This week, for a change, let's talk lawyers. And not just lawyers, because what applies to lawyers applies to everyone else who peddles expertise, including financial advisers, CFPs, management consultants and the like. Basically, the robots are coming for our jobs.
There is a robust literature arguing that the way we practice is unsustainable. In The Future of the Professions, Richard and Daniel Susskind argue that professional services have become unaffordable to many and unavailable to most. They fault professionals for being unwilling to fully exploit technology, for maintaining a mystique around their expertise that emasculates clients, and for being inscrutable in how they govern themselves. They basically agree with George Bernard Shaw, who claimed that the professions are "a conspiracy against the laity."
Richard Susskind has made a living from being the self-appointed Cassandra of the legal profession. His efforts to drag UK lawyers into the future earned him membership in the Order of the British Empire -- in other words, a knight. Sir Richard's basic point is that lawyers today practice a craft akin to that of a custom tailor, too elitist for a jeans-and-hoodie world. He even uses a Savile Row term -- bespoke -- to denote those legal skills that serve only the rich and powerful. Most of us can't afford bespoke legal services, and in all likelihood don't need them. And yet, to a very large degree, the profession offers only bespoke services, out of reach for all but the few. Sir Richard has argued for years, through several books, that the legal profession -- indeed all professions -- must reimagine themselves to remain relevant. His main idea is that market forces and technological innovations will force all bespoke professional services to become commoditized, and we'd best prepare for that.
Lawyers are not uniformly distributed; NYC has 14 attorneys per 1,000 people, while many counties have less than 1 per 1,000.
The average lawyer in big cities makes 5 times the average in small towns.
The vast majority of lawyers work long hours at least sometime, and almost half say they work long hours most (or all) of the time.
Lawyers have more mental health issues than the population at large.
Lawyers drink 4 times more than the US average.
These, I think, are the natural consequences of working to create bespoke services. Most of the people are under-served, while most lawyers outside of big cities are struggling because the vast majority of the people can't afford to pay for bespoke services.
Now, if ever there was a legal product that is typically more bespoke than it should be it's a contract. At its core, a contract is a simple thing. You give me that and I'll give you this. A quid pro quo, usually money is exchanged for something that's not money. Behind that give and take, a contract will provide details about how and when the exchange will take place, and certain promises and assurances that the things being exchanged are really what we say they are, and that we're not cheating some third party in the process. Of course, contracts can get very complicated, as the things being exchanged become more convoluted, or as conditions are added to the thing. But in the end, drafting a contract is like building a Lego house. They are all built out of standard blocks. Even the special blocks are built out of standard blocks.
No wonder, then, that contract work is the first legal product heading down the road of commoditization. Of course, there have always been standard form leases and contracts of sale that one can still buy from legal stationers like Blumberg's (see https://www.blumberg.com/forms/), which the literate consumer can just fill in herself. But I'm thinking more of RocketLawyer and LegalZoom, where answers to questions result in a computer-generated contract or other form. One can't get more commoditized than that. There isn't even a lawyer providing that service.
There will be more of this coming. Also last week, Casetext published a study of its Compose program for writing first drafts of litigation briefs. As I understand it, the program presents a list of possible arguments for certain standard trial motions. You click which arguments you want to use, input some facts, and the computer spits out a first draft of a brief with all your arguments and cases already there. When 13 litigators tested the software, they found a 76% reduction in time spent, resulting in either cheaper or better motion papers. https://compose.law/wp-content/uploads/2020/07/Compose-ROI-Report.pdf. Of course, this was not a scientifically valid study, but just the same.
Where will this end? One suggestion is provided by GPT-3, the most recent iteration of a deep machine learning program that can translate one language into another. But not like GoogleTranslate does. Demonstrated by the "research and deployment" company OpenAI, GPT-3 has the ability to code website features in response to plain English commands. But it could also be programmed to write a contract or even a brief, taking what RocketLawyer and Compose do to the next level. Imagine saying, "Alexa, write me contract with the following features," and, boom! it's done! https://www.artificiallawyer.com/2020/07/29/gpt-3-a-game-changer-for-legal-tech/.
Well, this is all very well and good, but I'm still skeptical that there will be so much change so fast. Compose does a great job of automating what most of us already have access to, in the form of well-organized brief banks. Those programs may save a little time at the front end, but I guarantee they'll take more time to hammer into final shape in later editing, which is when, if ever, writing becomes "good." When something like GPT-3 is perfected, it may indeed be a "game changer." But it would cost $4 million to $12 million to program GPT-3 to recognize the language needed to draft a contract or a brief. And for all that would GPT-3 really be that much better than RocketLawyer, LegalZoom, Compose or any human with good precedent forms?
In the end, all the gushing over these efforts smacks of the Eliza Effect. This is the "willingness to look past obvious errors in order to marvel at the wonders of AI. . . . a human complicity in a digital fantasy." Eliza was the first chatbot, programed at MIT in the 1960s to emulate a psychiatrist. Its creator was amazed at how willing its awestruck human interlocutors were to give it a pass on the Turing Test, despite its obvious flaws. See https://spectrum.ieee.org/artificial-intelligence/machine-learning/this-ai-poet-mastered-rhythm-rhyme-and-natural-language-to-write-like-shakespeare. Eliza simply took whatever was said to it and repackaged it into therapeutically-styled questions. The funniest thing Eliza ever did was match wits with PARRY, another early chatbot developed at Stanford that emulated a schizophrenic. The dialog that resulted is sheer poetry, albeit with a lunatic fringe. See https://tools.ietf.org/html/rfc439.
Eliza, PARRY, RocketLawyer, LegalZoom, Compose and GPT-3 all hang on a common evolutionary tree. They all must be pre-programmed with bits of information and certain rules for manipulating that data and for adding new data. That is what machine learning is about. Human learning is like that too, but it is more than that. Humans interact with and in the process change their environments. The changing environment provides the context through which humans process old and new data to understand themselves and the world. Machines will not do that until they are able to emulate human existence -- not just thinking -- more fully than anything around today. And if they can do that, we shouldn't call them machines anymore, and we shouldn't enslave them as such.
But until then, no matter how sophisticated, machines can only build Lego houses from the bricks we give them. The very best argument that guys like Sir Richard have for the commoditization of professional services is that when the bespoke is out of reach, Lego houses will have to do. That's true, and fortunately, Lego houses are usually good enough to keep out the rain. But they'll never be as good as the real thing.
Aegis Frumento is a partner of Stern Tannenbaum & Bell, and co-heads the firm's Financial Markets Practice. Mr. Frumento represents persons and businesses in all aspects of commercial, corporate and securities matters and dispute resolution (including trials and arbitrations); SEC and FINRA regulated firms and persons on regulatory compliance issues and in SEC and FINRA enforcement investigations and proceedings; and senior executives of public corporations personal securities law and corporate governance matters. Mr. Frumento also represents clients in forming and registering broker-dealers and registered investment advisers, in developing compliance policies, procedures and controls, and in adopting proper disclosure documents. Those now include industry professionals looking to adapt blockchain technologies to finance and financial market enterprises.
Prior to joining the firm, Mr. Frumento was a managing director of Citigroup and Morgan Stanley, a partner and the head of the financial markets group of Duane Morris LLP, and the managing partner of Singer Frumento LLP.
He graduated from Harvard College in 1976 and New York University School of Law in 1979. Mr. Frumento is a frequent author and speaker on securities law issues, and is often quoted in the media on current securities law developments.
NOTE: The views expressed in this Guest Blog are those of the author and do not necessarily reflect those of BrokeAndBroker.com Blog.