I had a conversation with a robot recently.
I had placed an order with Amazon and although I received a notification on my phone that it had been delivered, there was no sign of a parcel anywhere. The app said that the order had been delivered to a resident at my office, but no one in my building knew anything about it. I therefore visited the Amazon website and had an online chat with an Amazon customer service operative called Steffy.
After a couple of sentences bounced back and forth between me and Steffy, I realises that the slightly too perfect messages from her meant that I probably wasn’t talking to a human. I’m pretty convinced that Steffy is a robot.
In my more than slightly Star Wars obsessed mind, Steffy is a golden droid with a passing resemblance to C3PO, sitting in front of a computer, typing with metal fingers. Of course, in reality I suspect that Steffy has no physical presence. She (if robots have gender) is simply a computer program and a very efficient one too. She came up with an explanation for why the package not been delivered and offered me a choice of either a refund or waiting until the package was re-delivered. I opted to wait, and the following day, the package turned up.
Robots and advanced AI are going to be part of the legal landscape in the future. Last week I attended the 31st Annual Essex Law Lecture organised by the University of Essex School of Law and the Suffolk and North Essex Law Society. The lecture was given by Richard Susskind who is the leading thinker on the future of law and information technology.
Susskind described the near future to us. With every passing year, computers become more and more capable. Susskind outlined to us how online dispute resolution will become much more common and provide greater access to justice for people who cannot currently even consider affording a lawyer. For the foreseeable future, online dispute resolution courts, where there is little or no involvement from a human judge, will only be capable of resolving straightforward low-level civil claims such as debt recovery etc. However, who knows how sophisticated these systems will become in the long term?
Susskind cited the example of one solicitor who told him “My clients pay me to provide them with my judgement on their case”. Susskind’s response was that clients don’t pay for a solicitor’s judgement, they want to know what their chances of success are going to be. Legal AI is increasingly proving to be considerably more consistently accurate in its outcomes than humans.
I squirmed silently in my seat when he said this. I often explain to clients that resolving the financial aspect of a divorce is more of an art than a science. My advice is always an attempt to predict what I think a judge would do if he had to decide the case. In doing so, I am exercising my judgement. The court will take into account all the circumstances of the case and in particular a number of statutory factors when it decides how to divide the assets between the parties and whether to make an order that one party pays the other party maintenance:
(a) the income, earning capacity, property and other financial resources which each of the parties to the marriage has or is likely to have in the foreseeable future, including in the case of earning capacity any increase in that capacity which it would in the opinion of the court be reasonable to expect a party to the marriage to take steps to acquire;
(b) the financial needs, obligations and responsibilities which each of the parties to the marriage has or is likely to have in the foreseeable future;
(c) the standard of living enjoyed by the family before the breakdown of the marriage;
(d) the age of each party to the marriage and the duration of the marriage;
(e) any physical or mental disability of either of the parties to the marriage;
(f) the contributions which each of the parties has made or is likely in the foreseeable future to make to the welfare of the family, including any contribution by looking after the home or caring for the family;
(g) the conduct of each of the parties, if that conduct is such that it would in the opinion of the court be inequitable to disregard it;
(h) in the case of proceedings for divorce or nullity of marriage, the value to each of the parties to the marriage of any benefit which, by reason of the dissolution or annulment of the marriage, that party will lose the chance of acquiring.
This gives the court enormous discretion about how it can achieve fairness and meet the parties and their children’s needs. Inevitably therefore, it can sometimes be difficult to predict the outcome. No two judges will give you the same exact solution. In fact, if you put three lawyers in a room (the judge, the solicitor for one party and the solicitor for the other), you probably get three opinions. They ought to be all roughly in same area, but they often will not be exactly the same. The court’s discretion is designed to allow it to come up with a tailor-made solution to fit the parties’ circumstances. This very wide discretion is quite popular with family solicitors and barristers, which is one of the reasons why the proposed reforms being pursued by Baroness Deech have very little support in the profession, nor have they been adopted by the government.
Is it possible to come up with a computer program or algorithm (I’m not sufficiently technically aware to know what the difference is between the two, if any) that can divide a divorcing couple’s finances between them in way that achieves fairness and meet their needs?
