A Single Source of Truth for Strategic Planning and Execution
ON-DEMAND WEBINAR
Move to Create On Demand Landing Page
If you are a Forte or Innova Customer watch our webinar delivered by Litera experts to find out more about our new product Litera Create.
Watch NowMandy Neske, Sales Director for Litera’s Objective Manager offering, led a Changing Lawyer Summit panel that provided insight into the ways that technology is enabling more robust and well-executed strategic planning processes in law firms. Read transcript
Featuring
Mandy Neske
Sales Director for Litera’s Objective Manager offering
Amanda Bakken
Director of Business Development at Stinson LLP
Richard Punt
Managing Director, Professional Markets
Paul Devitt
Co-founder, NextLevel Leadership Coaching
James Lawson
VP Customer Success, Litera
Welcome to a special edition of LEGALTECH MATTERS devoted to important topics from Litera’s The Changing Lawyer research and report. In a series of special podcasts, we'll speak with industry experts with insights on the key takeaways from the report.
David Curle
Welcome to the latest Changing Lawyer podcast. This is a series of podcasts based on Litera’s Changing Lawyer report. I'm David Curle, Legal Content Research Lead at Litera. And my guest today is Casey Flaherty, co-founder and chief strategy officer at LexFusion. Welcome to the podcast, Casey. It's great to have you here.
00:00:39:01 - 00:00:39:24
Casey Flaherty
Thank you for having me.
00:00:40:06 - 00:01:03:03
David Curle
I'll give Casey a chance at a little bit to tell us more about LexFusion, which is a very interesting company. But I also wanted to point out that he's a prolific writer and a respected analyst when it comes to tracking legal industry changes. And really the origin of today's podcast is a report I'm currently working on, which is kind of a retrospective look at some of the findings and data from our Changing Lawyer research going back to 2017, we're taking a fresh look at what it means to be a Changing Lawyer.
But what I'm finding as I look back on the issues and challenges in those earlier reports is that we may not really have changed very much. Casey was one of the industry experts we quoted extensively in that first report from six years ago, So I thought I'd bring him on the podcast to get his perspective on some of the intervening years and what's happened.
Casey and his partners at LexFusion have also recently published a very well-received article at Legal Evolution, which was their second annual legal market in review, which I highly recommend. And the analysis is based on hundreds of conversations that LexFusion has been having with law firms and with corporate legal departments, and it provides a very frank and unvarnished view of the state of the market today, which might not be all that different from what it was in 2017.
We'll see what Casey has to say about that. So, Casey, you know, LexFusion has a very new and interesting model that I think deserves a little explanation. Can you give our listeners some background on yourself, but also about how LexFusion came about and how you operate in the industry?
00:02:13:07 - 00:02:38:19
Casey Flaherty
Sure. As for myself, I'm still a bar lawyer since I just paid my California dues before February 1, I started as a complex commercial litigator at a large firm. I then moved in-house while in-house, I started writing and speaking on the application of technology and process to legal service delivery. This got me involved in the that they sent legal ops movement.
I was a member of CLOC when we still called ourselves a book club and would meet in conference rooms at Google and Oracle. That led to even more writing and speaking. And ultimately I struck out on my own as a legal tech founder, and I supported that business by consulting primarily for law departments, but also several law firms.
This is legal operations consulting. So, process reengineering, tech selection and implementation, outside counsel management, etc. One of the law firms that I worked with was Baker McKenzie. They brought me in to build out the world's largest legal project management team, which I did for a couple of years. It was a great experience. I left behind over 60 legal project managers were fully utilized, but was also happy to return to the ecosystem as a co-founder of LexFusion, which is very much an ecosystem play and also defies an elevator pitch.
We keep trying different ones and none of them quite capture it. We exist to accelerate legal innovation. And the reason that the elevator pitch fails is because we're experienced differently by customers than we are by our actual clients. That is the people who pay us. So, we are paid by vendors. We consult to them, we work for them.
We're out in the market on their behalf. And we never we never hide that. We've curated a collective of vendors. We have a portfolio we're very proud of. We choose them as opposed to them choosing us. And we're certainly happy to talk about any of the vendors that we work with. But when we go in to speak to customers that's not our angle.
