Will AI decimate the legal profession? Will we care?
AI will decimate the professions. That’s the big idea that’s being debated by economists. In the same way that agricultural jobs were wiped out by mechanisation, working-class manufacturing jobs by robots and typing pools by word processors; so professions, such as managers, lawyers, doctors, journalists, architects and accountants and so on, will be wiped out by smart software.
Publishing as example
Let’s take one area of publishing as starter. Wikipedia came from nowhere. It wiped out the paper encyclopedia industry with a model that did four things:
Content was crowdsourced and is now being produced in many languages by people in those linguistic cultures. This was the converse of the ‘paid expert’ publishing model.
The publishers with their paper print and expensive distribution were simply cut out, as the content was available online with almost no creation and distribution costs.
As well as democratising the creation and editing process, access is free and democratic access to knowledge is available to anyone with an internet connection, on any device.
Knowledge was seen as corrigible and open to dispute through discussion, moderated creation and editing. Knowledge was no longer fixed and in the domain of experts only. This is an important, and often overlooked phenomenon and explains why educators and academics are often the most vociferous opponents of Wikipedia.
Lawyers as example
So let’s apply these four rules to the law. How will the law, and lawyers, be affected by AI and technology? Here are seven areas where the legal profession is already under pressure.
1. Direct access to legal documents
Repositories of legal documents give direct access by consumers, businesses and organisations to legal templates and documents. It disintermediates and cuts out the need for warm bodied lawyers at this basic level.
For example, Docracy provides free, open legal documents. You can even customise and sign these documents with electronic signatures. Their catchphrase is ‘legal for the people’ which says a lot about the process of disintermediation, democratization and demystification.
2. Direct access to legal templates
A step up from repositories of documents, are fully functional template services for shared access and the writing of legal documents. The idea is that the various parties can build upon a contract until complete, cutting down the need for expensive legal advice and/or lawyers.
ContractExpress provides a Microsoft Word interface to allow specific client details to be added, amended, noted and deleted. This makes the whole process of contract build that much more efficient. What is interesting is the way that intelligence is creeping into this process, to help increase speed, accuracy and lower costs.
3. New players
The profession has been opened up to non-legal entities, such as businesses, charities and other entities. This is where the legal process is being decentralised across a much wider set of players. Businesses, charities and non-qualified people are now, in many countries, allowed to play in the legal domain. This has resulted in a huge range of alternative sources for advice and help.
Rather than the hideous pro bono system, which was largely a PR exercise, organisations such as Citizens’ Advice have long provided free legal advice and services for the poor and this type pf service is growing. Even in the commercial sector, expert businesses, such as the AA, provide services that used to be provided by law firms. This has been accelerated by the use of online access to such services.
4. Workflow efficiency
A legal process can be a long drawn out affair; of get the facts, collect stuff from lots of different sources, then pay top dollar for someone to collate and reason using these facts to reach a conclusion and provide advice (often vaguer than you expected). You then have to execute that advice. There are few processes in life more frustrating and often scrappy than a legal process. It is ripe for automation.
Neota Logicprovide a workflow platform that does all of this, automating as much as possible. It asks specific questions, gets answers; collects from people, databases and the web; and here’s the rub, using a reasoning engine it provides advice, and can then execute the recommendation through screen, document or emails.
5. Dispute resolution
Disputation (and its prevention) is the cardinal driver of legal services and costs. Online services are now available that offer online resolution in divorces and other types of claims. The idea is to keep disputes out of the lengthy and costly system of lawyers and courts.
By avoiding lawyers and the courts, companies like Modria, deal with dispute resolution through straight disintermediation. Modria has a fairness engine that focuses on customer issues for commercial companies and has resolved more than 400 million cases for companies like PayPal and eBay. This lowers costs dramatically.
6. Reputation systems
Lawyers are in the judgement business and like many other professionals, are now subject to consumer judgement. Reputation systems such as AVVO disintermediate and demystify the law, as well as act as a filter for consumers looking for lawyers. AVVO is an online service that uses an algorithm to apply ratings. It has over 200,000 lawyers and a database of millions of questions and has become the de facto standard for reputation judgements in the legal profession. As a lead in, it offer free 15 minute sessions and free answers to submitted questions.
7. Predictive analytics
Legal judgements by very expensive human lawyers are under attack from software that uses smart algorithms, reasoning and probability software to determine outcomes. The process of capturing these processes is underway with the likes of LEX MACHINA, and will get better and better.
LEX MACHINA came out of a collaboration between the law and computer science departments at Stanford and delivers free data and advice on risk to academics, court and non-profits from its analytics engine. It collects large amounts of data and makes it available and searchable. This means that that for cases on patents, trademarks, antitrust and copyright, it can predict potential outcomes.
In ‘due diligence’, machine intelligence systems, such as Kira, promise speed and accuracy beyond that of a human lawyer. This is of huge benefit in M&A activity. eBrevia is a similar system with a diligence engine that completes as much as it can before the lawyers complete the rest of the process.
This is the area that most threatens traditional legal practice and jobs, as the human intelligence we used to pay for, is replaced by machine intelligence (to a degree).
Let’s revisit our four main criteria for our hypothesis, that the legal profession is under pressure in terms of employment, from technology.
All four are already at work and have already led to an overall loss of jobs. Note that the total number of jobs in the legal sphere is certainly reduced with technology but this is a complex equation, as some jobs are lost but new jobs created. However, it is not a zero-sum game. Fewer new jobs are created, as they handle services on scale. The net results is a large reduction in jobs.
Lest we think this is a bad thing, consider the redistribution of wealth from the expensive class of lawyers back into the pockets of the people who used to pay for these services. This particular profession relied for too long on mystifying their domain through abstruse language and ritual, charging for (often poor) document control and updates, keeping centralised control through expensive accreditation and services, refusing to lower costs through efficiencies of delivery and establishing a cartel that charged by the hour, for often menial tasks. The process of decentralisation, disintermediation, democratisation and, in my view the most important of all - demystification, is now well underway.