Reverse Engineering Legal Logic (Live from ReInvent Law)
This is a live post from the Reinvent Law Silicon Valley conference in Mountain View, CA. Please forgive any typos or errors in conveying what speakers say. Now up, in a six minute Ignite-style talk, is Kingsley Martin, CEO of KM Standards (formerly Kiiac) on Reverse Engineering Legal Logic.
Until today, we have followed inductive approach to law. But deductively, we can read documents to find out what they mean and what’s common. Kingsley says he is close to breaking the “subjectivity barrier”. Just as we will have driverless cars, we can have machines that draft contracts.
Some argue that we need intuition to think. Kingsley: when you measure by outcomes, does it matter what the mechanism is?
Find, analyze, and optimize are the three steps to make a decision or draft a document. Consider as example of buying a car. First, find info on all cars. In past, this would have required trip to library or consulting friends. Today, we solve by Internet search. Second, analyze the cars. With Amazon and other tools, we can use “faceted search” to narrow results and compare results (“Structural Classification”). To optimize, you need to be able to pull data from documents (from search results) and interpret.
Technology can find contracts – that’s easy. Most of Kingsley’s effort is to analyze similar sets of documents to identify relevant elements and come up with a checklist or template of common elements. Analysis also assesses how similar or dis-similar each clause is. “Standard” clauses have little variation in clauses across documents. Where there is heavy variance in language, it either reflects heavily negotiated clauses OR by lawyer personal preference. Some clauses are relatively rare but these have little variation. A fourth set of clauses are both rare and vary – these are deal specific.
A computer cannot draft a document from scratch – but it can pull clauses and combine. Computer can also reverse-engineer the logic to derive the questions that need to be asked. Beginning the journey to be able to data mine out of documents all of the necessary clauses. Will be able to benchmark, provision by provision, whether clauses meet market standards or raise issues in courts.
Let’s assume that today, humans are more capable than computers. But with computers doubling in capacity every 18 months… they will overtake humans. [RF: this is the Ray Kurzweil Singularity argument.]
- Alternative Legal Provider (32)
- Artificial Intelligence (AI) (38)
- Bar Regulation (13)
- Best Practices (39)
- Big Data and Data Science (7)
- Blockchain (7)
- Bloomberg Biz of Law Summit – Live (6)
- Business Intelligence (19)
- Contract Management (19)
- Do Less Law (36)
- eDiscovery and Litigation Support (165)
- Extranets (11)
- General (187)
- Innovation and Change Management (144)
- Interesting Technology (93)
- Knowledge Management (213)
- Law Department Management (12)
- Law Departments / Client Service (109)
- Law Factory v. Bet the Farm (26)
- Law Firm Service Delivery (99)
- Law Firm Staffing (24)
- Legal market survey featured (4)
- Legal Process Improvement (21)
- Legal Project Management (25)
- Legal Secretaries – Their Future (17)
- Litigation Finance (5)
- Low Cost Law Firm Centers (18)
- Management and Technology (177)
- Notices re this Blog (10)
- Online Legal Services (62)
- Outsourcing (132)
- Personal Productivity (37)
- ReInvent Law (10)
- Roundup (58)
- Structure of Legal Business (1)
- Supplier News (13)