Adam Torres and Alison Grounds discuss generative AI.
Subscribe: iTunes / Spotify
Apply to be a guest on our podcast here
Show Notes:
Why is generative AI different? In this episode, Adam Torres and Alison Grounds, Managing Partner & Founder at Troutman Pepper eMerge, explore generative AI and its impact on knowledge workers in industries like the practice of law along with Alison’s upcoming keynote at The State of the Woman Conference.
About Alison Grounds
Alison is the founder and managing partner of Troutman Pepper eMerge, a wholly owned subsidiary of the firm, which provides end-to-end, integrated discovery services for legal matters. Troutman Pepper eMerge’s attorneys and technologists combine legal strategy with the latest technology to manage data, reduce costs, and identify facts needed to resolve disputes. Their novel approach effectively addresses two common causes of unanticipated costs and risks: disconnected stakeholders and processes. The technologies used regularly include artificial intelligence, technology-assisted review, continuous active learning, predictive coding, analytics, email threading, data visualization, and custom applications. Troutman Pepper eMerge also creates bespoke staffing, technological, and project management solutions to ensure clients’ goals are achieved.
About Troutman Pepper eMerge
Troutman Pepper is a national law firm with more than 1,200 attorneys strategically located in 23 U.S. cities. The firm’s litigation, transactional, and regulatory practices advise a diverse client base, from start-ups to multinational enterprises. The firm provides sophisticated legal solutions to clients’ most pressing business challenges, with depth across industry sectors, including construction, energy, financial services, health sciences, insurance, private equity, and real estate, among others.
Full Unedited Transcript
Hey, I’d like to welcome you to another episode of Mission Matters. My name is Adam Torres, and if you’d like to apply to be a guest in the show, just head on over to missionmatters. com and click on be our guest to apply. All right. So today my guest is Allison Grounds, and she’s managing partner and founder at Troutman Pepper Emerge.
Allison, welcome to the show. Well, thanks, Adam. It’s a pleasure to be here. All right, Allison. So excited about today’s topic. So we’re going to talk about the impact of generative AI on knowledge workers. And of course, I definitely want to get into the state of the woman conference, which I understand that you’re going to be a, a keynote speaker at.
So I guess just to get us kicked off. Hey, are you excited for the conference coming up? I’m very excited about the conference. It’s in a beautiful Nashville and the speaker line up and attendee list is a great diverse set of women from wonderful backgrounds. I’m really looking forward to meeting some of the ones I have not met before and learning lots of fun things from the rest of the speakers.
Wonderful, and we don’t have to give away too many spoilers today, but maybe give us, you know, some high level of what you’ll be presenting as a keynote. Well, you know, funny, you should mention it. We’re going to be talking about a lot of the same things that you are talking about today, which is the, the impact of generative AI on, in particular, the legal field, which is the space that I occupy how women leaders and executives can have a seat at the table as this shift in how we.
Work and interact with 1 another and machines is upon us. So looking forward to that dialogue and hearing other people’s experiences and other industries, but definitely focused on this game changing technology that we’re seeing already play out. Certainly in the legal space. I mean, we’re hearing a lot about generative AI in the news right now.
I mean, AI in general, of course, but, you know, sometimes it can be confusing. I mean, why is generative AI different? Yeah, that’s a great question. And I think, you know, at our law firm, one of the things that we. Decided to do whenever the headlines started to shift and so much attention was being given to the chat GPT in particular, which you know, had its big launch and had a 100M users adopted in record time at the end of last year at the end of 2022 and the beginning of 2023.
So we thought, you know, this is interesting because people have been impacted by AI and using AI for 2020. Decades and just how do you even find define artificial intelligence, right? It’s helping suggest movies. You might like, or clothes you might want to wear songs. You might want to listen to and we’ve used it in our businesses to improve and automate certain tasks.
And analyze information and improve logistics, but generally, I think the reason that it’s got so much attention and still is right. I mean, probably by the time people hear this podcast, they are tired of hearing about it. It’s gotten so much attention. But my philosophy on why that is, is because it’s something that is.
Accessible to everyone chat is something anyone with a computer and Internet connection could download and interact with, and the rollout and the rapid adoption because of its popularity, it being incorporated into other social media platforms, Internet search engines, it really just had a quick impact because it was so accessible and the thing about generative AI, which is, Just another large language model natural language processing is that it is meant to mimic the way humans communicate.
So being able to ask questions and receive answers and fine tune that that sort of human like interaction, I think, is what drew so much attention and surprise people. People were used to having to interact with the I. T. department or someone like me in a law department as sort of a gateway to technology and meeting that learned intermediary, I will say, I think they still do, but you can certainly directly access chat and other AI language models and be able to instantly see what it can do.
