An Interview with ChatGPT on the Future of Law Practice

Episode Notes

Transcript

[Music]

Intro: Welcome to the Digital Edge with Sharon Nelson and Jim Calloway, your hosts, both legal technologists, authors and lecturers, invite industry professionals to discuss a new topic related to lawyers and technology.  You’re listening to Legal Talk Network.

Sharon D. Nelson: Welcome to the 180th edition of the Digital Edge Lawyers and Technology.  We’re glad to have you with us.  I’m Sharon Nelson, President of Sensei Enterprises an information technology, cybersecurity, and digital forensics firm in Fairfax, Virginia.

Jim Calloway: And I’m Jim Calloway, Director of the Oklahoma Bar Association’s Management Assistance Program. Today our topic is an interview with ChatGPT on the future of law practice.  Our Guest today is Julie A. Bay, the practice management advisor of the Oklahoma Bar Association.  She works with our OBA management Assistance Program to provide assistance to attorneys in using technology and other tools to effectively manage their offices and also teaches a wide variety of CLE classes.  Julie is also involved with the access to Justice Initiatives of the OBA such as Oklahoma Free Legal Answers.  Thanks for joining us today, Julie.

Julie A. Bays: I’m glad to be here today.

Sharon D. Nelson: Well, Julie, like you, I’ve been fascinated by ChatGPT, so let me give you my first question to ChatGPT.  If you were acting as a lawyer, you would be required to be a zealous advocate and what does zealous mean to you?  Would you consider lying or shading the truth?

Julie A. Bays: And Sharon I thought this was pretty interesting because over the last month and a half using ChatGPT, it’s starting to give better answers, I think, in cases like this.  So, what does zealous mean? Well, first ChatGPT says, as an AI language model.  I do not have personal opinion but I can provide information based on the definition of zealous and then it goes on to describe what zealous means.  But it then gives us disclaimer at the end of this, it says in short while lawyers are expected to be zealous advocates for their clients, they must always do so within the bounds of the law and ethical standards.  And so, ChatCPT basically pledges on this one because it just goes back to the rules of the model rules of professional conduct and just gives you a basic definition of zealous and explains that it does not have a personal opinion, which I think is pretty funny.

Sharon D. Nelson: Yes, it is.  Yes, it is.

Jim Calloway: Well, Julie given what we talked about in your introduction.  Do you think AI can help solve issues relating to access to justice?

Julie A. Bays: As somebody who deals with this subject on a daily matter here at the Oklahoma Bar Association, I do.  I think this is a great application and so what did Chat have to say about it?  Well, it says, yes, AI can play a role improving access to Justice.  Here are a few ways in which AI can help. Number one, automating legal processes.  AI can automate routine tasks such as document analysis and case management, freeing up lawyers and judges to focus on more complex issues.  Number two, providing legal information.  AI powered virtual legal assistants can provide individuals with access to accurate and up-to-date legal information.  Hmm, number two, we might need to think about, we’ll go back to that in a minute.  Number three predictive legal analytics.  AI can be used to analyze large amounts of legal data helping to identify patterns and predict outcomes which can assist judges and lawyers in making more informed decisions, true.  Chat bots, an online dispute resolution.  AI powered chat bots can help individuals resolve disputes online reducing the need for an in-person court appearances and increasing access to justice for those who may have difficulty traveling to a physical court location.

I agree with one, three and four.  I think that it’s already proving that AI could probably assist in automating legal processes, with chat bots and analytics.  For instance, in the predictive legal analytics, what we know now is Ed Walters with Fastcase recently said that he docket alarm which is their Docket product is now using GPT-3 to summarize large volumes of documents and so, i think that that is a way to increase productivity in a law firm.  But number two, providing legal information, that could be a problem because we found out over the last month and a half that sometimes ChatGPT sort of just makes things up.  So, you need a lawyer to analyze answers that are required up-to-date legal information and I think there’s a question later on that we’ll get to that a little more.

