Sunday, January 14, 2007

Automation of Legal Reasoning

Who doesn’t hate lawyers…? At least, I haven’t heard anyone praising them.

I think all of us have our own reasons to hate lawyers. When we approach them for a simple solution, they mire us deeper into a mess ; and pretending to get us out of it they’ll dish out loads and loads of legal bullshit and charge us by the hour for unleashing verbal diarrhea. It’s the same story everywhere.

Now that’s a global problem. Solution – Don’t go to them or need them. Make them irrelevant. Let systems based on Artificial Intelligence (AI) take it over from the black gown vampires.

Duh, Not so simple, huh…?!

Lawyers should know. From the days of Hammurabi, laws have never been written. They have been `coded’ – the legalese need interpretation as they cannot be decoded by normal humans armed with natural language prowess. It is this element of subjectivity which lawyers use to the hilt and drive you nuts. They instill the fear of unknown in you and push the horizon of your imagination to include wildest of outcome. When you are scared, you lose your perspective and hire him. This is his principal predatory tactic.

It is in this context the Plain English movement in legal writing is gaining a foothold, and experts are busy trying to debunk what they see as the myths of legalese. Considerable advances have been made in this direction, yet there’s a long way to go. Primarily because there’s a significant element of discretion involved in analysis of circumstances, stages of investigation, documentation, trial, arguments, evidencing and Judgements. If ever these processes could be structured and the element of discretion minimized, technology can make inroads and make life better for all litigants. We are looking at automation of jurisprudence itself.

Automation of Legal Reasoning explores the development which has led up to the formation of a joint field of artificial intelligence (AI) and law. In this undertaking, the basic foundations of AI and the methodological approaches found in jurisprudence are related to each other in a historical perspective. Discretionary domains challenge existing AI paradigms because models of judicial reasoning are difficult, if not impossible to specify. The systems for reasoning with this form of open texture can be built by integrating rule sets with neural networks trained with data collected from standard past cases. The obstacles to this approach include difficulties in generating explanations once conclusions have been inferred, difficulties associated with the collection of sufficient data from past cases and difficulties associated with integrating two vastly different paradigms.

It calls for introduction of an elaborated jurisprudential model of legal reasoning , reflecting different sub-processes and various types of legal knowledge exercising influence over them. The Swedish Law and Informatics Research Institute have done some research combining the two.

The investigation should lead up to the formulation of a design approach for advanced AI-systems for law, based on a functional decomposition of legal knowledge, the integration of various computational techniques and the structural integration of different types of small-scale AI-systems.
This I think would be one area which would find favor with venture capitalists, provided some seminal development can be undertaken by competent teams. VCs love large market sizes and if one looks at the huge backlog of cases pending before the courts, there cannot be a more salivating prospect. Every nation would love to have its legal machinery revamped so long as it would entail faster resolution of disputes, better conviction rates and efficient carriage of justice.