law and Economics from the National University of Singapore (NUS) and has recently completed an LLM at Harvard Law School (HLS).

Disclaimer: The views expressed here are his own and do not reflect those of SMU. This interview has been edited for concision and clarity

On Education

1. Hi Jerrold, glad you could join us today! Many congratulations on your recent graduation from HLS and NUS. Please share with us your experiences studying at these two institutions and how they differed in your opinion.

My experience as a graduate student in the US and undergraduate student in Singapore were very different. The Law and Economics faculties at NUS were quite different as well, so I might say I was very lucky to have studied at three great institutions (and campuses). Each differed in terms of people, environments, and focuses, so it’s hard to compare them without making the kind of broad generalisations that make university rankings possible. Perhaps I might just share my experience at each institution.

NUS Law, where I spent the most time, was very rigorous from day one. Coming into law school without any relatives from law and after two years in the military meant the learning curve was very steep. My first legal writing assignment received a C+. The next few were not much better. I was lucky then that one of my former secondary school teachers happened to be in the final year of a graduate law degree, and gave me very helpful pointers. I was also assigned an excellent torts tutor who, beyond teaching torts, taught us how to research and write about law properly. I continued to be lucky and got amazing teachers for criminal law, contract law, etc. NUS Law sets you a difficult task, but they teach you how to accomplish it. It was hard work, let’s be clear, though all my peers were working very hard (much harder than I) so it never felt lonely. I should add that the NUS Law campus, located in the middle of the Singapore Botanical Gardens, is probably one of the best places in Singapore to get lost in the forest trails of the law.

I only started at the Economics faculty in the third year. By then, I was a ‘seasoned’ senior far less touchy on topics like GPA. I started with first-year Economics modules, which lulled me into thinking that Economics was ‘easy’ or at any rate easier than law. Because of my double degree, I quickly started taking level 2, 3, and 4 modules. There was a world of difference. While Law’s learning curve was logarithmic, Econs’ curve was exponential. Mathematical courses presented a completely different game, and the economics professors were adept at designing exams which produced what they’d call separating equilibria.

By the time I did my HLS LLM, I had been teaching at SMU Law for a year. I had a different way of thinking and different priorities. As a Singaporean who’d never lived abroad, my first priority was surviving the Boston cold. Fortunately, last year’s winter turned out to be very mild (perhaps due to global warming?), and Harvard is of course very survivable. The graduate programme is expertly run, and you get help in all areas – curriculum, health, security, housing, and should I say visas – if you so much as begin to ask. After one gets through the initial starriness of, say, chancing upon an original portrait of Oliver Wendell Holmes Jr on the way to class, it really is a law school, with professors, students, exams, and deadlines. To be clear, that is a good thing, as my goal going there (on SMU funding) was to learn. There was a lot for me to learn, including from the local and international students. There was also the much-talked-about ‘cold calling’, but not every instructor does that. HLS has a mixture of both lecture and seminar-style classes, but I attended mainly the seminar-style ones.

The US has a very different approach to the law from the UK and Singapore that we can spend years talking about. For one, the substance was quite different. In the US, the law is taught in a more interdisciplinary manner. The torts class I took, for instance, went quite in-depth into economics and psychology. My NUS tort law class was a lot more doctrinal and case-heavy.

I also felt like people were less focused on grades alone. Most still saw good grades as necessary, let’s be clear, but they were also seen to be far from sufficient for a successful legal career. People are very active outside of academics: joining the Law Review, going out for work experience, being involved in clinics and causes, for example. One problem with some (and definitely not most) Singaporean students is the assumption that as long as they get good grades, everything will fall in place.

Overall, both institutions had outstanding teachers. Even in HLS, it is inevitable that some teachers are better than others. I happened to take very highly rated courses (inadvertently). I found out later that most of the courses I took were by professors who had won prestigious teaching awards.

2. How was the module selection and offering in HLS?

There were a lot of courses on offer in HLS – hundreds, really. It is a larger school, and given its reputation there are many visiting professors who drop by and teach courses on their areas of specialty. Part of the variety probably has to do with law being a graduate degree in the US, so students may already have strong foundations in political science, biology, etc. Courses can then delve in-depth look into related fields. This may not be something an undergraduate system can support as well.

