Certification regimes present a barrier to full automation, but with so many young and unemployed lawyers, you could easily get around that as well. Just do most of the stuff on the computer and let Mr. cheap-young-JD sign off on the last mile work on his phone. Each contract he goes through knocks another $100 off his student loan bill. Use the law schools' oversupply to break the legal industry and bring cheap legal services to all.
The main threat will be regulation. Putting lawyers out of work will mean they will try to rewrite the law to ban you. Depending on how you structure it, you might want to ally with a few firms in each vertical to rain fire upon their competitors. As long as you divide and conquer, and some lawyers are profiting from the new regime, it will make it much more difficult for the ABA to achieve consensus on the imminent threat your startup undoubtedly poses to the health of the republic.
EDIT: Even more interesting...as a dotcom you may also be able to market your services in a way that normal lawyers cannot.
http://en.wikipedia.org/wiki/Legal_advertising#In_the_United...
There has been an oversupply of legal gradutes for decades now. Today, half of law graduates don't get a job, and 90% can't get a job at a big firm. Yet, big firms are still paying $160k+bonus to people fresh out of school, because that's what they have to pay to get kids out of the top 15-20 schools or the tippy-top kids from the lower-ranked schools. Why do they continue to do this when lowering their standards for pedigree could cut their costs massively?
You can say that it's because law firms are stupid and invested in the status quo, but corporate clients surely aren't. Why do firms persist in putting so much stock in hiring Harvard/Yale/etc graduates, and why do clients keep hiring such firms? If there was really an opening in the market you could drive a truck through, what has stopped anyone from taking advantage of it thus far?
The work that is truly straightforward can be automated... and is. For example, in Australia at least, you can get a DIY wills pack.
The essential need for humans in legal work is integration: relating a fact situation with relevant laws. The fact situation is not pre-recorded according to a grammar - and the semantics of the law are not precisely defined. Both aspects and their connection are far more flexible and ambiguous and uncertain than any IT integration. Possibly similar complexity though.
tl;dr we'll have automated IT integration long before we have automated legal work.
Why do they continue to do this when lowering their
standards for pedigree could cut their costs massively?
If there was really an opening in the market you could
drive a truck through, what has stopped anyone from taking
advantage of it thus far?
Then your less-skilled JDs can review the documents mass produced by software, in a sort of legal assembly line. They may not be able to follow chains of complex reasoning, but that's ok: engineers + Stanford JDs already made sure the legal templates and algorithms worked on representative data.
In other words, we are making use here of the old joke - "What do you call the guy who got the lowest GPA in medical school? Doctor."
Similarly, the less-skilled JDs who can't get jobs today do have one very important asset: they passed the boolean threshold and are licensed to practice law. Hollow out everything else, replace it by software, and use this army as scalable last-mile reviewers with the thinking done by code. Their value-add is that they absorb the last mile liability, as they are doing final review before release to client.
This is no different than the way that Intuit replaces the CPA in many situations. Kind of a TaskRabbit for law. By giving jobs to (a) young tech-savvy jobless students and (b) a few top attorneys from the very best law firms to write the contracts, you can do a pincer attack and massively reduce legal costs.
Corporate clients will continue to hire the best lawyers and pay high fees when and because the company is on the line.
Most individuals, on the other hand, do not need a Yale Law grad billing $900 an hour to set up a living trust or LLC.
Obviously there's a whole spectrum in between. Just flagging that "Tech will utterly destroy the legal industry" and "We'll always have top lawyers starting at $160K" may be talking past each other a bit.
Apparently, if you have ten years experience in a specialized role, where you went to undergrad still matters. That's just crazy. So in light of that, it's possible that everyone's just putting so much stock in those graduates for no good reason at all.
Or maybe it's for the same reason as McKinsey and the other management consulting firms. They're billing these fresh grads out at 5-10X what they're paid, and those clients want to feel like they're paying for the best and brightest.
As for whether there's an opening you could drive a truck through.. it depends on how much of the work is 'routine' vs how much actually requires a seasoned lawyer to get done.
This is true, and if anyone is curious about this issue and the data behind it, take a look at Paul Campos's recent book Don't Go to Law School (Unless), which you should also thrust into the hands of anyone you know who is contemplating law school.
Law 'feels' like code, its kind of the "legal fiction" we need to believe in to stay sane, but when it actually comes down to it, its anything but.
You may think that the law is necessarily more like politics, and that there's little room for code. But I don't see any reason for that in your comment, which is more just a reaffirmation that the status quo exists.
We haven't even automated medical records for godsakes!
What would be 'expectedly naive'?
Was it surprisingly naive for the explorers and founders of the New World to want to build new societies based upon their ideas? They packed up everything they owned and risked their very lives to go to somewhere lacking an entrenched society so they could build something new.
The effects of their experiment have been disproportionately dominant in the culture and growth of the world for centuries.
Now we're in a world where all the land is spoken for by government bodies that enforce their monopolies on governance through force. Imagine if government said, "There can be no new startup companies. You'll have to influence the existing companies from within." How would entrepreneurs feel about being unable to start new enterprises?
Worst thing that happens - a bunch of people that don't want to live in the US anymore drown in the ocean. Best thing that happens - they experiment with new forms of government, some of which influence people back on land. Either way, we need more experiments in government, not less.
US law is supposed to be an axiomatic rule based system, but it is not consistent or complete, and in practice it ends up looking sort of arbitrary. But I don't think this means it is naive to assume that automation will make the system more consistent, or that it will help to simplify much of the complexity.
I am not suggesting that law will not always have some complexity and fuzziness - in fact I think it will - but I don't hear Thiel as explicitly saying the opposite - that we will ever get to a point where there is no complexity or fuzziness.
Without disagreeing with your larger point, I can't help but point out you're bringing the wrath of Godel down on yourself here.
And that is, of course, one of the key functions of the legal system as a system of dispute resolution: allowing society to continue to function by helping people get over disputes.
My personal opinion is that you can get both with a more transparent system. In theory, a transparent system would drive consistency, and consistent outcomes would hopefully converge around correct outcomes.
There has to be academic groundwork to build something like that, a mixture of CS and law. Maybe some students at a law faculty could create a project together with a few CS students just to get the stone rolling.
A programming language to describe who is responsible for cleaning the flat.