Our ability to consult to the people who pay us the vendors is predicated on our ability to talk to the buyers of their services that the legal market in your review comes from discussion with 450 law departments and 250 law firms, and we do not go in with I'm going to try and sell you something. Instead, it's what are you working on? What problems are you trying to solve?
And so, we have many conversations where none of the vendors we work with even come up because they don't have problem solution fit. We're there genuinely to help them to add value because adding value is the only way that we have repeat interactions. And so some conversations, none of our members come up, some conversations, only one of them come up.
It all depends on and not every person we talk to has to ultimately be a customer of directly of one of our vendors. They're certainly never a customer of us. We don't charge a lot apartments or law firms. For us, every node adds value to the network, right? So we are forever talking to law departments and law firms to get a composite picture of what's going on in the market, because having that knowledge is valuable to us, but it's also valuable to them, that is that we can bring that composite to bear in our discussions with the individuals.
So one of the reasons that everybody talks to us is everybody talks to us, and often one of the greatest services we provide is connecting peers with each other because they're working the same problem. It's not a very direct model and it defies direct description, but it enables us to be out in the market having conversations that are valuable to us and the people who pay us, but are also valuable to the people to whom we're speaking.
In fact, we will very oftentimes end up recommending vendors that are not in our portfolio, not because we dislike them, just because they're not a great fit or they don't have broad enough coverage or any of other dozen reasons. But they do have problem solution fit for the problem that's being presented to us, that they address something that none of our members do because no matter how much M&A Litera does, you're never going to cover every single area in the legal ecosystem, nor would you want to.
We have a vibrant ecosystem because there's so many different problems to solve and so many people trying to solve it different ways.
00:06:53:23 - 00:07:14:08
David Curle
Well, and I think that's what makes your analysis in the recent report so interesting is that you do have clearly the same depth of conversations with the firms as you're having with the clients, and it's bringing both sides to bear. And really, that's one of the things I want to dig into a little further out. But I want to look back at that first report that I referenced in 2017.
There was a chapter called Cheaper, Faster, Better about client frustrations with the service that they were receiving from law firms. And I'm just going to rattle off a few findings from that report. There were 68% of the firms we talked to believed they're losing business to corporate law departments. In other words, bringing work in-house. 33% of the firms didn't see regular input about what clients want and value.
So there was this sort of tension in the market that clients were expressing that they wanted more in the way of technology and innovation from the law firms. But there was a lack of a consensus or understanding about what that meant. You know, some of the other issues that popped up in that report were the lack of just basic technology skills, which is something near and dear to your heart.
You know that lack of ability to just use Microsoft Office applications to their fullest. Another one was that firms are just not really solution oriented. They think of legal work as a series of isolated matters rather than a system or a process that clients had needs for longer term solutions and that firms and lawyers just didn't have time to pursue that kind of solution because of the time constraints and the pressures of their of their day jobs.
And then finally, there was a bit in the report about the sort of the structure of law firms, the logjam at the top where decisions about long term investments and changes are being made by senior leaders who'd have to pay for them out of their pockets, but maybe not reap the benefits of them in the future. So, I guess my first question, Casey, is do you see the industry meeting any of those challenges any better today than in 2017 now?
00:08:55:22 - 00:08:56:16
Casey Flaherty
No. Next question.
00:08:58:04 - 00:09:06:10
David Curle
Well, you know, I was going to ask, well, what has changed and all that? I guess so. I guess if is it a categorical no or have there been certain things changed?
00:09:07:00 - 00:09:42:00
Casey Flaherty
Oh, there are always pockets, there's always examples. There’re always outliers. There are certainly things that are better than they were. And there are certainly specific firms that are doing specific things that get appropriate plaudits. Same with in-house departments. They're definitely department level innovations and efforts that are praiseworthy, meritorious and have excellent ROI. But overall, we have been locked in counterproductive pattern for a very long time and have yet to escape it.
And I don't think it starts with law firms. I think it starts with clients. And in fact, it doesn't even start with law departments. It starts with the businesses that law departments serve.