Hmm. How do you see this like further impacting, especially as it gets better. let’s put our, you know, crystal ball out there, or whatever, however you wanna say that. , like how do you see it? I impacting knowledge workers and even like industries like law, like how do you see that going forward?
Yeah, it’s interesting, you know, golden, golden sacks, I think was 1 of the 1st to kind of put out some research trying to figure out what that impact would be as of AI in particular. And certainly the legal space was identified as 1 of the top industries likely to have an impact. And I think they assumed something like 44 percent of legal work could be automated using AI.
And so, yeah, the, the law firm, we, we sat and looked at this explosion and thought, well, let’s. Let’s harness our existing skill set. We’ve got lawyers that are already using forms of AI in their daily lives, which would include my practice, which is focused on e discovery and information management, as well as other practitioners in the firm handling things like privacy issues, data security IP, and then we have a lot of business professionals.
We have an entire innovation team already dedicated to looking at custom solutions that would help us to do our work better and faster and more efficiently. And so we said, we need to take a look at this potential impact. And are we taking advantage of. All of our existing expertise and experience and harnessing that so that we can have a seat at the table and be able to leverage this next iteration of AI.
And the way we see it happening is, you know. What we did internally to law firm last year has been replicated by a lot of our clients and other law firms, which is really looking into what’s the skill set on the human side? Who are the people that should be at the table to evaluate this technology and make sure it’s deployed safely and securely and accurately and where can it be used?
And you mentioned sort of Where is it? And as it improves, what do we see it doing? I was just on a call with some colleagues where we were talking about, you know, continuing our effort and our task force initiatives to educate our lawyers on how to use the tools we have. And 1 of the things that kind of came to the forefront was just having people access and use these tools.
And we developed early on a product called Athena, which is an internal. Chat GPT like large language model that’s in our own private Azure environment. So our innovation team built that out and they had that actually in the works before chat GPT was launched. So luckily we were able to launch that pretty quickly and it was interesting to see how people were using it.
I mean, it’s, sort of a general application tool. So initially people were asking it, Yeah. Questions almost trivia like questions, and they were feeling frustrated that was getting them wrong because it didn’t they didn’t realize that the data used to train these large language models was, you know, a couple of years old.
And so we had to do a lot around educating people about. What tools like Athena or chat GPT could or couldn’t be used for and we made the call of of not allowing the use of publicly available chat GPT type tools because of the risk of, you know, we have a lot of ethical obligations to keep our clients information confidential and secure.
So, having an internal version where our data did not get used to train the model and our information speed. Confidential was really key for us, but it was fun to see the evolution of use of that. So, because we started early and had some early adopters, and we could see what they were using it for and kind of reach back out and give them more guidance.
It’s been nice to see lawyers using it to help stimulate ideas. I know in 1 example, I just heard about a partner in our IP practice, you know, taking a brief that she received from an adversary. On a motion and ingesting it into Athena and asking Athena to summarize the theme. She was very familiar with the case and had worked on it, but, you know, you’re in the weeds and you’re, you’re kind of, you start to develop your own bias about your case and your facts and having sort of a neutral model extract the themes from a brief to make sure that you’ve addressed all of them.
And to see how it might read to a judge or someone who’s more neutral with something was just an interesting strategy. So, and I think our marketing department and other administrative departments were some of the 1st early adopters for kind of safe uses, like press releases. I used it to help draft emails, asking people to submit annual reviews to my team.
innovation team used it and built out a skill for our partners retreat, where they trained it on a specific subset of documents. So the hotel information, the schedule, the agenda to be able to get people just get their feet wet. And, you know, how do you interact with this tool in a way that it can maybe shorten the time?
You know, we live in a billable world, which we could talk about hopefully is changing where. 6 minute increments are how we trap our lives. So if a lawyer can figure out the answer to a question in 30 seconds instead of 30 minutes, that’s revenue for the firm. So I’m really trying to figure out how to reduce the time spent on administrative tasks are looking for information.
I mean, we generate content and information for a living, which makes it. Generative AI, both a threat to what we do as well as a key solution to what we do. So being able to leverage these tools to minimize the time we waste on less valuable tasks and drafting exercises and really focus on the things that generate value to a client is something that’s it’s very exciting to me and to most of us in the practice.
Yeah, I find it. That was a really long answer for you, Adam. That was the longest answer to a question you’ve ever received in a podcast. Well, because the thing is, is that many people see it as or at least some of the arguments against it, right? I’ve seen it as kind of like limiting or decreasing you know, or maybe, you know, replacing jobs or this or that.