Regarding diversity, you are definitely more spoilt for choice with multi-disciplinary subjects. There are many well known legal economists in HLS, and also 3-4 courses on computer science and law including programming courses. For public law, which is one of the HLS’s well-known strengths, there are entire courses on specific amendments. So the 1st Amendment class is different from the 14th Amendment class. Whereas in NUS and the UK, your public law class may not split it up so granularly.

There are no compulsory modules for HLS LLM, though there are some broad requirements like having to take at least one American law course. For the year 1 JDs, there are compulsory courses like tort, criminal, public, contract, civil procedure. You don’t have to learn civil procedure in NUS, but it is very important in the US due to the federal system, and you can’t practice there if you don’t do it. They make the year 1 JDs do it in the first semester of the first year, and it is both very difficult and one of the defining characteristics of first year in HLS.

3. How’s the social life in HLS? What kind of activities are there to do?

Due to the size of the school, coupled with the numerous other universities in the area like MIT and Boston College, there are tons of events to attend. I think it is empirically true that every day there is a different guest speaker coming in, so you never actually have to pay for lunch if you want as there is lunch served everyday. This is quite remarkable. The Harvard brand attracts people flying in from all over the world.

There are also a lot of clinical courses in the senior years, where they embed themselves in the environmental rights or cyber law clinic, for example. These are taken quite seriously.

4. Many of our members are interested in pursuing further studies in the US and in the UK. Do you have any tips on the application process and advice on whether to pursue a Masters in the first place?

I think whether to do a Masters, especially in Law, is a highly circumstantial and personal decision, so I really can’t say. I did a Masters because my chosen career requires post-graduate qualification, and I managed to secure funding. Some do a Masters as a (pricey) holiday, and I don’t see anything wrong with that per se. Others wanted to work in the US. Perhaps if you actually manage to wade through the administrative trouble (and the copious application fees) of putting together something good enough to get accepted, you do have a strong enough personal reason for doing a Masters. Nothing to worry about, in other words.

On application tips, I did not apply to graduate programmes in the UK due to my research interests, so I don’t know about UK programmes. For most US schools, the process starts with the Law School Admission Council (LSAC) platform. Start as early as you can. There is a somewhat cumbersome process of getting LSAC to certify your transcripts. I almost missed the deadlines because of an administrative mixup that took me many troublesome phone calls to rectify, and I didn’t give LSAC enough lead time for it.

I applied to the US when international travel and visas were less of a concern, so do bear that in mind. I’m not sure how this affects acceptances – usually after accepting you the university will work hard to get you stateside. The Harvard graduate program was incredible at this in my year, but there are always things beyond one’s control.

As for substance, I can only echo the received wisdom that the personal statement matters (part of my research is on causality theory, so I find it hard to give tips on what works when all we have to go on are selected, successful observations). One of my SMU colleagues who had done the Harvard LLM before me said, very helpfully, “please don’t write that you’ve done this well in school, then got this prize, this award, then did this prestigious placement – every application will be like that”. I didn’t get multiple school awards anyway, so my strategy was to emphasise the different path I took into legal academia – through economics and programming. The first line of my personal statement read “I launched my first website a year before law school”. I also made sure to specifically address “why School X” in each application by identifying specifics of those schools – people, courses, etc. – which were attractive to me. I have also heard that having referees who are alumni helps.

I don’t know if any of it actually worked. Not every school I applied to accepted. One never really knows. Do look for unsuccessful examples too, as counterfactuals are necessary to infer causality.

5. You managed to balance your law and economics degree while executing data science research and programming tasks at your startup. Why did you decide to pursue these disciplines and how do you feel they intersect or complement each other?

I said during my NUS admission interview that I am interested in how the world works, and that law represents the social and political side of that while economics represents, well, the economic side. I was way out of my depth then, and have since learnt that the world is more than the social/economic/political framework one uses to write good A-level essays, but I think it largely still holds.

My parents also encouraged me to do the double, thinking that I’d have a stable career if I could speak economics to lawyers and law to economists. I doubt they had known then how many people in the US have doctorates in both. Call that Asian parenting, but in hindsight they clearly understood me better than I did at that point, and knew I would enjoy more diversity (and maths) in my university education.

How law and economics intersect is a strange question. Every lawyer I asked said economics was pointless for practising law. Everyone else thought the compatibility was obvious. There is no question in the US that the economic analysis of law is central to legal theory and doctrine. Views around the world and in Singapore are mixed, but law and economics are far from orthodoxy outside of the US.