00:09:54:21 - 00:10:27:24
David Curle
Right. And that's the interesting thing about the perspective of I think your legal marketing review piece is that it's sort of that's the source of everything, right? Is the is the corporations that are driving this industry with their needs and getting what they need or not spending on how it's delivered. Let's maybe dive into that piece. You know, I think one of the things I took out was that you do see when you're talking to law departments on the corporate side, a real sense of urgency or pain around the issues, that there's a lot of pressure to cut costs and get more for less and all that stuff.
But there's a tension between that and sort of the longer-term solution to the problem. It seems to be there's this tension between sort of the short term just get the work done versus rethinking how to get the work done. And the section of your piece that I think was most interesting was where you started talking about spend optimization and scale.
Those are sort of industrial terms that legal work isn't very well scaled today, it's labor intensive. And that that seems to be the crux of the whole problem. Is that a good statement of what you're trying to say there?
00:11:00:10 - 00:11:24:20
Casey Flaherty
Yes. Legal work suffers from what's known as Baumol cost disease, which traditional theory is that the less productive you are, the less money you will make because others are becoming more productive. And that might be true within a sector, but not across sectors. In fact, the more labor intensive a sector is, the more it costs to deal with that sector.
And that doesn't mean everyone in it gets rich. But schools are expensive, medicine is expensive, concerts are expensive, legal is expensive because they are in many ways labor centric. They're built. It's not always 1 to 1, but it might be 1 to 30 or it might be 1 to 3,000, or it might be 1 to 30,000 in the case of concerts.
But it's not infinitely scalable. You know, a live music performance can be transmitted all over the world, but it's not live in the same sense of being in the concert hall is and since the limited capacity there, the ability to scale again and then the size of the venue is similar, the more intimate the venue, the more the tickets costs, You know, yes, you could have recorded lessons and flipped the classroom, but the actual classroom, we found that the most impactful interventions are 1 to 1 tutoring. Doctors can be augmented by all kinds of technology, but it's still meeting with the doctor or the surgeon. There are limits to scale, and law fits in what is called the stagnant sector that is the areas that see the least amount of productivity gains from the advancement of technology. Some of that is inherent in legal work, say, showing up in court, but not all of it.
And we have found ourselves incapable of decoupling labor from the delivery of legal services such that as legal needs go up, there is a proportionate increase in the need for legal labor.
00:13:05:07 - 00:13:29:12
David Curle
I was going to say that all that's happening at the same time that as you point out in the piece, there's this incredible increase in the complexity of business today, both most obviously as it relates to legal, just the regulatory environment, but other things as well, other sort of social issues are playing into it and just the increasing reliance of most industries on various kinds of complex technologies and processes.
That's what's driving the biggest tension, right, is that the more complex things get, the more difficult is for the legal to operate in an environment where they can't operate at scale.
00:13:39:23 - 00:14:09:14
Casey Flaherty
So the operating environment for business is becoming more complex. The more complex that operating environment becomes, the more valuable quality legal guidance becomes to the enterprise. That is in direct tension with the desire to slash resources spent on legal, at the very least as a percentage of revenue. But even companies that are stagnant, both in terms of their revenue growth and their spend on legal, are operating in a more complex environment.
It's exogenous. It's outside the organization. Those new regulations are being put in place regardless of whether you are in cost cutting mode. And so to the extent legal resources are being kept flat and the relationship between business needs and legal labor remains static, then all we will see is a delta between the legal needs of the business and the ability of the law department to meet those needs.
And so, they will accumulate operational risk and eventually bad things happen and bad things are almost inevitable. But their frequency and their severity can be impacted by the ability of the law department or the quasi-law department. So we would call compliance or tax or regulatory affairs to meet the needs of business at scale and pace. And again, fundamentally, we are still labor centric in how services are delivered.
And so that Delta only grows, as does the danger, as does the negative consequences.
00:15:13:14 - 00:15:30:06
David Curle
So yeah, that Delta is growing. It seems to me that that's pretty sort of clearly unsustainable. What's going to be the breaking point or what's going to be the adjustment that will finally address that Delta? Where do you think the first sort of expressions of doing legal work at scale are going to come from?
00:15:30:15 - 00:16:00:03
Casey Flaherty
I suspect that most law departments will become increasingly irrelevant to their organizations. They'll still exist, but corporations are going to set up rival teams and rival silos. And over time have those more oriented towards scale. So, you know, the regulatory team that stood up for the first time is likely going to have more of a scale orientation than the law department, to which they're being set up as an alternative.
