Will Artificial Intelligence Replace Lawyers?

A friend recently asked (via social media) if artificial intelligence (AI) would replace lawyers. Here's how I answered:

Much of what I personally learned in law school could be handed off to today's intelligent machines. So yes, New World machines will likely take over many of the functions of the Old World retail-level lawyers (i.e., many of the tasks that are currently charged by the hour could be resolved by smart machines programed to handle transactional and common inter-personal conflict-resolution services in milliseconds). But will governments give up the human element in deciding who goes to jail? And won't data monetization (spurred by the rise of the IoT) also create a new role for government-approved humans as arbiters of disputes? How will threats to the established order be dealt with? How will the rights and obligations of 21st century data flow be monetized? How will complex multinational disputes be resolved? Will we trust machines to regulate machines? When people hurt each other or threaten the social order, when political systems clash, when the new algorithms fail - and blood and data are spilled - who will get the call? Will we send in machines or people to clean up the inevitable messes?