Perhaps the only way to synthesise all this is to say that law and economics mostly intersect in theory and not practice, but even that depends on what we mean by ‘law’. Holmes Jr wrote that the life of the law is not logic but experience; Posner added that the logic of the law might be economics. I’m not by any means an eminent jurist, but for what it’s worth I think both law and economics are, as I’ve said, complementary ways of thinking about the world. They’re functional languages (or technologies, in the broad sense of that word) for reasoning about rights and wrongs. Pioneer legal economists found tremendous success translating legal doctrine into economese, which then let us use the full range of economic techniques – calculus and statistics included – to analyse the law. I’ve often found that trying to express legal doctrine in formal economic (and computational) terms forces me to be very clear about my argument, and quickly exposes fallacies.

I’ll add that, as the Harvard professor Jon Hanson taught me, law and economics go very well together in theory because both are built on a dispositional ‘will’ model of a rational person. Because one is assumed to act per internal (moral) reasoning, we can backwards-infer one’s internal morality from observing one’s behaviour. One who steals is bad; one who gives is good. There are many problems with this, as behavioural psychology points out, but I digress.

On Careers

6. Could you give an introduction to the career prospects as an academic?

Perhaps I could give a background on academic roles in SMU. Generally, there are two tracks, one is the research/tenure track, and the other is the teaching/practice track. If you are on the tenure track, you focus on research. You do less teaching that faculty on the teaching track and are evaluated primarily based on your research publications and other research-related activities. Practice-track faculty follow a mirrored (but, to be clear, no less important or prestigious) path. The key difference, other than the balance of research and teaching responsibilities, is the concept of a tenure review. As with academia anywhere in the world, tenure-track faculty will put their research portfolio ‘up for review’ a number of years (it varies with institution) post-appointment. The institution then goes through an elaborate process to determine if the academic is worthy of tenure – that is, lifetime employment subject to dismissal on very narrow grounds.

7. Please tell us more about your regular day as Assistant Professor of Law at SMU and the research projects you are undertaking.

There isn’t really a regular day in academia. My time is generally split across three big containers of teaching, research, and service (which includes service to the university, also known as administration). Some weeks I will exclusively work on preparing or delivering teaching materials. The month after exams is dominated by marking. Other weeks I spend primarily on research, particularly when submission deadlines approach. Before COVID-19 there would be weeks of conferences. The only daily constant is reading and replying to emails which could be on anything from student questions to grant applications to, well, interviews.

My focus has been trying to figure out how to use tech for law. This goes into 3 sub-questions: representing law formally, analysing law empirically, and the normative question of when we should and should not use tech for law.

As you might expect, my research intersects law, economics, and computer science. One of my projects involves studying how natural language processing can be used for legal analysis, particularly to extract outcome-causal legal factors from legal documents like judgments. That’s basically what common lawyers do, but there are in fact many empirical problems with relying on judgments, which may be written after outcomes and factors are determined, to infer this.

Another area I care deeply about is in applying new empirical legal techniques to study Singapore law. Numbers never tell the full story, but they can shed light on things we’d otherwise miss, and Singaporean legal scholarship has traditionally been highly doctrinal. This is a missed opportunity as Singapore actually has some of the highest quality legal data in the world. Given our jurisdiction size, the data is large enough to be useful, but also small enough to be wieldy. The first article I got published was about using network theory to empirically identify central appellate decisions in Singapore law.

I have other projects examining topics in AI law, such as autonomous vehicle liability and algorithmic bias, that I’m working on through SMU’s Centre for AI and Data Governance. We have very recently set up a Centre for Computational Law at SMU which will see computer scientists working with (and, in fact, in) the law school to tackle some big challenges at the intersection of law and computer science. The kind of problems that require deep computer science knowledge, way above what I have, to even begin thinking about. For example, how might we write a formal language for expressing, drafting, and verifying contracts?

8. You have multiple legal work experiences under your belt. What made you decide to depart from a career as a lawyer and instead venture into academia?

