Some people think Large Language Models will transform the practice of law. I think it's bigger than that.
Excellent read, agree with a lot of your points. We launched www.spellbook.legal last summer and have onboarded nearly 1000 law firms/solos onto the platform. We have very much taken a "lawyer-augmentation" approach like GitHub Copilot has for programmers. We are not so much an assistant or co-pilot, but a precision power tool that enables a lawyer to do bespoke tasks much more effectively. The feedback from our users has been incredible.
One trend we've noticed is that small firms are much, much more receptive to this. Many of them have been billing on a flat fee basis for years at this point, and they need all the help they can get to improve their margins. It is BigLaw hourly billing that is at stake, but literally 1000's of small firms/solos are adopting this technology en masse. This tech also makes it more viable to operate as a solo, since you would no longer need so many support staff.
Looking forward to reading more of your takes!
If A.I. replaces lawyers I'm not sure anyone other than lawyers will care (which I say lovingly to my fellow lawyers because it's just true, even if we don't like to be believe that). 😏
Critical thinking will not be replaced by A.I. and most of what good lawyers do relies on this hard-to-develop skill. Here are the 3 levels of modern knowledge work in a nutshell:
1. Information & Data processing (low-level, automatable, increasingly harder to monetize, except at scale)
2. Pattern Recognition & Synthesizing of Information (mid-level for now, and also getting harder to monetize, especially now that GPT is rapidly developing)
3. Complex problem solving based on critical thinking (high-level and will be for a long time; lucrative and largely immune from large scale disruption)
Really interesting article, thank you. As the technology may increasingly be deployed in review and response contracting it will be fascinating to watch the output of the battle of the bots - where will the output of two powerful models, largely unconstrained by resource tend to end up? Increasingly long and complex arrangements or cutting the knot?
Completely agree with you, Jordan, that we will look at pre and post GPT as an inflection point. I'm not sure it will be GPT that is the ultimate tool, of course, so perhaps it is "pre-what GPT portended" and "post-what GPT portended), but it's the same idea.
"Gaming the system" has become a perjorative term, but let's face it, a large part of what good lawyers do is just that. We sometimes frame it in terms of finding loopholes, but what really good lawyers do is figure out the rules, find any rules or lack thereof which our client can benefit from, and use that information. In other words it is the creativity and idea aspect which provides the ultimate value. It will be a while, maybe a long while, before AI has similar capabilities to humans in this way, but any human lawyer that can create a unique idea and then apply it, probably using AI, will rise to the top. Lawyers won't be replaced, but they will need to find a value proposition that AI can't achieve.
I’m pretty sure you’ve long believed that “… lawyers will need to re-imagine who we are, what we do, and what we’re for” but I agree that these recent AI developments have likely made that need all the more real and pressing.
I wonder what will happen to law societies across the board, since lawyer license testing seem to be their main source of income. And if a bot can collect and analyse information within minutes, think of what that will do for a company or small business that now employs legal compliance enforcement, employment law standards, OH&S laws...the implications reach far past lawyers into every business entity that is regulated by fed/prov laws. Lawyers should be working on their sales personalities...because clients are always looking for that voice they trust. That voice doesn't come from a bot.