Right. And so over time and corporations do this and often to their own detriment because they're adding layers, they're adding complicatedness as a response to external complexity. But those new layers will disintermediate the law department from a lot of areas that are legal in nature, and a lot of law apartments will become less and less relevant in the operations of their businesses.
I probably the most extreme example and while I don't think it will happen tomorrow, I do think it will happen in the next few years is when consultants, you know, insert big name here McKinsey, Accenture, Bain, Deloitte, whomever is talking to the C-suite about digital transformation, especially as powered by these new large language models and transfer arm based neural nets, etc...
And they won't be talking about the lot apartment because the lot apartment isn't big enough for that kind of conversation. Instead, they'll be talking about business-critical processes like contracts and how much the business can improve quality and velocity while reducing labor costs by removing low value added labor from the process. Because it's now that much easier to automate.
And when they go through that exercise, a lot of the touch points that will be cut out will be legal again. Now, not with legal being the target, the business processes and the business outcomes that are the target, right. But that low level labor in a lot of cases currently resides in the law department and no law departments ever want to hear that because what we're calling low value add doesn't mean that it's unskilled and doesn't mean that these people aren't smart and well trained and working hard and actually adding value.
It just so happens that it is something that can be graduated up the chain. One of my favorite authors is Roger Martin. His books were introduced to me by The Amazing Jane, and among them is a book, The Design of Business, and he discusses how we evolve through kind of three stages of getting work done.
One is mystery. We have no idea how to do something and we just need to figure it out. And there is an enormous amount of experimentation involved there. The next is heuristic. We develop some wisdom and kind of shortcuts and informal rules on how to approach whatever it is. And then the ultimate one is algorithm. It is a well-defined process that doesn't necessarily run automatically, but runs very smoothly and is optimal or can be optimized over time.
Reminds me of one of my favorite quotes from Whitehead, which is civilization advances with the number of functions that can be performed without thinking and a huge focus of what legal work should be is compliance by design that is moving upstream into the business and embedding legal knowledge into business processes. So you can get them to the point of algorithm that is, you delawyer them without delegaling them so that you are getting higher quality, more consistent outcomes at pace.
Yeah, but in his newest book, Roger Martin observes that knowledge work in general is in no way focused on lawyers. It's actually more focused on managers, marketers, salespeople that we make this mistake in that we try to organize his knowledge work the same way we organize physical work, which is people in flat fixed jobs who do basically the same thing over and over.
And we may assemble them into teams and those teams come together to produce things. Knowledge workers, oftentimes those things are decisions and that is that is what you do. But from the knowledge worker perspective, to graduate from your heuristic to algorithm is therefore a threat and you do everything you can to avoid that. Thus, the proper approach isn't to organize knowledge work like physical work, it's to organize knowledge work around projects that the thing to be produced is not decision after decision after decision, but knowledge.
And then that knowledge as part of the project is embedded into the business, into those other algorithms that are ongoing. And then you move on to the next project. And that project orientation ultimately is what will enable you to meet the needs of business at scale. But we do not have that project orientation in law departments, mainly because that's not what lot apartments were originally designed and asked to do.
It's unfair to just put this on law departments, businesses start law departments as a less expensive, more accessible alternative to law firms. Come in and do the work that we would have paid outside lawyers to do and also help us spend less with outside lawyers by being a sophisticated consumer of their services. And that is the original premise on which law departments are founded and they are fulfilling.
And so, we have had massive insourcing. I don't know who the 32% of firms who didn't think they were losing work to clients are. With very few exceptions, they're probably quite wrong. The story in legal since the middle of the 1990s has been the explosion of in-house counsel. In-house departments have run at seven times the rate of law firms.
They're now more in-house counsel than their lawyers in the domestic offices of the AmLaw200. Not only that, but corporations with one or more in-house counsel account for 52% of the purchases of all legal services in the US. That is, in-house lawyers are the number one segment of suppliers of corporate legal services, and they are the number one buyers of all legal services, period.
And so, there's been an absolute explosion because insourcing does in fact work in terms of being less expensive. It is less expensive. It's just not sustainable over the long term if you're trying to meet the evolving needs of a business.