(00:04:55)

But at the end of this question, the funny thing is it gives a new disclaimer which I had not seen until I asked this question is, well AI can help address some of the challenges faced by the justice system.  It is important to remember that it is not a substitute for human judgment and should be used in conjunction with the expertise of legal professionals.  And I think that’s pretty good.  I think over a month and a half of legal experts playing with Chat, it has started to use those disclaimers a little more.  What would you say Sharon?

Sharon D. Nelson: Yes, absolutely.  In fact, I’ve actually asked the AI itself.  Why do I ask the same question and get a different answer every day?

Julie A. Bays: Exactly.

Sharon D. Nelson: And it tells me that it is changing and evolving and getting input and so forth.  So that’s the answer as much as we would like it to say stable.  It can’t stay stable.  That’s not its nature it learns.  So, let’s go ahead with another question, Julie.  Can AI replace lawyers with simple legal test for consumers?

Julie A. Bays: This is where we get in where chat has evolved.  It says, AI such as myself, can assist with simple legal tasks, but it is not yet advanced enough to replace lawyers.  AI can provide information and guidance, but it cannot provide the same legal level of analysis and judgment as a trained and experienced human lawyer.  Additionally, the legal system is complex and constantly changing making it difficult for AI to keep up with the latest developments.  And that’s true.  And something that I had asked it regarding a legal question that I knew the answer to that was easy for me to answer was used car sales.  I asked it, but if you buy a defective used car, what is your remedy.  And ChatGPT, originally, the first answer was to go to Oklahoma State’s Lemon Law.  Well, the Lemon Law in Oklahoma does not apply to used cars.  So, I asked it again not telling it that and it gave me the correct answer and so I could never get it to go back to the wrong answer.  But it is true that the law changes so often it would be difficult for ChatGPT to keep up and so I’m glad that it’s giving these answers about legal analysis.

Jim Calloway: But let’s note, it’s important that it said yet, Julie.  Can’t replace lawyers yet.

Julie A. Bays: Yet.

Sharon D. Nelson: It also uses the word currently.  I cannot currently do this legal test.  I focused entirely on the word currently when I read its answer.

Julie A. Bays: Well, what’s funny it does say in the future may play a larger role in the legal industry, but it is unlikely to fully replace human lawyers.  Instead, AI will likely be used to support lawyers and make their work more efficient and effective.

Jim Calloway: Is there research in developing AI that can understand a process ethical considerations in a similar manner to humans?

Julie A. Bays: ChatGPT’s answer on this was, yes, there is ongoing research in the field of artificial intelligence, ethics and decision-making that aims to develop AI systems that can understand and process ethical considerations in a way that is similar to humans.  This area of research is interdisciplinary and involve computer science, philosophy, psychology and law.  Some of the key topics being explored in AI ethics include moral decision-making, explain ability, accountability, bias and fairness, privacy and more.  The goal of this research is to create AI systems that are transparent, responsible and aligned with human values and ethical principles.  Well, that answer that almost sounds like an iRobot answer to me.

Jim Calloway: Thanks, Julie.  Before we move on to our next segment.  Let’s take a quick commercial break.

[Music]

Male: Built for lawyers, Nota’s cloud-based business banking is perfect for your solo or small law firm.  You want to spend your day helping clients not struggling to reconcile bank statements, Nota’s customer service specialists are here to help you.  They only support attorneys so they understand the tools you use and the requirements you’re up against and take your business as seriously as you do.  Don’t miss out on exciting new member benefits including our partnership with Law Line to earn ethics credits for your CLEs. Online at trustnota.com/legal.  Nota, banking built for law firms, like yours.  Terms and conditions may apply.

[Music]

Craig Williams: Today’s legal news is rarely a straightforward as the headlines that accompany them.  On Lawyer 2 Lawyer, we provide legal perspective you need to better understand the current events that shape our society.  Join me, Craig Williams, in a wide variety of industry experts as we break down the top stories.  Follow Lawyer 2 Lawyer on the Legal Talk Network or wherever you subscribe to podcasts.