I never actually worked as a lawyer or trainee, only a legal intern. Frankly, I entered law school not knowing what I’d do after it. I went along with the internships-to-training-contract flow and was lined up to practice. Sometime in 3L, there emerged a trifecta of bad news for the legal industry: too many law graduates, bad economic conditions, and impending tech disruption. It was a stressful time to be doing law (it may have gotten worse). I figured that since I’m learning economics with some programming background, if anyone was to adapt to this, it should be me. I started by simply trying to learn more about this pesky ‘AI’ thing that, according to the Law Society President then, was going to replace junior lawyers. Law school 101 is to understand the other side, right? Econs had taught me mathematical notation and data analysis, so it turned out a lot more accessible to me than expected.

I also like to write. I always thought I eventually would want a job where I could write for a living, such as academia. What I didn’t expect was to join academia this early. The usual case is to practice law for a while, build up your expertise, then join academia and write about what you learned in practice.

How I ended up in academia early is a confluence of different factors, some personal, some external due to market conditions. Through the work I was doing with Lex Quanta, I learned that SMU law school was building capacity in law and technology, which coincided well with my interest. I also got to speak with lawyers and regulators from many levels of the legal practice, and the more I learnt about legal tech, the more I realised that this was a very exciting time to jump in. It felt like the wave was coming and I didn’t want to join too late after it had already swept everything away. I decided I’d just apply to SMU and see if they’d take me, and they did.

9. What advice would you give for someone interested in pursuing a career in academia?

Get a good idea of what a career in academia actually means. Many, including myself when I first started applying to academia, have misconceptions about this job. The more common ones: it is all about teaching; you get to do whatever you want; it is a highly-paid job; no prior experience is needed; you won’t travel much. Much depends on field and institution. I have heard people half-jokingly saying that in certain older institutions papers must go through the Associate Professor and be signed off by the Professor before it may be submitted to a journal, but that’s extreme.

I’d suggest speaking to people who are in the field and institution you’re applying to, but do remember that culture and expectations change over time too. What the ACJ is doing with its interview series is useful.

10. It’s been 4 years since you co-founded Lex Quanta. What have been some of the highlights and what are the startup’s goals in the years to come?

The best part has been working with a capable and multi-talented team. Legal startups are difficult because knowing how to code isn’t enough. You need to know both the theory and practice of law. Every lawyer will tell you that legal practice is special. This is true not just for what lawyers do, but also for the social and economic systems within legal practice. I enjoyed working with my co-founders to overcome the multi-faceted challenges thrown at us. I learnt a lot from a range of skills they demonstrated whether in engineering, business presentations, or financial accounting. All of us were law students when we started Lex Quanta, and you’d know that law students are not always known for so-called ‘extralegal’ skills.

We have worked/are working on a number of projects including with academic institutions, government organisations, and courts. I have to be vague here due to confidentiality. As our core offering – analytics – is very novel in law, it took us some time to get people to appreciate its value. Plus the sales cycle for legal organisations and firms tends to be very long. We’d like to think that we’ve found something that works, though. Our focus now is on delivering good work on our current projects and using them as a foundation for new ones thereafter.

11. What is Lex Quanta’s USP and how does it overcome issues that legal tech startups face?

All 4 co-founders had legal training and multi-disciplinary backgrounds in computer science and business. We could convince people that we knew the industry and application side. There are drawbacks though, as it was not easy to pick up the engineering skills to pull some of the projects off. But it is a lot easier to self-learn programming than law. Firstly, there are so much more resources for programming, but for law it is hard to pick up the basic principles yourself.

Every startup has this problem with user trust. This problem is even more acute in the legal industry where people are trained especially to distrust, to question, and be sceptical. If you are trying to make a sales pitch to judges and lawyers, you either need very high qualifications on the tech side, or you need qualifications in law so that you can be seen as an “insider” who understands the nuances of legal practice.

12. What has been the most defining moment in your career so far?

I think careers are made by a series of small sprints, rather than a big, definitive moment, but if I had to choose one episode, it will have to be presenting on legal technology and AI to a council of Singapore Supreme Court judges. That was the hardest I’ve prepared for any presentation I’ve made, and probably the only time I wrote the entire speech, which took many drafts, and stuck substantially to it to minimise the chance I’d embarrass myself. At my Dean’s suggestion, I rehearsed it tens of times, a few times before a live audience of colleagues whom I must thank.

One of the facilitators greeted me with “good morning, Prof” and I recall looking over my shoulder to find out who she was addressing. This was barely 3 months after starting at SMU and 2 months after my official NUS graduation, so I was (and still am) very green. That experience drove home the responsibilities an academic has as someone to whom people give attention and respect, if not only because of a title, and reinforced how carefully and conscientiously we should work to discharge that (public) duty.

