This article is not only for friends who feel anxious about the future of the legal profession. It is for anyone who is young, uncertain about what lies ahead, and still determined to do something about it.

Sahil Lavingia once wrote on X:

The best jobs aren't publicly listed. You have to dig for them, invent them, or convince someone it's worth creating just for you.

There was a time when a reasonably clear road seemed to lie ahead of us. Study hard. Enter a respected profession. Progress through a series of recognisable stages.

For many people, that road is becoming narrower.

Technology is changing the value of established skills. Institutions are becoming more selective. Entire industries are questioning how many junior people they will need, what work those people should perform, and how quickly yesterday’s expertise may lose its value.

When the roads laid down by others begin to narrow, we have to become better at creating roads of our own.

I do not have an authoritative answer for how to do this. I am still working it out for myself. But three principles have gradually shaped the projects I choose, the subjects I study, and the risks I am willing to take.

1. Start where the information barrier is lowest

I find it useful to distinguish between two kinds of knowledge: public knowledge and institutional knowledge. Borrowing loosely from the technology industry, we might also think of them as open-source knowledge and closed-source knowledge.

Public knowledge is available to anyone willing to look for it. It includes books, court judgments, corporate filings, research papers, open-source code, technical documentation, public datasets and recorded lectures.

Institutional knowledge is acquired through access to a particular workplace, client, transaction or professional network. It includes internal discussions, negotiation dynamics, feedback from experienced colleagues, interactions with clients, and the practical judgement that develops from repeatedly doing the work.

The distinction is not absolute. Most professions depend on both. But the difference matters because the barriers to acquiring them are very different.

Consider legal practice.

• In capital markets work, a prospectus is public. The process of coordinating banks, issuers, regulators and other advisers is largely institutional.

• In mergers and acquisitions, some transaction documents eventually become public, but much of the real work, including negotiation strategy, internal deliberations and communication between the parties, remains private.

• In international arbitration, procedural rules and academic commentary are publicly available. Learning how to communicate with opposing counsel and a tribunal, or conduct advocacy at a hearing, requires access to actual cases and the experience of being in the hearing room.

The same pattern exists outside law. A software developer can study documentation, inspect open-source repositories and participate in technical communities without being employed by a major technology company. By contrast, learning how a particular company makes product decisions may require access to its customers, internal data and organisational debates.

At different stages of life, we have access to different forms of knowledge.

• A student may have almost no access to institutional knowledge. But that does not mean the student must wait for someone to provide “real experience”. A serious project built from public information can demonstrate curiosity, judgement and execution far more clearly than a line on a CV saying that one is interested in a subject.

• An employee may be in the opposite position. If their organisation possesses exceptional institutional knowledge in a particular field, concentrating on that field can accelerate their development. They are temporarily standing inside an information environment that would be difficult to recreate alone.

We should therefore ask ourselves:

In which field do I currently have an information advantage, or at least face the smallest information disadvantage?

For myself, this is one reason I have chosen to build personal projects around emerging technologies.

Law is a profession built on decades, and sometimes centuries, of accumulated experience. Junior lawyers inevitably compete with people who have spent much longer developing professional judgement, client relationships and institutional authority.

Emerging technology has a different time structure.

When a new technical system, business model or regulatory question appears, nobody has thirty years of experience in it. Younger people may begin much closer to the same starting line. In some cases, they may even have an advantage: fewer assumptions to unlearn, more familiarity with new tools, and greater willingness to explore an untouched field.

These fields also contain an unusual amount of valuable public knowledge. The official documentation of a technology project may explain its design in remarkable detail. Research papers, developer discussions, regulatory consultations and public repositories allow an individual to investigate questions that would once have required access to a large institution.

This creates a place where experience can begin to accumulate regardless of where we are or who we are.

2. Learn to distinguish good information from convenient information

Projects and research are only as good as the information on which they are built.

With institutional knowledge, quality is often assessed through experience, supervision and feedback. A senior colleague explains why an apparently sensible approach will fail. A client reveals what actually matters commercially.

Public information presents the opposite problem. There is far too much of it.

The internet has made information abundant without making judgement abundant. A widely shared explanation may simply repeat an earlier misunderstanding. A confident commentator may never have examined the primary material.

For young people at the beginning of their careers, the ability to search for, compare and evaluate sources may be one of the most important variables in their development.

Suppose someone wants to understand artificial intelligence.

• One route costs nothing. They can begin with searches in the languages they know, identify the questions they are actually trying to answer, and choose materials appropriate to their technical background. They might read the documentation published by an AI company, study a research paper, begin with an accessible explanation such as Andrej Karpathy’s “Deep Dive into LLMs like ChatGPT” on YouTube, or, more directly, start their own project with an AI tool and ask questions as they build.