[Music]

Sharon D. Nelson: Welcome back to the Digital Edge on the Legal Talk Network.  Today, our subject is An Interview with ChatGPT on the Future of Law Practice.  Our Guest today is Julie Bays, the practice management advisor of the Oklahoma Bar Association.

(00:09:59)

She works with the OBA Management Assistance Program to provide assistance to attorneys in using technology and other tools to efficiently manage their offices and teaches a variety of CLE classes.  Bays is also involved with the access to Justice Initiatives of the OBA such as Oklahoma Free Legal Answers.  Jim was asking about the ethical considerations.  What specific institutions or individuals are working on this new technology to help develop AI that can understand and process ethical considerations?

Julie A. Bays: Well, I asked ChatGPT this because I was curious who is working on the AI to make it more ethical or be able to understand ethics.  And so, the answer it gave was, yes, the development of this technology is being pursued by various institutions and individuals including universities, research institutions and tech companies.  Some notable examples are the Machine Ethics Lab at the University of Cambridge, the Center for Human-Compatible AI at the University of California. Berkeley, the AI Ethics Lab at the Montreal Institute for Learning Algorithms, the Ethics and Society Unit at Google AI.

The partnership on AI, a nonprofit that includes major tech companies as members.  It continues to say that there is an ongoing debate among researchers and practitioners in the field about the feasibility and desirability of creating machines that can have ethics and the specific approaches to do so, which I think is very interesting subject because there have been some studies showing some bias in some of the AI products out there.  And so, it’d be interesting to see how that research continues.

Sharon D. Nelson: The biases, you’ve put garbage in, you get garbage out the old GIGO and that has been particularly true Because if you put in all the decisions from 200 years ago, they’re going to show racial bias and we have most of the programmers for AI are young white males and they’re going to they’re going to have their own set of biases.  So, all of this is sort of going in there without anybody over seeing it because there’s all these different brands of AI and their melding them together at this point which I’m not sure what that’s going to do.

Jim Calloway: It’s quite interesting.  Let’s talk about hypothetical like some of our listeners might want to know about.  Suppose, I’m a lawyer and needed to do list for a car accident trial.  I represent the car owner who was injured and now I’m suing the insurance company.

Julie A. Bays: This is where I think AI could really help attorneys and small practitioners are specifically and small law firms.  For instance, this to-do list was a great list.  It just helps, for instance if you need a summary or you need to get your thoughts together on a new case or you have some interesting fact patterns that that AI can help you organize your thoughts by doing it for you, basically, and then, of course, it’s up to you to analyze it.  So, in this instance, what Chat came up with was the suggested to do list for a car accident trial is number one, gather all relevant documentation and that includes medical records, police report, blah, blah, and then investigate the accident, peak to witnesses, determine the defendant’s liability, calculate the damages, prepare a demand letter, negotiate with the insurance company, prepare for trial and then it goes on and list what you need to do to prepare for trial.  Try the case, making opening statements and closing statements, calling witnesses.

Anyway, I think that this is a perfect example of it and it really does give a lot of detail on what you need to do to prepare for trial.  So, let’s say that you needed a to-do list for a new project you’re starting.  You can explain what kind of project and then AI can list out a to-do list.  Do you need everything on that list? No.  Maybe you might want to add to that list, yes.  But it’s a really great starting point and it saves time.  And I think that’s where we need to learn to appreciate these new AI products that are hitting the market.

Sharon D. Nelson: Well, it certainly is a challenge even to swallow all of what’s coming at us.  I don’t know about, Julie.  But I’m exhausted from trying to keep up with this thing.  What does the AI answer when it’s asked, how should a lawyer effectively network to build their business?  Because I got to tell you, Julie, before you answer.  I got to tell you that I’ve been playing with us in our marketing and it is a marvel, what it recommends and how it has sources that I have no access to, to tell me what to do.

Julie A. Bays: Oh, I know.  I thought about this question because I was thinking, what if I — because I do give CLEs, what if I’m going to start a new CLE on a new subject matter and I just need it tweet.  So, I need a list or a way of going about it.  And so, this was — how do I effectively network to build a business.  ChatGPT came up with this in seconds and it was really a great list.  So, number one, attend legal events and conferences.  These events provide a platform to meet other professionals to make connections.