There is already a formula for calculating child maintenance which is used in this country by the Child Maintenance Service. It replaced a much more complicated formula that was used by the old Child Support Agency and has been in place now for about 16 years. The old formula was so complex you needed a computer program to calculate it. However, using the current formula, I can usually do it with a pocket calculator.
However, the formula is a blunt instrument. It provides certainty at the cost of being able to provide a bespoke solution. The formula is quite controversial in some respects; for example, it gives the non-resident parent a discount if the children spend 52 nights per annum or more with that parent. It also gives the non-resident parent a discount to reflect the fact that he may have a child living with him, even if that child is his new girlfriend’s child and he is not the father, and even if the child’s mother is receiving child maintenance from the child’s father. It also does not take into account income that is received by the non-resident parent from his or her assets (although changes are being introduced whereby the non-resident parent will be assumed to receive income from it, which may be more than the actual income is in reality). The resident parent’s financial circumstances are ignored completely, whether he or she is a pauper or a billionaire.
The inflexible nature of a formula makes it far less attractive to family practitioners. However, this may be what the future brings. I am aware of least one app developer who’s already struck it rich with an app that is used to challenge parking fines and who is currently working on app to resolve financial disputes in divorce proceedings. He wants to become the Uber of divorce.
Uber is very controversial among some taxi drivers; black cab drivers hate it. I love it. About 18 months ago, when the Mayor of London announced that Uber would lose its operating licence in London, I tweeted some mild scepticism about whether this was likely to last and whether or not it could be challenged by Uber through the courts. I took the time to read Transport for London’s reasons for taking away the licence and they struck me as being issues that could be easily resolved by a decent HR department (and indeed, in due course, they were; Uber continues to operate in London).
Although at that time I had not actually used an Uber, I had read Brad Stone’s fascinating book “The Upstarts”, all about Uber and AirBnb. Uber’s business model struck me as being brilliantly conceived and implemented. Taking away Uber’s licence to operate seemed to be a futile attempt to turn the clock back.
For the rest of the day, I then received numerous tweets from black cab drivers, many of which were very obnoxious and some downright insulting. I have to say that I didn’t feel the love. Their response did not make me want to use black cabs and simply made me more inclined to use Uber. Their objections to minicab drivers using Uber seemed to be that they were dangerous, but as far as I was aware Uber drivers are regulated by TfL and they are insured. One of the black cab tweeters argued that as a solicitor, I couldn’t possibly know what I was talking about because, unlike a taxi driver, if I did my job badly, nobody dies. I resisted the temptation to tweet back that training to be a solicitor is a lengthy and much harder process, whereas I could become an Uber driver tomorrow as I have a driving licence and I know how to use a satnav. The only real difference between an Uber driver and a black cab driver, as far as I was concerned, was that Uber drivers aren’t too proud to use satnav. Passing the Knowledge is an impressive achievement, to be sure, but also completely pointless in this day and age when a mobile phone has the Google Maps app on it. Mrs Armstrong and I once took a black cab to a hotel in a obsure backstreet in Shoreditch and the black cab driver wracked his memory desperately to remember where the street was. He was clearly infuriated that he couldn’t remember it. I admire his pride in his training, but it reminded me of a conveyancing partner, when I was a trainee solicitor in 1994; he was only really ever happy with a conveyancing matter if it involved an unregistered title;. Ploughing through an epitome of title and dusty old title deeds was his idea of proper conveyancing, whereas I considered it far too complex and I much preferred dealing with a straightforward registered title.
Uber plan to replace their drivers in due course with robot vehicles. Again, in my sci-fi obsessed mind, I want them to look like Johnny Cab in Total Recall, but of course they won’t. Taxi drivers and lorry drivers are likely to become a dying profession. I recently saw a tweet where a father proudly announced that his son had qualified as an HGV driver. The tweet made it very clear how pleased he was that his son had learned a trade which he could use to support himself and his family in the future. I resisted the temptation to burst his bubble by replying to his tweet pointing out that HGV drivers will cease to exist long before his son reaches retirement age. But it may not just be drivers; the prospect of AI and other sorts of advanced IT replacing lawyers does makes me wonder if family lawyers may one day join taxi and HGV drivers as obsolete relics of the past.
16 March 2019