00:22:28:21 - 00:23:00:04
David Curle
One of the other aspects that you address in the pieces as well as has to do with sourcing? And you just mentioned, you know, what types of providers out there. One thing I wanted to drill into a little bit is what you think of the development of LASPs looks like recently, even back at the time of the 2017 report, a lot of people believed that lots of work would be shifting to alternative providers and I know TR Thomson Reuters just published its annual or semiannual report on alternative providers just this week, I think, and it shows an upswing. Is that a significant and an appropriate solution to at least part of the problem faced by the corporate side is just finding a different source provider for certain types of services.
00:23:11:19 - 00:23:43:15
Casey Flaherty
Yes and no. Okay. I know this is going to shock you, but I have a long, long, long answer on this subject. So not only has the growth of law companies, LASPs or whatever managed services, whatever you want to call it, not only has there been substantial growth and that growth has outpaced the growth of the AMLaw 100 is on a percentage basis that TR goes back to 2015, so they trace it 2015 to 2020.
I compared their statistics to the AMLaw100 substantially higher growth percentage, but the 20 billion that they found being spent with and we use their term now, LASPs is an understatement. And the reason it's an understatement is because there is a substitution effect. The great Bruce McEwan, Adam Smith, Esquire once estimated that for every $1 that moves in a house big law loses three, and that the number for LASPs is likely higher conservatively, let's say that's four that would be $80 billion.
Last year, the AMLaw100 came in at $127 billion. So that's a lot of work. But by volume that, you know, back in the old, old days, pre and post revolution would have gone to law firms. But in fact, if you treat 2007 as the inflection point and look at the pre 2007 period which goes back to 1993 to the post 2007 period, that goes to 2021, the AMLaw100 is 54% off revenue trend.
That is if you would simply extrapolate it out that 93 to 2007 growth rate to 2021 there 54% of that revenue trend. And so, yes, I think LASPs are a substantial part of the future and the present. I think how substantial it is understated and I also think they're underutilized. That said, going back to being a sophisticated consumer, it is one thing to use ALSPs as simply as a source of labor arbitrage.
You know, in-house lawyers are less expensive and outside counsel and ALSPs personnel are less expensive than in-house personnel. And there's nothing fundamentally wrong with labor arbitrage, right? Sourcing is absolutely part of this overall puzzle. But it's still the case that that's fundamentally about labor. And so, you still have to ask questions. How is that labor being augmented through process and technology?
And for much of the work that we're sending, how are we moving the work up the chain from pure labor to pure tech? I'm not that all of it will get there, but do we have projects in place again to progress our ability to deliver at scale, or have we just simply made a labor arbitrage play? That's a one-time lift, but we are going to need to add people in the low cost center, whether we own it or someone else does as volumes go up.
And so, you don't get away from this question of how do we change the way that legal services are delivered just because you found lower cost labor? And in fact, I think even in the ALSP space, finding maximum savings on day one is short term understandable and a long term mistake because you want to have a relationship in place where there can be investments in changing the way that work is done and that you want the relationship to be significant enough that it's worth your time as an in-house department to actually pay attention to that partnership and to make it a strategic partnership.
The ultimate goal, I believe, and I've been convinced of this by my good friend at Stone and Factor, who I think you may have known in a past life, is what's now being called integrated law, where the managed service provider, it's more than just a managed service. It's really and part of the value chain that's in close proximity to business.
People who work there can have the domain emails. So that people in the business don't know, right. When they're dealing with someone who works in a lot apartment versus someone who works at the integrated law provider that there's very close partnership in thinking through the systems, playbooks, process, the tech, etc. and that that is a great way to deal with the high volume, but also what is a very real allergy to headcount that oftentimes it is easier for law departments to spend money externally than it is internally on headcount, no matter how compelling the savings math is, because eventually you get large enough that you're asked, well, you already have so many people, why can't you make do with them? So, I think this integrated law concept is only going to become more prominent and more important. But it can't just be labor arbitrage.
00:28:31:20 - 00:28:42:15
David Curle
Yeah, I mean, you're talking about a world where the labor arbitrage is still there, the scale problem still exists, but at least you're shifting, right? Sourcing shifting the work to different places. There's advantages there.
