3 ms·

As a lawyer, I am always excited to see innovation in the legal space. This seems like an interesting concept: using an algorithm to summarize a document (cont

As a lawyer, I am always excited to see innovation in the legal space. This seems like an interesting concept: using an algorithm to summarize a document (contract) and to flag unusual clauses. Put the document in, and out comes the legal blessing. However, I am a little concerned about the viability of this concept in terms of producing a valuable analysis. A true analysis considers the context of a document. For instance, a one-year term is a common residential lease clause. However, what if the tenant was under a 6 month employment contract and needed to change locations? The algorithm would not know this and would not flag the clause. Or what if the tenant was 17 years old and couldn't legally sign contracts in the first place? Sometimes interpretation of a contract comes down to the exact wording of a contract. Most boilerplate language is the result of years of evolution and precedent. A summary might subtly change or obscure the meaning. Finally, you are getting very close, if not into, the unauthorized practice of law. I recommend contacting your state bar association for feedback. That being said, I input my email address into your website and look forward to monitoring your progress.

Balgair 12y ago
Think of it as a first pass. Yes, you'll have to go over it again anyways, but the 'flags' it pulls out give a little bit of landmarks and a tiny bit of extra awareness that you may have missed. It's probably going to have a error rate much higher than yours, but it's errors have a chance of being not yours and it may catch things you missed. The real question is how much better of a lawyer does it make you? Can it save you 1 hour a day, 1 hour a week, 1 hour a month? And if it can, how much does that cost you in terms of billable hours or whatever your rates are like?
- joshuaheard 12y ago
  If this service is for lawyers, then it could have value. However, most lawyers pass their time costs onto their clients, so they have no incentive to be efficient. I would aim at corporate lawyers, although it's a very small market.
  - Balgair 12y ago
    I can then see the aim to the mass market. It is something that in the very least gives you confidence in the contract in some small way. If the software gives it a pass, you can then sorta assume it is somewhat legit. However, if it gives a pass with conditions, you can then have something to push back on the other party for. It'll give people some confidence to argue their own rights, even if it is terribly misplaced and wrongheaded, it might get people to read their contracts.
Chevalier 12y ago
Fellow JD here. I have to disagree with you re: viability. Not only does most of our work lend itself well to automation (discovery, researching precedent, and predicting case outcomes), but this particular startup is aimed squarely at the most obvious domain where humans mostly aren't needed anymore: formulating contracts. Translating boilerplate legalese into natural language and running both parties through checklists of essential elements seems like an obvious solution to an expensive problem. I really don't see why 99% of contracts can't be automated outside of high-level practice. Nor do I really understand your examples. I can't recall the last time anyone I know hired a lawyer to sign a rental agreement, nor do I see any reason why contract software wouldn't first check that both parties are legally competent. Finally, I'm pretty uncomfortable with the idea that "if you automate our profession, our protectionist guild will go after you." You're not wrong, and it would be wise for LawGeex to stay wary -- of all trades, the lawyers are going to fight tooth and nail to stay relevant -- but startups like LawGeex or RocketLawyer are clearly far more in the public interest than our ludicrously expensive middleman services. We may as well ban self-driving cars because the Teamsters want to keep dangerous, error-prone human truckers.
- will_brown 12y ago
  >I can't recall the last time anyone I know hired a lawyer to sign a rental agreement I know the OP was talking about residential leases; however, in the span of ~3 years I would say I have been retained ~100 times to review/negotiate/revise commercial lease (rental) agreements. The real value would be comparing the results of LawGeex with a lawyer's legal opinion for the same contract. Even then, it is easy enough to say one could compare 2 attorney legal opinions of a contract and have drastically different opinions, or that while a LawGeex contract review may be identical to an attorney on one contract, it may not on another.
- joshuaheard 12y ago
  Yes, you would have to have some sort of questionnaire for the user to complete, along the lines of the TurboTax wizards or legal document assembly software, with questions based on the algorithm's analysis of the contract. I didn't see anything like that on the website, but it is certainly doable. At the moment, the website is only accepting residential real estate contracts which is why I used them as an example. As a member of the State Bar of California, I can only wish it acted as a guild protecting my interests. Quite the contrary, it has an adversarial relationship with its members from overseeing the attorney discipline system to enforcing continuing education rules. It is more about consumer protection than attorney protection. I see your point, though.

Balgair 12y ago

Think of it as a first pass. Yes, you'll have to go over it again anyways, but the 'flags' it pulls out give a little bit of landmarks and a tiny bit of extra awareness that you may have missed. It's probably going to have a error rate much higher than yours, but it's errors have a chance of being not yours and it may catch things you missed. The real question is how much better of a lawyer does it make you? Can it save you 1 hour a day, 1 hour a week, 1 hour a month? And if it can, how much does that cost you in terms of billable hours or whatever your rates are like?

joshuaheard 12y ago
If this service is for lawyers, then it could have value. However, most lawyers pass their time costs onto their clients, so they have no incentive to be efficient. I would aim at corporate lawyers, although it's a very small market.
- Balgair 12y ago
  I can then see the aim to the mass market. It is something that in the very least gives you confidence in the contract in some small way. If the software gives it a pass, you can then sorta assume it is somewhat legit. However, if it gives a pass with conditions, you can then have something to push back on the other party for. It'll give people some confidence to argue their own rights, even if it is terribly misplaced and wrongheaded, it might get people to read their contracts.

Chevalier 12y ago

Fellow JD here. I have to disagree with you re: viability. Not only does most of our work lend itself well to automation (discovery, researching precedent, and predicting case outcomes), but this particular startup is aimed squarely at the most obvious domain where humans mostly aren't needed anymore: formulating contracts. Translating boilerplate legalese into natural language and running both parties through checklists of essential elements seems like an obvious solution to an expensive problem. I really don't see why 99% of contracts can't be automated outside of high-level practice. Nor do I really understand your examples. I can't recall the last time anyone I know hired a lawyer to sign a rental agreement, nor do I see any reason why contract software wouldn't first check that both parties are legally competent. Finally, I'm pretty uncomfortable with the idea that "if you automate our profession, our protectionist guild will go after you." You're not wrong, and it would be wise for LawGeex to stay wary -- of all trades, the lawyers are going to fight tooth and nail to stay relevant -- but startups like LawGeex or RocketLawyer are clearly far more in the public interest than our ludicrously expensive middleman services. We may as well ban self-driving cars because the Teamsters want to keep dangerous, error-prone human truckers.

will_brown 12y ago
>I can't recall the last time anyone I know hired a lawyer to sign a rental agreement I know the OP was talking about residential leases; however, in the span of ~3 years I would say I have been retained ~100 times to review/negotiate/revise commercial lease (rental) agreements. The real value would be comparing the results of LawGeex with a lawyer's legal opinion for the same contract. Even then, it is easy enough to say one could compare 2 attorney legal opinions of a contract and have drastically different opinions, or that while a LawGeex contract review may be identical to an attorney on one contract, it may not on another.
joshuaheard 12y ago
Yes, you would have to have some sort of questionnaire for the user to complete, along the lines of the TurboTax wizards or legal document assembly software, with questions based on the algorithm's analysis of the contract. I didn't see anything like that on the website, but it is certainly doable. At the moment, the website is only accepting residential real estate contracts which is why I used them as an example. As a member of the State Bar of California, I can only wish it acted as a guild protecting my interests. Quite the contrary, it has an adversarial relationship with its members from overseeing the attorney discipline system to enforcing continuing education rules. It is more about consumer protection than attorney protection. I see your point, though.