Human value, trust and limits of automation

When I sit down with Richard Tromans, he has just published a piece on the Anthropic legal plugin, along with a flurry of other updates on the day’s legal tech developments - another busy morning in the world of Artificial Lawyer.

He started the business in 2016 after a career in journalism and consulting - inspired by the change he saw coming and wanted to be a part of shaping.

Nearly a decade later, he says things are finally moving.

In this conversation, we cover his diverse background (he’s done more jobs than most people you know!) and how that informs his writing. We cover what is different about the latest developments in legal tech, whether the law firm pyramid is about to be replaced by something else, and whether AI risks making us all a bit dumber.

A decade of travel that still shapes his work today

Richard left university with one goal: never have a proper job. While his peers were lining up training contracts and summer placements, he went in the opposite direction. A decade of travel, factory shifts, bar work, cinema tickets, salad prep, and even a stint as a telephone tarot card reader (yes really), which he quit once he realised the callers were people who couldn’t afford a therapist.

He doesn’t see those years as squandered. Working in factories taught him how different parts of the economy actually operate and the assembly line gave him a perspective on the economics of law firms. Growing up in the Black Country, the historic heartland of the Industrial Revolution, shaped everything. His school history teacher skipped Waterloo and the Napoleonic Wars and spent years teaching the Industrial Revolution instead, in a town where it had literally happened a few doors down.

By the time Richard eventually landed in the City and then legal tech, he had a completely different lens.

From journalist to consultant

Richard’s first real job in law was as the world’s first international legal reporter at Legal Week magazine, starting in 1999. His beat was everything happening outside London: globalisation, the expansion of the magic circle firms, the BRICs coming online, the EU taking shape.

The job meant sitting down with managing partners of enormous firms and asking them to explain their strategy. He says the first couple of years he probably didn’t understand what they were saying and his articles probably weren’t much good either. But he got deeper and deeper into the business of law. About seven years in, a managing partner stopped their interview mid-conversation because he said Richard now knew more about the subject than he did. That was his first consultation. It was free, because he was still a journalist.

He spent the next decade or so as a strategy consultant in the City. By 2015 he was running his own practice, strangely dissatisfied, and increasingly intrigued by AI - which, he admits, he initially mocked. He says the scepticism masked a deep-seated interest. He just wasn’t getting good answers.

Then someone invited him to see it work. A company called RAVN (later acquired by iManage) showed him their contract analysis tool in action. He watched it zip through real estate documents and pull out key clauses. And he thought: this is the industrial revolution of the legal world. He changed his LinkedIn title to “legal industrialist” in homage to his Black Country forefathers, and launched Artificial Lawyer.

The billable hour problem that never went away

Richard tells a story from early in Artificial Lawyer’s life. He was giving a speech to hundreds of people at a law firm network event in Berlin, right next to the Brandenburg Gate. It had gone well. Then a woman at the back raised her hand and said: there’s one problem. I sell time for a living, and this will destroy my business.

Every head in the room turned. Richard admits he hadn’t thought deeply enough about the billing question at that point. He was more amazed by the technology. But she was right. She owned a private business, she wasn’t a charity, and this thing was not going to help her.

That was nearly a decade ago, and the problem hasn’t gone away. Almost every issue anyone ever raised about AI in law — training juniors, the billable hour, time to value — still exists. The base technology has changed dramatically, but the structural environment hasn’t.

Why firms aren’t rushing to change

On whether big law is productising its work, Richard is blunt: no, and we shouldn’t expect it. Even Big Law, he points out, contains about twenty different constituencies. Shipping firms in London are nothing like private equity firms in Manhattan. They share a pyramid structure, but they’re radically different businesses.

The economics are too compelling. An equity partner can work out on the back of a napkin what a workstream will produce in billable hours, at what rates, for what profit. It must be incredibly reassuring, Richard says, to know at the end of each quarter that you and your team have made millions doing essentially what you did last year. Without much client pushback, without much threat from new entrants.

He’s not surprised they’re not queuing up to disrupt themselves. Some firms are doing interesting things at the edges. If you went to an equity partners’ annual meeting and proposed a radical redesign of the business, why would they vote for it? There’s no need yet. They don’t feel it yet.

Richard thinks we’re waiting for a Cravath moment: a leading firm or small group of firms that seize the moment, change the model, and everyone else lines up behind them. He doesn’t know whether it will come from the US, the UK, or somewhere else. But he thinks it will take three or four years before partners start truly feeling it — losing clients, being told by buyers that they won’t pay for anyone below eight years’ PQE.

The pyramid is eternal, but it will operate differently

Richard pushes back on the idea that the pyramid is going away. The pyramid, he says, is the eternal structure of all human labour and organisation. Even organisations that say they’re flat are kidding themselves.

But the way it operates is going to change. He makes an interesting historical point: law firms were one-to-one for hundreds of years. A partner and an apprentice. It was technology - Word, email, the internet - that allowed leverage to scale. The question now is whether technology shrinks it back down again.

What won’t change is that equity partners are owners with client followings or irreplaceable niche skills. How they design their businesses around that is up to them. He wonders how many partners are actually having that conversation, versus how many are simply behaving as if this is the way it always was.

Cognitive surrender and the risk of getting dumber

Richard raises a concern he calls cognitive outsourcing. He shares a recent experience where ChatGPT confidently told him that a type of AI plugin didn’t exist - and nearly convinced him, until he pushed back and the model admitted it was wrong. He’s seen it with health questions too: the model assured him an edge case was extremely unlikely, and it turned out to be the correct diagnosis.

The danger, he says, isn’t just hallucination. It’s that when AI tries to be clever, it leads you down the wrong street entirely. And if people - or governments - outsource their thinking to systems that aren’t good enough yet, the consequences could be severe.

What’s next

Richard plans to keep Artificial Lawyer going for at least another twenty years. He can’t wait to see what legal tech looks like then. The Legal Innovators events are expanding - London, Paris, New York, California, which he calls the perfume bottle lineup. And he’s quietly launched The Robot Times, covering the intersection of robotics, business, and law, because he believes the robotics industry will become huge in the next decade and will need its own specialist legal ecosystem.

Through it all, the thread is the same one that started in the Black Country: how do complex systems change? He’s been watching this one for a decade, and he thinks we’re finally passing through a gate into something new.

Final Note: Turing & Partners, an AI Law Firm (in 2016)

In preparing for this discussion, I stumbled upon this post from 2016 about an AI law firm. I did a double-take when I saw the date!

Turing & Partners, an AI Law Firm

* Driverless cars

* AI-powered law firms (heard that term recently)

* New leverage models

* Big data centre developments

Links

* Artificial Lawyer

* The Robot Times

* Legal Innovators events (London, Paris, New York, California)

* Follow Richard on LinkedIn

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.agents.law