00:28:42:20 - 00:29:11:15
Casey Flaherty
And also, and this is one of the reasons that I consider the insourcing push to be counterproductive is precisely because you don't ever want to make changes that cause your friends and colleagues to lose their livelihood. And that means that your ambitions to automate are going to run into internal roadblocks for very obvious, very human reasons. Again, that graduation from heuristic to algorithm that Martin talks about.
And if you put the high-volume work that you want to move up that automation chain over time with a third-party provider, there's far fewer internal political issues. And to pursuing a project that will lead to more automation. Not only that, but because of the alignment of business interests, it's easier to convince in an ALSP, which is actually a business with a pal to invest in automating in a way that may cause temporary displacement of certain jobs than it is to convince a law department, which, for all the rhetoric of running a law apartment like a business.
It's not a business, it's a function. And the reason that the distinction matters is because it when you run it like a business, then its interests become primary interests of the business. When it's a function, they are subordinate, they are downstream from the needs and objectives of the business itself. And so, they're run legal like a business only bolsters what is very natural but very problematic department centric thinking.
00:30:17:14 - 00:30:44:04
David Curle
Right. Let's wrap this up by talking about that third level, the algorithmic stage that you described yesterday. And today you published a two-part article in Legal Tech News about everyone's favorite topic these days, ChatGBT, I'd be remiss if we didn't talk about that a little bit, and you address kind of the hype around the issue, the fact that people are spending their time trying to show that ChatGBT doesn't really work or obey what it can do, but also what its shortcomings are.
And I think you do a pretty good job of sort of putting this in context of what the sort of the longer-term implications are. And you actually repeat the prediction you've made before is that in 2023 AI will be capable of generating near instantly a legal opinion or contract superior to the work product of 90% of junior lawyers.
And I think there's a lot in there, and I'd like to sort of unpack that prediction and talk about it in terms of, you know, sort of where we are with things like ChatGBT and other large language models and where the sort of specific impacts are likely to be in the next few years.
00:31:17:23 - 00:31:43:02
Casey Flaherty
Yes. Well, and importantly, that better than 90% of junior lawyers, junior lawyers is doing quite a bit of heavy lifting in that particular statement, because the work product of junior lawyers is generally not particularly good and it's part of their learning process that they produce things that aren't that great, that then go to senior lawyers for refinement. And I have a piece I wrote long ago about how much of lawyering is being a copy and paste monkey, right?
That that speaks directly to this. And so, I am not predicting an AI apocalypse for lawyers. You know, the dawn of robot lawyer age. I do think that these large language models show an enormous amount of promise when they are supplemented with domain specific document sets, domain specific language and tuned and trained. And that training includes reinforce learning from human feedback, and in our case, expert human feedback.
And as lawyers, they show an enormous amount of promise when again targeted to specific tasks to perform the lower-level version of that task extremely well. But it's more than just pointing at something, right? So one of the things that I took aim at in the articles is all the talk about what ChatGBT can do and what it ChatGBT is a new interface for a preexisting language model, ChatGBT three, and it was presented as a preview of progress by Open AI, the Microsoft backed startup. They said from day one, don't use this for anything other than experimentation. It's a general large language model. There's lots of specific things it will not be good at because it is not been tuned or trained to do those, let alone combined with other pieces of complimentary tech.
When you just let LMNs run wild, it gets wild. And so, the example I point to is Galactica, which is a large language model released by Metta that they had to pull down after three days. It was supposed to help with scientific writing, and it could produce some really excellent sounding scientific content. Unfortunately, most of it was fabrication, but it was compelling fabrication.
You had to be an expert in order to know that it was wrong that it was nonsense. I give the example of if the prompt is put into ChatGBT, who are the leading female CEOs in legal tech mining companies? And I've never been a CEO and I've never identified as female. But unless you're deep into legal tech, it's a plausible sounding list, even though not a single person on it is a legal tech CEO.
In fact, I'm the only person on it actually associated with the company that I'm listed next to. And so, these models are very powerful, but they are not godlike in their ability to produce things that are correct or useful. Why don't you approach them the proper way they can be? My favorite example is the going from Web M.D., which we've all typed symptoms into and come away horrified.