But if you think about like just progress in this country, worldwide, whatever, not just this country, but in terms of, you know, once upon a time, there was a room full of individuals that were sitting in front of a typewriter. Right? Like once upon a time there was, I mean, if we go further and further back, like, but think about all the innovation and all the things that, you know, those now individuals that might’ve been doing a, a rote routine for the sake of, you know, playing their role.
And that’s what that time, you know, called for now that that person has the opportunity for higher level thinking and to express themselves in the workplace to a higher calling than just following, you know, the. Rote routine of again, using my example, which might not be the best example of when there was a room full of individuals, you know, in front of a typewriter.
Right? Oh, absolutely. And I think that’s definitely true. What you’re seeing in the research and Goldman Sachs and others is that certainly there will be a large number of roles who will have. Percentages of their daily job responsibilities automated through various forms of AI. But to me, it does open up opportunities.
I mean, 1 of the biggest things you hear about the risk of bias and discrimination and hallucinations in the data set. So really, especially as lawyers who are sort of. Train to be logical, analytical thinkers and those of us who’ve already been sort of dealing with the legal implications of data, whether that’s through privacy or data security or discovery kind of have already been thinking along those lines.
We’ve already been using technological solutions to be able to do work that was frankly cost prohibitive to do without that technology. So my in my space, we’re doing the collection and analysis of. Documents and data for litigation or government investigations, that’s electronic discovery or discovery.
I’m old enough that whenever I started doing litigation back in the early 2000, there was still some focus on paper documents and you go to the client’s office and scan a bunch of paper in. And I also was just. I was thrilled when I saw that we could use databases and analytics to help collect data electronically.
I remember the first case I worked on. I remember we received the production of electronic documents on a CD through the young members of the audience. This is a physical medium, something that had been. Yeah, why would you do that? Go ahead. Tell us more. Yeah, so, so because the people dealing with the data we’re used to dealing with in banker’s boxes, right?
Flipping through page to review documents, the, the older partners on this case just sent these compact disks away and had them give us boxes. So then I have this sitting in a room, you know, buried in giant piles of boxes, which we then reviewed. And labeled and scanning back in. I mean, it was totally inefficient.
And so, and we didn’t do that intentionally. It was just what people were comfortable with. They were used to working with paper. They were used to reviewing documents and paper. So we very quickly adapted it and said, you know, why are we downgrading something that was digital into paper and then converting it back?
I mean, it was a very complicated process. And as more of the information we generated was generated electronically, it made sense. That you would then collect it electronically, process it, put it in a database and analyze it. And that technology has used some form of AI and artificial intelligence and technology assisted review for some time it has the same impact that you’re starting to hear now.
I remember. When I started our practice and, and asked that we bring the technology in house and have specialists that know how to use it, analyze data, they were like, well, what are the associates going to do? That’s what they always did was document review. Well, there’s still plenty of documents to review.
If you think about your examples, a great 1, if content used to require, you know, physically typing a hammer, you know, for the typewriter, the paper and then pulling that it’s a lot. Okay. Faster to generate content and we’re generating more than ever. So you need technology to even digest what’s happening.
I had a case where the partner was like, alright, Allison, you’re assigned this particular witness. So I need you to go to that fancy database viewers and read all the emails that this person sent in a 10 year period. And I was like, so let me let me explain why that used to be possible for you to read the quote unquote file of a witness.
And everything that was in their office, but if I were to read, I mean, I had to kind of break it down. Like, think about how many emails you receive a day, even if we’re filtering those with certain terms. Right? It would take me 5 years to read 10 years worth of filtered email. So you’ve got to use some sort of technology to get to the things that really matter.
So the way I view it is. Yes, associates still review documents, but they review more important documents and the technology helps to eliminate the relevant information, prioritize the review and we’ve seen the generative AI solutions and document review. To me, it’s been some of the most exciting and fastest to be sort of in the call.
I mentioned earlier where there’s certain generative. I pulled it. Quote unquote aren’t there yet, right? Like, yes, you can have it summarize a document and you can help it prioritize your outline, but it’s not necessarily being trained on a subset of data. And we have examples and test cases where we’re doing this, where the output is super reliable on its own, but we’re working with it in the discovery space using generative AI to train.
And feed the information to the machine to say, here’s the information you should be looking for in this document set in that kind of dialogue Q and a type way has been very impressive and amazingly accurate. Pretty early on. It’s still expensive because those tools, the back end engines and the computing power necessary to do that work certainly not free or easy.
And you have to. You have to understand a little bit about engineering, but the tools are really getting there pretty rapidly and we’re excited by the results we’re seeing. And I don’t know, we, I guess, as a lawyer who works on the technology space, it would seem strange for me to be scared of technology.