00:14:58

Number two, join legal organizations.  Joining professional organizations such as Bar Associations and Chambers of Commerce can help lawyers expand their network.  Number 3, utilize social media.  Platforms like LinkedIn, twitter and Facebook can be used to connect with other lawyers, clients and potential referral services.  Number four, volunteer and give back to the community.  This can establish a lawyer’s reputation and connect with individuals who may need legal services.  Speaking of that, Jim teaches a class where he talks about that type of marketing and how you network to build your business is give back to the community.

Number five, offer to speak in events.  Public speaking is a great way to demonstrate your expertise.  Number six, build relationships with other professionals such as accountants, financial advisors that can lead to referrals.  And number seven, maintain a high level of professionalism.  Providing excellent service to clients and conducting oneself in a professional manner will enhance lawyers’ reputation and make them more attractive to potential referral sources.  Now that is mind-blowing.  It basically took — I’m not sure if it caught it out of some book that’s on the internet or what do you think Sharon or Jim?  I mean I think that’s a good list of exactly what I would tell a lawyer on how to build their business.

Sharon D. Nelson: I think it’s — I can’t even say that word.  Say, the word for me that I want to say, Julie.  Synthe–

Julie A. Bays: Synthesizes.

Sharon D. Nelson: Thank you.  Thank you.  I knew I could do it 10 times and it would never be right.  But anyway, it does that.  It does it really, really well.  And so, I had been impressed by how many sources it seems to consult before it comes up with some of its answers.

Julie A. Bays: And the speed.  It never ceases to amaze me the speed of this type of AI.  Because it must go to more than one source at the same time.  Oh, and I’ve got a perfect example of that.  I’ll go back to that question I asked earlier.  I asked it about used car sales and if they’re defective and at first gave me that Lemon Law.  Well, I asked it to answer it again, of course, and it must have found that the somewhere on the web that Oklahoma’s Lemon Law does not apply to used cars because then it came with the right answer.  And then I could not get it to go back to the Lemon Law.  So, it was a little misstep in there but it was only seconds and so, once it does find the right answer, it seems to stick with it.

Sharon D. Nelson: That was true for me when I asked about the ethics and competence and needing to understand the technology a lawyer is using.  At first, it said that, there was nothing to command that and then somehow the next day, it is toward about legal ethics.  And now it’s cited, the legal ethics, and Rule 1.1 in technical competence.  So, amazing how fast it got to there.

Julie A. Bays: I just wonder where we will be six months from now.

Jim Calloway: Well, what really is interesting, it’s like it got every article that was ever written about this and then it just scored the answers, right? Oh, this is an 8 of the 10 articles.  So, we’ll include each one.  Let’s talk about legal research for a second.  Can ChatGPT link to online cases?

Julie A. Bays: So, I talked about this earlier.  Ed Walters has now incorporated GPT-3, not ChatGPT, into his Docket Alarm where it can breathe those federal cases and those state cases that they’ve included in Docket Alarm.  But as for ChatGPT, I don’t think it’s ready for prime time.  We have our dockets here in Oklahoma online and it’s called oscn.net and anybody can access the dockets and can also access cases.  But for some reason, Chat came up with a 404 error.  So, I guess that it must either recognize that it was a robot trying to access the cases and it wouldn’t it in or it just can’t.  So, I just thought that was pretty interesting.

Now, products, AI can apparently because the Docket Alarm people are using GPT-3 now to summarize case documents and I think that’s a wonderful way to save attorneys’ time.  And this is why attorneys, especially solos and small firm attorneys, if they do decide to embrace AI realize that this is a good way to be able to serve more clients using less time because you could do summaries of cases and narrow down things that you need to do faster which gets to what am I talking points is why they should learn alternative fee structures instead of hourly rates especially with AI now.

Jim Calloway: Before we move on to our next segment, let’s take a quick commercial break.