• Another route is to purchase a cheap, heavily advertised “complete AI course” or download a “free AI guide PDF” encountered on social media. Often, you even have to leave a comment or follow the creator in order to get a copy. The problem with this route is that we are outsourcing the choice of what to learn to whoever is best at attracting our attention.

One day, I had a conversation with Claude about building my “credentials” in tech. It said something I now fully agree with: start by building, not by taking courses. For smart professionals coming from law, the instinct is often to “qualify” through coursework before doing the real thing. In this field, the causation often runs the other way: a concrete project forces you to learn the theory, and that is why it sticks. One working artefact may say more than a pile of completed courses.

We do not need to understand everything immediately. High-quality, first-hand information is the foundation of our knowledge map and, eventually, our professional moat. We build that foundation by getting our hands dirty, not by collecting neatly formatted PDF guides that we save immediately and never open again.

That leads to the third principle.

3. Send signals to the world

We should not only collect knowledge from around the world. We should also send signals back into it.

That means choosing some questions that matter beyond our friends and immediate professional circle. It also means publishing our work where people outside that circle can discover it, and allowing our ideas to be examined by people who do not share our assumptions.

Law is an unusually local and fragmented profession. A lawyer’s expertise may be tied to one jurisdiction, one court system, one language or one category of transaction. This local knowledge can be extremely valuable. But it also leaves a career exposed to changes in a particular market.

One response is to identify subjects that require local knowledge but also participate in an international conversation.

The regulation of artificial intelligence, data, digital assets and financial crime provides obvious examples. Every jurisdiction has its own legal framework, but regulators, courts, academics and practitioners constantly observe one another. A development in one place can become a reference point somewhere else.

For example, a short analysis published in a column run by a Shanghai court discussed whether Bitcoin could possess the characteristics of property. The point was later picked up by international blockchain media and ultimately found its way into a footnote of the UK Law Commission’s consultation paper on digital assets. A local legal observation had travelled far beyond the platform, language and audience for which it was originally written.

Not every article will travel. Most will not. That is not the point.

We rarely know which observation, project or argument may later become useful to someone we have never met. A small contribution can enter a conversation in ways its author could not have planned.

Sending signals to the world shows people how you think and what you can do. A signal can be an article, a small software tool, a dataset, or an experiment.

Even a failed project can be a signal if it shows that you investigated a worthwhile problem and learned something concrete from the attempt.

There is another way to interact with the wider world: occasionally take the smaller road that departs from the direction your life is already moving.

This does not require abandoning your job or making a dramatic leap into uncertainty. It may mean choosing an unusual conference, writing about a subject outside your formal specialism, helping with a project whose value is not yet obvious, or accepting an invitation that connects you with people you would not otherwise meet.

During the final summer of my student years, I was offered a place at a workshop in Europe on law and natural language processing. I chose instead to undertake an internship at a court.

The court internship was, of course, more directly connected to the legal career I expected to pursue. It was the conventional choice, and not an irrational one.

Still, I sometimes think about the road I did not take.

At the time, legal AI was a much smaller academic community. That was before the boom of large language models. The discussion was still centred on forms of natural language processing that now seem traditional compared with the extraordinary range of work taking place today. Attending that workshop would probably not have transformed my life overnight. But it might have introduced me earlier to a group of people and questions that later became much more important.

The slight regret is not that I made the “wrong” career decision. It is that I evaluated the two opportunities mainly according to their immediate relevance to the path I was already on.

A life can be understood as a portfolio of “call options". Some choices offer stability. Others preserve exposure to unexpected upside.

The sensible strategy is not to maximise volatility in every part of life. Most of us have responsibilities and cannot behave as though failure carries no cost. But when a choice has limited downside and opens a genuinely different future, it may be worth selecting the less predictable road.

The age of the super-individual

People sometimes describe this as the age of the “super-individual”.

The expression can sound exaggerated. No one is truly self-sufficient. Meaningful work still depends on institutions, collaborators, communities and trust.

But the underlying change is real.

With today’s technology, one person can research a subject, analyse data, build a prototype, publish an essay, create a website, reach an international audience and collaborate across borders at a scale that would previously have required an organisation.

You do not need to wait until you have the ideal position, the right title, perfect credentials or permission from someone more senior. Begin where the information barrier is lowest. Learn to recognise information worth trusting. Make something that demonstrates what you care about. Put it somewhere the world can find it.

It may not lead anywhere immediately.

It may not receive much attention.

It may be imperfect.

But it will exist, and it will always belong to you.

Starting today, leave a small trace of yourself in the world.