Thoughts on lawyering and AI
Welcome to Ideothetic Flow, my newsletter sharing my reflections on finding balance, sufficiency, and security.
If this resonates, do subscribe so we can stay connected. I post every 4 weeks.
Hi!
My thoughts on lawyering in a world of generative AI
The last few months, it's impossible to go through a day without hearing about AI and ChatGPT. I'm not sure how many comments I've fielded about chatGPT being able to pass law school, or being used to create wrong legal submissions.
Robot takeover aside, what scares me about generative AI is the content overload that will follow once we can infinitely generate information with a click.
What happens when we solve supply-side challenges
I'm too young to know the industrial revolution, but I remember how, as a child, many things were not available to buy in Singapore. People going on holidays to America or Japan would get my wishlist of Pokemon or Warhammer merchandise that I'd never be able to get my hands on in Singapore.
This is a rare problem today. Improvements in production, supply chain, and market entry have made almost anything available in Singapore. The only souvenir I ask for now is coffee beans.
This is the same for content too. It's easy to make and distribute content, until even someone like me can write an email newsletter. There's far more to watch than we can consume, and sadly, much more of it in low quality. Most people find themselves figuring out how to stay productive and not fall into the trap of doomscrolling feeds, which we all admit wastes time, but do it anyway.
Legal practice also solved supply side issues. With email and digital word processing, its too easy to create and edit documebts. We have a deluge of NDAs, cookie consents, and t&cs for everything. These are accompanied with back and forth communication to perfect the document, or quibble over words until we get to “finaldraft_v7.3b(trk)”.
Generative AI = infinite documentation?
The worst thing we could do is make use of AI to go even further in practising law this way. We can now make documents infinitely long, and generate something new for every human interaction. Buying a fast food burger could come with its own unique AI contracts.
The same could be said of all other things AI touches. We can already see automatically generated music, videos, or books. At least for these, their consumption is the point. We still must interact directly with them, and can make a decision on whether their output is valuable to us.
Law is different in that, no one actually wants to interface with what we produce. There is no world that is better simply because it has more contracts, more rules, more legal principles. It exists because it is necessary to fix the imperfections of human interaction.
Legal work is thus only as important as how seriously people take it. If we enter a world of infinite legalese, that seriousness is lost. The harder it gets for the average person to interface with a legal document, the more likely it will be disregarded.
Lawyers are important now for our ability to bridge that gap, and interface with the law on behalf of everyone else.
I often remind colleagues that I dislike reading contracts as much as they do, I just happen to be paid to do it. We survive it by our training making us more tolerant (or even appreciative) of the unnecessary complexity and detail better than everyone else.
Can we keep up though? If we constantly expand output, at some point we won't be able to interface with pur own material at a reasonable rate. I could work a 400 page one. But what happens when buying that cup of coffee (an essential tool for reading a 400 page document), comes with 4 million pages of legal text underlying it? Even if it were possible, maybe with other AI tools to help, the value of the legal interaction is now dilute.
Recalibration and editing
I think as a profession, we need to calibrate away from complexity and over-inclusion. To stop trying to solve things by doing more. We have to be comfortable with making things simpler, and more usable. I recall explaining a contract to a business manager once. She then said “i understand what you're saying, but why didn't you write it that way?" Truth is, if I did, my own bosses would probably think I did a bad job.
We also need more standardisation. So much time is wasted because everyone practices slightly different, often without any reason, or sometimes just for a marginal advantage. Two documents could say the same things but in a completely different manner. Take a hundred lawyers and we get a hundred different sets of boilerplates to the same effect.
Standardisation is crucial so we can make the tools work. We have a deadly mix of infinite (and likely unnecessary) variance, and pinpoint precision in language, which slows down our ability to create and adopt tools that can support our work. Right now, even a tool to automate the review of something simple like a NDA involves alot of initial effort to set up and make it work to a level where it would be useful.
Google's CEO talked about the future of AI as something working alongside humans. It's pretty cringe for any lawyer to bring up the drama Suits, but that's also how I imagine things could be. Harvey Spectre strategizing, finding solutions, dealing with people, supported by Mike Ross’ supercomputer brain. We now have a chance to actually have that supercomputer brain next to us, as we work in what (at least in my view), is the fun part of lawyering.
Or, we could go the other way. We use all this power to generate mountains of legalese until we drown in it too.
I believe law remains an important tool to shape behavior and make social life possible for humans. Just look at any hawker center and see how we have become a nation where trays are returned in less than a year. But, its an open question if lawyers can continue to be useful in this discipline, or could we end up as the first victims of a robot takeover, when we build institutions we can no longer navigate.
Other things…
Semantic Satiation
The phenomenon of looking at a word too much until it suddenly loses meaning. I experienced a few weeks back with the word “Schedule” as I was trying to edit a document. This in part inspired my post above.
For tarot based reflections and journal prompts, check out @thecenterline_ on IG
An open invitation for conversation.
If you would be interested in having a short, targeted chat about anything, drop a reply and i’d be most happy to schedule a coffee or a video call. Perhaps there was a post that resonated which you want to unpack, a problem you want to talk through, or just want someone to listen to whatever is on your mind.
The agenda is purely up to you, and I promise a duty of confidence over anything shared.
Thanks for reading! I’d love to hear your thoughts, start a conversation, or simply connect over a chat. You can reply this email, leave a comment, or reach me at jameschanwz@hey.com.
If you enjoyed reading this and would like to support my writing, do subscribe or share this with a friend.
Take care and have a good week!
James