Last week, I had the chance to use Westlaw’s Precision with Artificial Intelligence. I’m glad I tried it. And I’m impressed with its capabilities. But I’m also skeptical that it can keep up with a skilled attorney.
Lexis and Westlaw have been around since I was in law school. In fact, I never learned how to use the physical reporters to conduct research. So I’ve been researching online for about 20 years now. Last year alone, Matt and I spent over 1,200 hours researching on Westlaw, and that doesn’t even include the time we spent in other sources, like Google Scholar or PACER. And it doesn’t include time spent reading hard copies of cases (I still print cases I want to read later).
I know that I’m nervous to change the way I have researched for the past 20 years. Asking Westlaw a plain language question seems wrong. I’m conditioned to think in boolean terms. How many words must whistleblower be from elements or summary judgment? That’s a question I’m comfortable asking. And I’ve gotten good at using my methods. For that reason, I’m tempted to avoid generative AI to assist in research. But I also know that 10 years from now, I won’t have the choice.
Over the next decade, I think that good lawyers will still conduct a lot of research. And certain things will remain; good lawyers will do good research while mediocre lawyers do mediocre research. But I also think that AI help good lawyers get to the point faster and save time. I’m just not sure that the time is now.
For now, I see the advantage of lawyers doing their own research with minimal help from generative AI. Surely all research uses some AI. Google, Bing, Lexis, WestLaw, they all use AI in some way. But I’m not comfortable letting that AI lead the way. ChatGPT is only one year old and still going through growing pains. It’s just too early to trust one of the most important legal tasks to an unproven technology. But a year from now, I may be writing a different blog post. Stay tuned . . .
~David
