Speaker 1: As a lawyer, I deal with a lot of information, from the facts related to a client's case, to statutes and regulations and case law. It can be really overwhelming, but it's not just lawyers, it's all of us. As we live in this digital age, more and more information is accumulating. Whether you're a teacher or a doctor or in some other industry, we're all dealing with information overload. And even though in this video, we're going to talk about information and generative AI from a lawyer's perspective, even if you're not a lawyer, the information in this video might be helpful to you. You might need a lawyer one day. You might need a lawyer for a will or a traffic ticket or to buy a home or for some other matter. But also, lawyers are business people, and the way that lawyers are leveraging large language models may also give you some ideas how you might be able to use it in your business. So during this video, we're going to talk about the way lawyers are using these emerging technologies for gathering facts, researching the law, analyzing the facts and law, and creating documents. All right, so we're going to talk about gathering facts. So first, lawyers are problem solvers. We counsel clients, we give them advice, we generate documents, and we advocate for them in court if necessary. And the first step in lawyers solving people's problems is to gather the facts. Now, sometimes gathering the facts is pretty straightforward. Like if you're seeking a will from a lawyer, the lawyer would be able to gather that information from the client pretty effortlessly. It becomes exponentially more challenging if the information and the facts are included in digital documents or stored electronically. So we have electronically stored documents, or ESI. And this information and data is found in computers or in text messages. And because of the sheer amount of this electronically stored data, it can be really challenging for a lawyer to gather up all of those facts. And so lawyers have been using tools related to e-discovery for a number of years. But with the advent of generative AI and large language models, it really does help to extract the information more efficiently and also to oftentimes summarize the information that's found. So these large language models are able to help us accumulate the facts in a much more efficient way. So after gathering facts, lawyers need to determine the law. And so they need to find the relevant legal authority that addresses the problem that they're trying to solve. Now, when we say law, what do we mean? Well, we have statutes. And statutes are created by state legislatures or in the case of national law, by Congress. And we also have regulations. And regulations are promulgated by administrative agencies, and they provide the details and some specific instructions on how to fulfill the obligations of the statute. We also have case law. And so case law is judge-made law. And so when we think about all of the statutes that we have, all of the regulations that we have, and all of the cases that we have, that's a lot of information to process. And lawyers have to be able to do legal research, again, to be able to determine what law determines the outcome of the case that they're handling. Now, artificial intelligence has been used in legal research for decades, but generative AI has really supercharged legal research. So generative AI tools with their natural language processing capabilities are able to provide more precise search results and also provide more nuanced analysis of the laws, again, statutes, regulations, and cases. These legal research tools are also able to suggest potential arguments and interpretations of the law. But lawyers or anyone using generative AI or large language models to do legal research, they need to know the limitations, because these tools still have the tendency to hallucinate and create laws that don't exist. So the rules of professional responsibility require that lawyers review the law, make sure that the law is good law, which means that it hasn't been overturned, and to also ensure that the law supports the advice that they're giving and the arguments that they're making. So after a lawyer gathers all the facts and does legal research and determines the relevant law, the lawyer needs to analyze the facts and the law. And generative AI tools can really help a lawyer to determine the strengths and the weaknesses of the cases. And whenever we talk about lawyers analyzing cases, part of it involves prediction. And so lawyers need to think about whether their case is a strong case, a weak case, what strategies need to be employed. And in order to determine or to predict a case, there's pieces of information that a lawyer needs to consider. So the type of the case, the jurisdiction of the case, and even the judge, if this is a case that is being reviewed in court. Taking all of this information and using a generative AI tool can be helpful when the lawyer is trying to make these predictions. Now, one thing that's really important to emphasize is that generative AI tools, large language models can assist and augment the work of lawyers, but never replace lawyers. We need to make sure that at every step of the way, you have a lawyer, you have that human judgment, and you have the work that's ultimately being done by a person. All right, now, after the lawyer has gathered all the facts, done the legal research and found the relevant law, and analyzed the facts in the law, it's time for the lawyer to draft the documents. And large language models can be incredibly helpful with the initial drafting of documents. But the keywords there are initial and draft. Because any document that's not created by the lawyer has to be reviewed by the lawyer, refined by the lawyer, and the lawyer has to ensure that the document meets the specific needs of the client. Because a lawyer's expertise, judgment, and ethical responsibilities can't be delegated to artificial intelligence. All right, so we've talked about how lawyers are using generative AI and large language models to gather facts, to do research to determine the relevant law, to assist them in analyzing the facts in the law, and to assist them in creating documents. So I just want to cover a couple of final, really important considerations. The first is that anyone who's using large language models needs to make sure that they understand the privacy implications of using these tools. And lawyers in particular have to be especially vigilant because they have an ethical obligation to protect the confidentiality of their client's information. And again, that goes back to their professional responsibility. So it's important to be aware of the AI governance systems associated with any models that are being used. I also want to emphasize that a lawyer needs to be involved at every stage of the process of using generative AI tools, and that any output that's being used needs to be reviewed by the lawyer carefully. While these tools are really helpful to create efficiencies, they absolutely have their limitations. Thank you for watching. Please remember to click like and subscribe. Thank you.