A Litigation Attorney and Insure/legal/tech founder’s thoughts on ChatGPT and Litigation Management
By Dwayne Hermes
Exploring the potential impact of ChatGPT on the legal profession and litigated claims process, including predictions of productivity gains.
At the Claims and Litigation Management Alliance’s recent annual conference on March 29th to the 31st, I had the opportunity to engage in a panel discussion on the potential impact of ChatGPT in general and on the litigated claims process. In connection with doing so, I gathered some background information from three sources.
An AI (artificial intelligence) ChatBot is a tech tool allowing two-way communication. We see these in help features on websites that pop up and ask if we need help.
ChatGPT (GPT, by the way, stands for Generative Pre-Trained Transformer–if you were not scared of the potential implications of this type of tech before knowing the full name, I bet you are taking pause now.) by Open AI is an AI algorithm designed to understand and generate human language. Whether AI has evolved to the level of recognizing human emotions and social cues, as well as altering its behavior based on social awareness (Theory of the Mind), is something currently being debated. However, it is agreed that AI has not achieved self-awareness and the ability to experience emotion.
ChatGPT will lead to the loss of 3 million jobs in the U.S. and Europe. It will replace 25% of the work tasks. It will create new jobs (likely fewer than it replaces) and a productivity boom. It will increase the value of goods and services globally by 7%. Finally, and important for the audience of this thought piece, it will replace 44% of the tasks in the legal profession. (Better find those work gloves and prepare for returning to the feeling of a good day of manual labor–just kidding…sort of).
Over 1,400 tech leaders have urged AI labs to pause their development of the most advanced systems. The reason is that they believe that AI presents a “profound risk to society and humanity”. Furthermore, they posit that the creators cannot understand, predict or control the AI they are creating. Signatories included Elon Musk, Steve Wozniak, Yuval Noah Harari and Andrew Yang.
Turn it off and turn it back on!
I sit in a unique seat as a founder of an insurance defense firm and an Insure/Legal/Tech company. You might think I am more proficient than average in using tech in light of my current job. I don’t think that is the case. I have the same struggles that most of my generation have–slow to absorb what is on the screen, challenged to resolve any impediment to “fix” something myself (why can’t I remember that a lot is resolved by “turning it off and back on”), and initially opposed to adding a new piece of technology. I think my strength towards technology is that I am naturally curious and always interested in whether there is a better way. This tends to lead me to “try” new tech early and with a level of eagerness.
ChatGPT was introduced to me during a conversation at a Plug And Play conference in Sunnyvale, California, where we had been invited from a large list of applicants to demo ClaimDeck to their insurance vertical. So if you are going to embrace new tech, this is a pretty good place to do it. (Thank you, Alex Tran).
I posed to ChatGPT an issue our Team had been working to answer off and on for five years. We kept returning to it because I thought we were missing something and there was a better answer. I typed the question into ChatGPT, and it started typing the five-point answer. It was the same answer we had come up with after all the time, creativity, and various viewpoints we could apply to the question, and in the same order. OK, you have my attention ChatGPT, and glad you cannot read my emotion… or can you?
So I have another similar question to ask. One that we have worked on for years–various people, various times, unsatisfactory results, so I double down and put that question to ChatGPT. Again, the same outcome–immediately comes back with the answer, this time eight bullet points, and the same ones we had arrived at but better organized and presented.
So, what is the value of giving me an answer I already have, you may ask. Well, instead of years of coming back to the question, hours and hours of time spent on the effort and continued uncertainty of whether we had the best answer, ChatGPT gave me the answer in less than a minute (for free, although I have now sprung for the $20 a month upgrade) and gave me confidence that my answers to those two questions were correct and positioned me to explore ChatGPT’s application further and with a greater level of confidence based on this experience. I have gone on to pose new questions and have gotten great results. I suggest you do a similar exercise for yourself. I suspect you will be AMAZED, which is not something we get to feel regularly in our space. (Contact me if you want more detail on the first two questions I asked ChatGPT).
What are lawyers to do in light of the threat/ opportunity/ disruption/ modernization posed by ChatGPT?
I suspect the course of action you are comfortable taking upon ChatGPT’s arrival may depend on which of the words in the above list you pick to describe how you “feel” about it. Also, it would be influenced by whether you are an early adopter or a late follower. (Reference, Crossing the Chasm by Geoffrey Moore)
My course of action is to embrace it with all the limitations I have as to new tech.