You know, its accuracy level is about negative infinity. It's, you know, do I have a cold, or do I have brain cancer, too? There's actually a chart of all these different large language models trained on, you know, PubMed and other datasets. And until very recently, the highest performer was Med Palm from Google. Palm is pathways language model. It's a large language model.
It's actually bigger than GPT three and it got to 67% accuracy in terms of its ability to take a set of symptoms and suggest a diagnosis. That's nothing compared to clinicians who are I think and I'm going to get these wrong, but it will be closer to 90, 92.7% accuracy. You getting get the same set of symptoms. Along comes Google's Med Palm, the successor to Klein Palm and it's at 92.4.
So, it's point 3% below the accuracy of the clinicians importantly from Palm and met Palm use the exact same underlying large language. So this was about the tuning training and all the complements to it to get it to a level of accuracy that's acceptable. And then for me, the question isn't, oh, well, it's as good as the doctors.
Is it better than doctors? No! For me, it's what happens when we equip those doctors who also have an error rate. And medical error is the third leading cause of death in the United States. What happens when we equip the doctors with this tech combined? Can they surpass where each of them are alone? Yeah. Again, I think the focus on ChatGBT is understandable, but it's missing the point of what's possible and also what it takes to get there.
I'm actually worried by all these legal tech vendors who are suddenly putting out press releases about how they've integrated GPT three into their phone. I know we're running out of time because I can.
00:36:13:17 - 00:36:27:17
David Curle
Well; I just want to I want to bring it back to that junior lawyer in your prediction. You know, one of the things you said in the review was you made the point that the widespread use of calculators didn't eliminate the need for math skills. Right. It's the opposite.
00:36:27:18 - 00:36:34:20
Casey Flaherty
How even more than that computer was a human occupation anyway, Lawyers and occupation dentist is an occupation.
00:36:35:03 - 00:36:36:00
David Curle
So, our figure is an.
00:36:36:00 - 00:37:03:18
Casey Flaherty
Occupation for 200 years. And these were people who did long form by hand. This was absolutely necessary work that can contribute was vital to the advancement of science and technology. But eventually silicone surpassed carbon and its ability to do these calculations, both from a speed and accuracy perspective. And we overall became more numerate. There was more work in science and technology.
It accelerated faster, but business and even law became more empirical in its approach. And not all of it is good. Not everything finance has ever done has been good for the world, but it wouldn't have been possible without the silicone computers. But the silicone computers interacting with people who have fluency with numbers, you know, even if we think large language models, we're now passing that a similar inflection point with language.
The facility with language does not become a less valuable skill. There is just a lot of lower-level language intensive tasks that we will no longer have to do. Again, the Whitehead quote Yeah, the number of operations we can perform without thinking.
00:37:46:12 - 00:38:02:10
David Curle
All right. We'll leave it there. We're going to get to the point where we can perform without thinking. There's a lot here. I really appreciate Casey taking time to talk to me. To go back to my original question. I think this last bit of the conversation tells me how much has changed or at least how much change is on the near horizon.
So even though sometimes we feel we're still mired in a static industry, things are coming. So thanks a lot. I want to thank you, Casey. And I also should mention if anyone's interested in the changing lawyer series that I'm referencing, it's at litera.com/tcl. The main report is there, and there's also some regional reports and some other materials, including additional podcasts are there.
So, thanks again, Casey. And appreciate you everyone who's listening. Thanks.
00:38:32:13 - 00:38:33:09
Casey Flaherty
Thank you for having me.
00:38:34:08 - 00:38:42:11
Thank you for listening to LEGALTECH MATTERS. Be sure to subscribe wherever you get your podcasts.
Ready to hear more?
Listen NowVision
Solutions
- Litera Check
- Litera Create
- Litera Compare
- Litera Desktop
- Litera Litigate
- Litera Review
- Litera SecureShare
- Litera Transact
- Clocktimizer
- Concep
- Contract Companion
- DocXtools
- DocXtools Companion
- Foundation Firm Intelligence
- Kira Systems
- Litigation Companion
- Metadact
- Micron Systems
- Objective Manager
- Prosperoware
- PS/Ship
Customers
- Licensing Terms
- Privacy Notice & DPA
- Subprocessors
- Preference Center
- Contact Us
- Modern Slavery Statement
© 2023 Litera Corp. All rights reserved.