[Music]

Adriana Linares: Are you looking for a podcast that was created for new solos?  Then join me, Adriana Linares, each month on the New Solo Podcast.  We talk to lawyers who built their own successful practices and share their insights to help you grow yours.  You can find New Solo on the Legal Talk Network or anywhere you get your podcast.

[Music]

(00:20:00)

Jud Pierce:  Workers’ Comp Matters is a podcast dedicated to exploring the laws, the landmark cases, and the true stories that define our workers’ compensation system.  I’m Judd Pearson, together with Alan Pears, we host a different guest each month as we bring to life this diverse area of the law.  Join us on Workers’ Comp Matters on the Legal Talk Network.

[Music]

Sharon D. Nelson: Welcome back to the Digital Edge on the Legal Talk Network.  Today, our subject is An Interview with ChatGPT on the Future of Law Practice.  Our Guest today is Julie Bays, the practice management advisor of the Oklahoma Bar Association.  She works with the OBA Management Assistance Program to provide assistance to attorneys in using technology and other tools to efficiently manage their offices and teaches a variety of CLE classes.  She is also involved with the access to Justice Initiatives of the OBA such as Oklahoma Free Legal Answers.  So, at this time, Julie can ChatGPT summarize case law?

Julie A. Bays: Not that I know of.  Now they’re offering a new paid service of course but I don’t know if they’re going to offer more space but the free service could not download cases that were more than a page or two long and so in most cases are going to be more than a two page and so it is too much text for it to handle right now.  There are programs in GPT-3 that can do it but so far ChatGPT can’t do that right now.

Jim Calloway: This is a bit of an oddball question, but since all three of us have played with this, can ChatGPT be as creative with poetry as a human?

Julie A. Bays: I have to say this really quickly.  So far, it just really does not understand poetry or humor at all, but it gave the right answer which is, ChatGPT can generate poetry but it may not be as creative as humans.  While ChatGPT has been trained on a vast amount of text data and can generate poetry that is grammatically correct and follow certain styles, the model may lack the human like nuance, emotional depth and originality that often-made poetry so powerful.  Additionally, poetry is often a deeply personal and subjective form of expression and it may be difficult for a machine to truly capture the incense of the human experience and emotions.  And so far, that’s been the case.  I had to try to some jokes and the jokes were written by a six-year-old.  It seemed like it.  They were pretty bad.

Sharon D. Nelson: Well, I know, the poetry was pretty bad indeed too and I asked it to describe this podcast and that was pretty funny and also my Digital Detectives podcast and the poetry there was terrible.  So, Robert Frost and Emily Dickinson you may rest in peace, there is no challenge to you by AI at this time.  Okay.  So, thank you so much for joining us today, Julie.  It’s been a great pleasure.  I know you and I are so excited about this.  We could bore an entire room by going on and on forever, but actually people are very excited about it and you have done an amazing job with just asking questions relevant to lawyers and providing us with the answers.  And I know that a lot of lawyers are surprised by what they heard today and maybe a little nervous.  So, thank you very much for being with us.

Julie A. Bays: Oh, no problem.  Thank you for having me.

Sharon D. Nelson: That does it for this edition of the Digital Edge Lawyers and Technology. And remember, you can subscribe to all of the editions of this podcast at legaltalknetwork.com or on Apple Podcast.  And if you enjoyed our podcast, please rate us in Apple Podcast.

Jim Calloway: Thanks for joining us. Goodbye, Ms. Sharon.

Sharon D. Nelson: Happy trails, cowboy.

Outro: Thanks for listening to The Digital Edge, produced by the broadcast professionals at Legal Talk Network. Join Sharon Nelson and Jim Calloway for their next podcast covering the latest topic related to lawyers and technology. Subscribe to the RSS feed on legaltalknetwork.com or in iTunes.

The views expressed by the participants of this program are their own and do not represent the views of, nor are they endorsed by Legal Talk Network, its officers, directors, employees, agents, representatives, shareholders and subsidiaries. None of the content should be considered legal advice. As always, consult a lawyer.