Like so many other lawyer’s offices, my office was once cluttered with random piles of paper. Articles that I wanted to keep were in piles in one area, and various case files were in piles in others. It looked like a mess, but I could walk over to a pile and usually in a few moments find the document that I was looking for. This is the way nearly all lawyer’s offices were for centuries. Indeed, a lawyer who had a nice and clean office would be subject to suspicions about his competency by his colleagues.
Then one day this thing called a scanner appeared. The papers were converted into .pdf files and the piles moved onto hard drives. They were still piles, mostly disorganized, but at least they were digital piles. At first, the digital tiles were moved around physically. Mailing CD-ROMs and flash drives was common. Then, on another day, these piles moved to the cloud where they could be accessed from anywhere.
They were still piles. Instead of a stack of disorganized papers a foot high, now there was a file folder containing a bunch of disorganized .pdf files. Trees were being saved, but the disorganization persisted. It was sort of possible to search through these files, but this was limited to a word or a phrase. Even then, the search result would usually be just a smaller pile of documents to search through. This was not merely inefficient, but it was often impossible to find what one was looking for without a prolonged examination of numerous documents.
I’ll use myself as an example. Over the years, I’ve collected numerous articles, court opinions, books and even multi-volumed treatises on a variety of legal topics that I follow. My file on charging orders contains literally hundreds of these materials. I might easily find something by pure recollection, but that’s about it. Finding everything else is usually laborious. I’ll look at the file name and then decide if that is something that might hold the research treasure that I seek. Better than the piles of paper, but not by much.
Those days are over thanks to Google’s NotebookLM feature. All that is required is for me to open a new notebook for charging orders and then upload all my charging order documents into that notebook. This means court cases, law journal articles, and even whole books. Once those documents have been uploaded, then I can use Google’s Gemini AI to process all those documents pretty much any way that I want. More than just searching for terms through documents, Gemini will apply AI logic to digest those documents so that I can have a conversation with AI about specific topics.
For example, instead of just asking for all documents that discuss the foreclosure of a charging order, I can ask Gemini to put together a historical summary of the law relating to the foreclosure of a charging order lien. After a few minutes, Gemini will crank out what amounts to a structured memorandum which discusses that topic. By further questions, I can drill down to the exact issue that I want to examine, such as the foreclosure of charging interests by a bankruptcy court.
Here is something that is very important: When using NotebookLM, Gemini will only look at the documents that I have uploaded. Gemini will not go outside my documents to look for additional sources. This all but negates so-called GIGO (“garbage in, garbage out”), since presumably I have only uploaded credible sources. It also helps to prevent “hallucinations” by AI where sources are just made up. The Gemini output will also reference in a footnote-like fashion the materials upon which it has relied so that I can double-check that the source has actually said what Gemini has told me it says. It is like having a group conversation with judges and law professors who have written opinions or articles about charging order issues.
But that’s not all. Nowhere close. If I am invited to lecture on some topic, NotebookLM will take my materials and create a slide deck. NotebookLM will also develop a logical flow chart if I desire. NotebookLM can create a quiz that I can handout at my lectures. Even beyond all that, NotebookLM can create a podcast and even a video involving AI characters who chat about the issue that I want in a Q&A-style format.
Bottom line: NotebookLM is amazing for legal research. It’s like having the world’s fastest legal research assistant on hand 24/7.
There is a catch to all this. To get the answers that you want, you first have to ask the right questions. The questions are known as “prompts” and it takes a little bit of time to learn how to put together the right prompts. Prompts are sort of a combination of instructions and questions. This is not too dissimilar to 30 years ago when lawyers had to learn how to use the Windows 3.0 computers which later became ubiquitous. So, there is a learning curve, albeit a relatively shallow one, but it is also somewhat fun to learn the art of prompting by experimentation. If you want a really good prompt, Gemini will also help you develop that prompt if you ask it.
In future articles, I intend to discuss the deeper ramifications of this within the legal profession. There is a generalized fear in the legal community that products like NotebookLM will eliminate the need for younger lawyers and thus deprive them of their opportunities to learn the ropes of a legal practice. Another fear is that the ability to deeply analyze legal issues may atrophy. We’re already seeing lawyer being sanctioned for using AI to write their legal briefs without cite-checking the output, so these fears have a solid ground in reality.
If you’re wondering, I don’t have any special relationship with Google other than as a mere user of their products, I don’t own any Google stock, etc., and I did not talk with anybody at Google in relation to this article. I got into NotebookLM and Gemini all on my own. There are some good Youtube.com videos out there which you can watch that will make using NotebookLM and Gemini pretty easy. I watched some of those videos before starting to use these products. Also, competing AI platforms may offer equivalent or maybe even better services (I don’t know). Like all such things, you’ve got to find the one that jives with you the best. Oh, and no part of this article was created with AI.
If you haven’t tried NotebookLM ― or a product like it ― you should give it a spin. You’ll see that the world of legal research has changed dramatically.
Read the full article here
