« Is this our next Supreme Court justice? | Main | The Talk »

Wednesday, February 24, 2016

I, for one, welcome our new robot Law Lords.

Friends, I've been a terrible guest-blogger so far this month. My apologies. Life (and teaching... mostly teaching) intervened.

But one of the things I'm teaching is an experimental yearlong project-based seminar called the Policy Lab (link is somewhat obsolete), where students spend the first semester learning about an area of legal policy, and the second designing innovations to work on it. And for this first run-through, students have been thinking about legal technology and access to justice. They've learned about things like predictive coding, multijurisdictional tech-driven delivery of legal services, and artificial intelligence, and they've had virtual as well as physical visits from experts and people making waves in the area, including Dan Katz, Jake Heller, Stephen Poor, Tim Hwang, and Craig Cook, as well as more local folks---and now they're working on designing (though not fully implementing) technological tools to provide legal knowledge to nonprofits, as well as policy analyses of, e.g., the ethical implications of such tools. I'm really proud of them.

I'm also a confirmed parking and traffic scofflaw, who once beat a parking ticket with a procedural due process claim, and also once beat a speeding ticket by getting testimony about the laser evidence chucked on the good-old Frye standard (back in grad school, when that standard applied in California). So imagine my delight when I saw this story: "A 19-year-old made a free robot lawyer that has appealed $3 million in parking tickets". A Stanford kid, Joshua Browder, has written a webapp that (as far as I can discern without trying it out or seeing the code) quizzes people about their parking tickets (U.K. only, alas) in natural language, invokes what is sometimes called an expert system to discern a defense for them, then provides an appeal for them to file. Obviously, I have lots of questions and thoughts about this after the fold.

First, is this legal in the U.K.? How do folks feel about the unauthorized practice of law on the other side of the pond? And what about California? On some aggressive interpretations of UPL rules, we might think that the awesome kid is practicing British law in California. As this kind of service, and the services provided by companies like RocketLawyer, LegalZoom, and the like become more customized, and interact with people more like lawyers interact with clients, the UPL questions are going to get harder and harder. The natural language aspect of the parking ticket thing feels to me more like legal practice: you can easily imagine a client trusting an interactive, English-speaking app more than they might trust a more web 1.0 or 2.0 system of drop-down menus and such. Are the regulators going to quash this (especially now that he's looking to expand to New York), or are they going to get out of the way?

Second, to me, this level of legal tech innovation seems like an unmitigated good. Is there anyone scrutinizing the behavior of parking enforcement authorities right now (given that it's far too small-fry in most cases for lawyers), or is the parking ticket system in many cities nothing but taxation by another name, buttressed by the total lack of any real opportunity to challenge them? Browder might look closer to his temporary home, given that San Francisco is kind of notorious for its abusive parking tickets and they've been resisting the use of other automated systems to squeeze out a droplet of due process from the machine. As I've argued previously on this blog, nickel-and-diming people to death with penny-ante law enforcement directed at ordinary day-to-day behaviors is a threat to the kind of ideas underlying the rule of law, and maybe software can fix it where lawyers can't.

Third, to fellow prawfs: as folks like Dan Katz and Oliver Goodenough keep reminding us, this is coming to the rest of the law. Right now, the advances seem mostly to be looming over the discovery process, with stuff like predictive coding threatening to be the second level of the inexorable process of stripping the legal profession of the rents generated by document review (where outsourcing and offshoring were the first), as well as to relatively small-scale stuff like parking tickets, leases, etc. for small players. But as the technology gets more sophisticated, it has the potential to supplement or replace lawyers in more areas of law. (Right now, the most hubristic claims are being made by an early-stage startup called Ross... but what happens if those claims turn out to be even sort-of true?) What can we as law professors do about it?

One option is to get a lot better about teaching our students to be more comfortable with technology, as users as well as creators, even to the point of trying to teach them programming and machine learning. That's a strategy I'm interested in exploring further, but I also have some skepticism about it. It doesn't obviously follow from the danger of technology supplanting lawyers that the lawyers who will be best positioned to survive are those who are capable of operating in both domains. Whether that's true depends on the shape of the ultimate market: will it actually demand people with both legal skill and technological skill (perhaps to translate from one to the other), or will it favor people with pure technological skill plus a handful of really good lawyers to handle the most high-level work? My crystal ball isn't sharp enough to tell me, though I'm encouraging my students to tech up to the extent possible in order to hedge their bets. But what else can we do?

Posted by Paul Gowder on February 24, 2016 at 06:55 PM in Life of Law Schools, Science, Teaching Law, Web/Tech | Permalink


Post a comment