Like, I’d rather. Find ways to do more and a more accurate way, and we’re arbitrarily limiting what we’re analyzing because of the limitations of time and space and the human brain. So, I think being able to have that sort of. Amped up version of a, I to interact more. You know, Efficiently with the documents, figure out what’s relevant, build out chronology, key events that have occurred.
We can also save a lot of wasted time. I mean, I know I’m not going to shock you Adam when I say people don’t love spending money on lawyers, but they don’t and they enjoy spending money on or a new product. I know. So, but we can limit that. Right? Like, I don’t want you to spend money on me if you don’t need to.
So, if I know. Yeah. Early on, by analyzing your data, I can find the key information I need to help you settle the case before you spend 2 years fighting about it. Then that’s a much better use of my time and your time and our knowledge of your data and the case and our judgment and strategic legal thinking to know what to look for in the data.
Right? It’s really pretty essential to what we do. So, yeah. I’m excited about how it can reduce time spent on wasted analysis to really get to the heart of what matters. And even though we have a billable our model, which is makes a lot of this challenging. And this is kind of the topic. I wanted to make sure we covered is because we have built around charging for our time and 6 minute increments.
The legal industry as a whole is often reluctant to adopt to things that maybe make things more efficient, not because they’re trying to make things more expensive than they need to be, but because they just they’re used to a time model. And so we’ve always thought that we would have more unit based pricing or value pricing or deliverable based pricing.
But I think I haven’t seen that happen as quickly. At least with prior forms of technological advancements, because the measure of whether or not that pricing is fair or accurate is still the hours based alternative. So, if the question is, I’m going to charge you 10, 000 dollars to do this legal task.
And the answer is, well, but also tell me how many hours it took and what your billable rate is. So I can see if that’s a good value. We’re never going to get away from that. And so that’s the, I feel like that’s the. A barrier we’ve seen in the, in the legal space is how to figure out what is the right value or price for a particular legal deliverable.
And as long as that measure for value is, but how long did it take you? That’s going to discourage. The investment in technology and efficiencies, right? So interested in how we, how we account for that and how we price it. And we’ve got our task force working on it and thinking about it. And and how do we convey the value of what the clients are receiving?
How do we make sure we also get our investment back? We’re, we’re investing a lot in these technology solutions and in the people and processes to build around them to make sure we’re deploying the right solution for a particular need. To get the right results. So it’s it’s an exciting time as you can tell, because I can’t stop talking so fast about it.
But I like being here at the time. I’m here, which is having seen the technology steadily advance having been here during a more analog period and then seeing how we adapt to a fundamental shift in that technology. I think that’s a great way to end it, Allison. Allison, if somebody is listening to this or watching this and they want to follow up, continue the conversation, learn more about your work or follow your journey, I mean, how do people do that?
Well, they can certainly check out our website TrautmanPepper. com. That is our law firm’s website. And let me actually make sure I’m not lying to you because I don’t ever really just. Type in our website like that. It might actually be trotman. com and we’ve got a whole dedicated page on our website focused on our innovation efforts within the firm.
It’s even easier. It’s just trotman. com. So we’ve got a page where we’re tracking the legal developments in this space as well as what we’re seeing in terms of what our clients are doing to apply. Generative AI and other AI solutions in their own industries and then it’s really a multidisciplinary team coming together on that.
So there we have several different not as good as yours. Adam. Your podcast is the best. We certainly have podcasts focused on the impact of generative AI and different industries, health and life sciences, the financial industry, other clients that we serve and industries that we serve in addition to the, to the legal.
Practice itself, so we are excited to have folks check that out and to reach out. You can also find my contact information to trauma dot com and reach out. I’d love to hear for you. Fantastic. and we’re big fans of the podcast community in general. So I want everybody to go check that stuff out for sure.
No, I seriously, I love the podcasting community and all the people that I’ve had the privilege of working with. I mean, love podcasters out there and for everybody listening, by the way. We’ll put all that information in the show notes so that you can just click on the links and head right on over to Troutman.
com. I just checked it out and then Allison was obvious. It’s very easy to find on LinkedIn as well. So feel free to check any of that out. And if this is your first time listening to Mission Matters and you haven’t hit it, That subscribe button yet. I don’t know why. I mean, here goes your invitation, hit the subscribe button and if you’re a long-term listener then leave us that review.
If you sure haven’t done it yet, we would appreciate it. And Allison, I’m looking forward to your, your appearance and your keynote over at State of the Woman Conference and all the great work you’ll do over there just like you did today. Appreciate you bringing us that content to, to our audience.
Thank you. Well, thanks Adam. I appreciate the opportunity.