On Artificial Intelligence: Part 1
About the last thing you might want right now is one more commentary about AI – artificial intelligence. But humor me, because there is one small corner of the discussion which is mostly overlooked.
Anyone who has used ChatGPT or any of its cousins to create a “photograph” of a horned albino monkey riding a green bull, or write a sonnet about a dinosaur in Toronto, or create recipes based on the contents of your refrigerator, knows that this stuff is mind-blowing. And of course, as humans, we always begin with over-reaction. That’s the crocodile brain talking.
Lawyers, especially in Common Law jurisdictions, understand the foundations of AI perhaps better than anyone else, because we have always done what AI does: rely on history to solve the present problem. We call that stare decisis, or precedent. What should we do if a landowner fails to take reasonable precautions to prevent the escape of dangerous things onto a neighbor’s property? Well, we look to the Rule in Rylands v Fletcher, derived from a case in 1868.
As it turns out, AI is very good at doing that “research the past” part of the lawyer’s work. Even five or six years ago, rudimentary AI was doing legal research faster and more accurately than law students. So, why don’t we just “kill all the lawyers”, as Shakespeare proposed, and replace them with AI?
Well, that’s because there is something else that lawyers do (and you do, too). AI is not there yet, if it ever will be. While good lawyers (and all good thinkers) rely on the past for the grist, they go further and ask the “What if?” questions. While AI is poring over all the books in the library, the creative amongst us are gazing out the window, wondering about connections that nobody has made before.
In my personal panoply of legal gods, none stands taller than the late Gordon Henderson, who was famous for his opening line in many a difficult case, “Your Honour, I know that the law is against us this morning, but we are going to change that.” In other words, he was inviting the judge to re-think precedent and its premises and to consider solutions that better fit the case at hand, both practically and morally.
This is the stuff of great advocacy, and it’s the stuff of all higher-level human creativity. The past may be crystal clear but the future is always murky. We proceed only by our intuitions, our hunches, and our dreams, with faith and morality our only compasses.
In the unfettered ferment of our subconscious is a rambunctious tumble of “what ifs” and “why nots” and wild creation, sometimes painting the absurd canvas of our night-time dreams, sometimes spilling out in Eureka! moments.
AI is very, very clever when it slices and dices the past into a new present, but for now, at least, it doesn’t own the future. We do.