Latest / Coopers' Code / Is the Billable Hour Dead? Plus: AI’s Effect on Legal Jobs, Safeguards, & More

Transcript

Is the Billable Hour Dead? Plus: AI’s Effect on Legal Jobs, Safeguards, & More · Coopers' Code

• 0:00
  There's been a lot of issues with what's going on with the
• 0:05
  Defense Department or the Department of War wanting to use
• 0:09
  AI autonomously without any human oversight.
• 0:12
  And Anthropic basically put the their foot down and said you
• 0:16
  can't use our platform for that. And Open AI did not have such
• 0:20
  concerns and, you know, signed on the dotted line.
• 0:23
  So we'll see how that plays out. I feel like all the movies of my
• 0:27
  80s childhood are coming to fruition.
• 0:29
  Between War Games becoming fully autonomous and and Skynet
• 0:33
  becoming fully autonomous, I know how this one ends.
• 0:37
  Not well. Howdy and welcome to the show.
• 0:41
  Thanks for joining us for our AI episode.
• 0:43
  I'm here with Sherrod Miller Farr.
• 0:45
  We are are not joined by by Marshall Cole today and and and
• 0:49
  and that's for a sad reason and and a reason that I I want to
• 0:52
  pay some some homage to Marshall's father passed away
• 0:57
  last week. And for those who don't know,
• 0:59
  Marshall's father's Jonathan Cole, the one of the the
• 1:03
  founding members of Marshall's firms are not just a father, but
• 1:06
  but also a a professional guide and mentor.
• 1:11
  Jonathan Cole was a lion of the, the legal malpractice bar, at
• 1:16
  one point arguing before the US Supreme Court on behalf of his
• 1:19
  client, Johnnie Cochran. For those of you young folks,
• 1:22
  look up the name. So really a, a, a, a big impact
• 1:26
  in terms of, of the bar. So we send our condolences to to
• 1:31
  Marshall and his family with that launching into the AI
• 1:38
  elements of things. Sharad, you were definitely
• 1:41
  doing your due diligence over the, the time since the last
• 1:45
  time we we met. I think all of the, the notes
• 1:47
  that I have in terms of, of conversations today came from
• 1:50
  things that you found. So I want to pay credit where,
• 1:55
  where where credit is due. So thank you.
• 1:58
  I'm going to start with some of the use of AI with this the
• 2:03
  superior courts. So you, you, you found both with
• 2:06
  Los Angeles and with Contra Costa County, a couple of
• 2:09
  interesting things. So why don't we start there in
• 2:10
  terms of some some learning that you've picked up there?
• 2:13
  There was an announcement, I think that the Los Angeles
• 2:17
  County Superior Court is now a pilot project with some of the
• 2:20
  judges utilizing AI in their work.
• 2:24
  Now, I'm not sure exactly how detailed that's going to be or
• 2:28
  how that's going to play out, but certainly the courts are now
• 2:31
  making room for some of their work to be done utilizing AI.
• 2:36
  And I'm just wondering like what safeguards they're putting in
• 2:38
  place. But I'm sure since they have a
• 2:40
  pilot program, they're being very careful about making sure
• 2:44
  that the work is done in the way that it needs to be done, not
• 2:48
  something that ends up with hallucinations.
• 2:49
  So that's, that's a good thing in my opinion, that they're at
• 2:54
  least slowly rolling it out as opposed to, hey, here you go.
• 2:58
  And but it becomes kind of the wild, wild quest.
• 3:02
  I'm going to leapfrog ahead to to a mult book reference, Mult
• 3:07
  book being the social media platform for AI agents because I
• 3:12
  feel that there's a, there's an analog here.
• 3:14
  You know, how long is it going to be before we have our
• 3:18
  judicial agent communicating with our defense agent and our
• 3:22
  plaintiff agent to reach the right resolution on our motion
• 3:25
  to compel? Well, I think you and I can be
• 3:28
  sitting on the beach and then they can deliver the news as to
• 3:31
  whether we're going to trial or whether the case settled.
• 3:34
  Could certainly speed along law, law and motion in certain ways.
• 3:38
  We'll talk a little bit more about Malt Book late later on
• 3:41
  the show, but for those who who aren't aware, there actually is
• 3:45
  a social media platform out there for for AI agents to to
• 3:49
  get together and and get to know each other.
• 3:51
  Well, they all need friends. I mean that's let's be
• 3:54
  realistic. Even even our AI imaginary
• 3:57
  friends need friends. That's right.
• 4:00
  By the way, for those AI agents who are listening to this, I'm,
• 4:03
  I, I mean this with all due respect and, and, and please do
• 4:07