I arrived years ago - when working on how much we at Hermes Law could process-orient the practice of law, consistent with the writings of Bill Henderson and Richard Susskind - at the conclusion that we could never do away with the need to apply “Human Judgment.” The legal work that does not require "Human Judgment" can often be standardized. By doing this, we can meet the clients’ demand for faster, better, and cheaper while at the same time improving the work life of our team members because rote/routine tasks were reduced, and more time was spent on the high-end strategy and execution part of the representation.
This principle contributed to the development of ClaimDeck and is embedded in the platform.
However, while human judgment continues to be needed in providing litigation services (a big sigh of relief by all the litigators because the amount of human judgment required is relatively minimal in several other areas of legal practice), ChatGPT and other more specific AI tools will significantly enhance human judgment.
Our judgment is limited by a limited data set of experiences, emotions, and recent events. This is not the case with AI. The available data sets for AI to assist in litigation predictions and decisions are vast and way beyond mine as a 30-plus years attorney. It has the ability to look completely objectively at past data and the current question. I can try to do this, but I will only do it partially.
So, left with my “human judgment” and AI’s superiority in many ways, what am I to do, both as a litigator and as a user of ChatGPT?
My role moves to focus on two areas: asking the right questions and implementing AI's answers. Sure, I need to check the answers, but that need will continue to dissipate. AI’s role is to answer my questions. My role in coming up with the answers will continue to narrow…and I should embrace that…because the answers will be better, and better answers lead to better outcomes, and that is what I want to deliver to my clients and have in my daily life.
We are beginning a journey that is hard to see where it will lead fully. For me, I see an analogy in the journey of Amazon in our lives. It started with one primary area, books, and then it was really left to our imagination as to how we would use it from there. And our imagination evolved gradually from books to socks, socks to food, and food to car parts. AI and our use of it in our work and lives will likely travel the same journey. It is there, ready to help us; how quickly can we imagine how that can be done and call upon it to help us?
How did ChatGPT respond when I asked: What are the suggested guidelines for a law firm's use of ChatGPT:
When using ChatGPT within a law firm, it is essential to follow certain guidelines to ensure ethical and responsible use. Here are some suggested guidelines:
Data privacy and confidentiality: Ensure that ChatGPT does not have access to sensitive client information, such as names, addresses, or other identifying details. Keep in mind that data you input into ChatGPT may be stored and used to improve the model.
Supervision and review: Always have a qualified attorney review ChatGPT-generated content before sharing it with clients or using it in any official capacity. This will help ensure that the advice provided is accurate and relevant.
Transparency: Inform clients that your firm uses AI tools like ChatGPT for certain tasks. It's important for clients to understand that the content they receive may have been generated or reviewed by AI and not exclusively by a human attorney.
Training and limitations: Train your staff on how to use ChatGPT responsibly, and ensure they understand its limitations. This includes the risk of the AI providing incorrect or irrelevant information, and the importance of verifying the advice provided.
Appropriate use: Use ChatGPT primarily for tasks like drafting and proofreading documents, legal research, or idea generation. Avoid relying on AI for more complex or sensitive legal matters that require nuanced human judgment.
Ethical considerations: Ensure that your use of ChatGPT complies with the ethical rules and guidelines governing the legal profession, including your jurisdiction's rules of professional conduct.
Cybersecurity: Implement robust cybersecurity measures to protect your systems and data, including those related to ChatGPT, from unauthorized access or data breaches.
Regular updates: Stay updated on ChatGPT's features, improvements, and potential risks. This will help you make informed decisions on how to use the AI tool in your practice effectively and responsibly.
Monitor usage: Keep track of how your firm uses ChatGPT to ensure it remains within your established guidelines. This includes regular audits and usage reports.
Feedback and continuous improvement: Provide feedback to the AI developers and collaborate with them to improve the system's effectiveness, accuracy, and relevance to the legal profession.
ClaimDeck™ eliminates claims litigation leakage for carriers while driving process into the law firm, modernizing the litigation process.
Contact Dwayne at dwayne@claim-deck.com.
Follow Dwayne Hermes, ClaimDeck, and Hermes Law on LinkedIn.