On Current Affairs

13. What are your views on the burgeoning legal tech space in the US, UK, and in Singapore?

Legal tech is past its ideation phase and moving into execution and perhaps consolidation. In the early days we had a wide range of people starting up a wide range of ideas. Many did not work or fizzled out for other reasons (see: UpCounsel; Atrium), but some survived and are now thriving. I’m thinking of companies like Everlaw which have been years in the making and who have recently secured significant funding from mainstream VCs.

More importantly, we’re seeing higher market adoption rates (see the articles on ArtificialLawyer). The big players – magic circle firms, Big Four accounting firms, even many in-house departments – are now well into the legal tech scene. We might compare this to fintech – which is now dominated by banks and credit card conglomerates. Professor Ron Dolin at Harvard describes this as adaptive innovation, which is a hybrid between disruptive and incremental innovation where creative, young upstarts get co-opted (i.e. adapted) into big incumbents.

In short, the legal tech ‘battle lines’ have been drawn. It’s no longer the cool new kid on the block, but the college senior everyone expects something from.

The challenge for Singapore is remaining relevant amidst all this, given our small domestic legal market. Why should legal tech care about us? Singaporean legal tech startups and founders have many reasons to pay more attention to the UK and US legal market. To be sure, the UK may have this problem too particularly post-Brexit. I am thinking of DoNotPay, which is now US-based. But I doubt London’s strategic legal importance will wear off so quickly. Singapore’s strategy has, as always, been to position itself as a hub and launchpad for the Asian side of the world, and with strong government, judicial, and industry support I think we’ve done well to attract important players to set up legal technology efforts here.

Most would acknowledge the US’s leading position in legal tech. That’s been enabled by forward-thinking, multidisciplinary legal research coming out of their law schools, and the mature tech/VC infrastructure they have. The US JD system creates a larger pool of cross-disciplinary legal talent. But it does come with far higher tuition costs. Of the US law students I have interacted with, let’s just say not many are in a position to turn down high-paying attorney jobs out of graduation. Overall I’d say the US is clearly advanced in this area, but perhaps not by as much as the hype suggests.

At SMU I led a research collection surveying legal innovation efforts across the Asia-Pacific region, and China really stood out in terms of the number and variety of legal tech applications – some developed by the courts themselves. The first research paper to use neural networks for legal prediction was written by Chinese researchers, as far as I know. Things are both complicated and sensitive given US-China tensions but, regardless of one’s allegiances, it would be a mistake to ignore China’s rising and under-reported legal tech significance. I’m currently working on an updated 2020 version of this legal tech survey and am looking forward to what we’ll find.

14. How do you think COVID-19 will impact the adoption and advancement of legal tech?

The easy answer is that it will accelerate adoption, and market signals will in turn accelerate advancement. Legal tech is what economists call an experience good: you only know how good it is after you experience it. Unless you’ve tried it, you’ll underprice its value. But once you’ve had a taste, you start to see how useful it might be, and perhaps even wonder how you lived without it. COVID-19 has forced everyone to try. I think certain quarters of the industry are now thinking about why we haven’t done some of this sooner. I’ve always thought that law in particular is a job well-suited for working from home, since all you need is e-mail and Word. And, at least from my limited internship experience, the dominant model of teamwork in legal practice is divide-and-conquer, which means you don’t often need to discuss things with colleagues.

The difficult question, which I cannot answer, relate to exactly what technologies will be furthered, and perhaps whose technologies. The main product on trial is online dispute resolution and video-conferencing. I don’t know how far other services like e-discovery and analytics will be affected, but they could benefit from improving attitudes towards tech, as well as a much-needed shortening of the legal tech sales cycle (how many startups can afford two years of back and forth before closing?).

Many world leaders, including Singapore’s PM, have written about the dangers of escalating US-China tensions, in part caused by (political reactions to) COVID. I think we can expect some of this to play out on the legal tech stage as well. There is an increasing awareness, particularly, but not only, in the US, of how technology can exacerbate social biases (aside: legal tech will soon have to confront this). A paper written by Chinese researchers on using AI in the context of criminal sentencing, based on Chinese court data, triggered very harsh critiques from the AI ethics community. I suspect that the US and China’s legal tech trajectories will only diverge more and more.

395 views0 comments

Recent Posts

See All