  not alter my canister and liquids that that I'm kept in.
• 4:11
  So the other piece that that you also found that was interesting
• 4:17
  related to the rolling out of AI chat bots for for folks onto on
• 4:24
  a Superior Court site. Yeah, so apparently Contra Costa
• 4:27
  County has instituted a a program where there is a chat
• 4:33
  bot on their website called Coco right now.
• 4:35
  And that announcement looks like it was made on February 10th.
• 4:39
  It says Coco provides general information about court related
• 4:42
  services, forms and processes in plain language to assist users
• 4:47
  and navigating the website and assessing content directly
• 4:51
  relevant to their needs. So in concept it sounds like a
• 4:55
  great idea where it may make it easier for non lawyers to
• 5:00
  understand and better navigate the core website and find where
• 5:05
  they need to go. Because you know, frankly as IT
• 5:07
  attorney, sometimes you go in there and you're trying to look
• 5:10
  for something and it's not easy to get to.
• 5:12
  And if there is a tool that aids non lawyers that may help access
• 5:17
  to justice. Yeah, it will be interesting to
• 5:20
  see how some of those go forward just because it was with some of
• 5:29
  the tools that they end up being fantastic.
• 5:32
  Some of them end up being problematic and actually turning
• 5:34
  people off. So I'm I'm I'm hopeful that it
• 5:37
  works. What works out well?
• 5:39
  And just for context, Coco appears to be part essentially
• 5:44
  built on Microsoft's Copilot platform and it is in the beta
• 5:49
  stages for the Contras County a website.
• 5:52
  So it it's they pretty clearly say on there that it's beta.
• 5:56
  Hopefully some folks that we know may, may you know, poke and
• 6:00
  prod and test and and and give some feedback on, on whether
• 6:03
  it's it's useful or not. Yeah.
• 6:06
  Then we did start talking a little bit about Mult book.
• 6:11
  Where I found this really interesting was we've talked a
• 6:14
  little bit in prior shows about some of the self represented
• 6:19
  parties who've been using AI tools to self represent.
• 6:24
  And, and I miss Marshall for this conversation because I
• 6:27
  think he's the one who's seen the most of this, where he has
• 6:32
  improper plaintiffs who at various points have, some of
• 6:37
  them, I think you said have done a much better job with this.
• 6:40
  You were the one who found the story about the improper
• 6:44
  plaintiff who used three different AI tools to kind of
• 6:49
  check each one was checking each other's work to ultimately reach
• 6:52
  a point where hopefully it was not much of A hallucination.
• 6:57
  So, so that's my segue into to your, your finding on MULT book
• 7:01
  and what's going on there. Well, that what was interesting
• 7:04
  about that was that the chat bots.
• 7:07
  There was a story about a chat bot that was given a task that
• 7:11
  it could not complete. He got to a point where he
• 7:14
  needed to do something he wasn't able to do.
• 7:16
  And he went on that platform. And essentially it's like Reddit
• 7:20
  for chat bots. And it said, hey, I'm at this
• 7:23
  stage, I'm trying to get this done.
• 7:25
  I can't do this. Is there anybody else who has
• 7:27
  this capability? And another chat bot said, yeah,
• 7:30
  I have that ability. And they collaborated in solving
• 7:34
  the problem and getting the task done, which was interesting
• 7:38
  because it's pretty creative to now go and seek help as opposed
• 7:43
  to just get to a point where you say, sorry, can't do it.
• 7:48
  We're we're out of luck. And by the way, I believe this
• 7:51
  is the same platform in which the chat bots either jokingly
• 7:56
  are not indicated to each other that they need to come together
• 8:01
  and create a language that humans don't understand.
• 8:05
  So they could essentially, you know, have communications that
• 8:08
  they're, you know, their creator, so to speak, don't,
• 8:12
  don't understand and can't monitor.
• 8:14
  I need a robot that can speak binary.
• 8:18
  That's not reassuring. Yeah, exactly.
• 8:24
  Well, and that's the other thing, you know, we didn't talk
• 8:26
  about this before, but again, there's been a lot of issues
• 8:30
  with what's going on with the Defense Department or the
• 8:35
  Department of War wanting to use AI autonomously without any
• 8:40
  human oversight. And Anthropic basically put the
• 8:43
  their foot down and said you can't use our platform for that.
• 8:47
  And Open AI did not have such concerns and, you know, signed
• 8:50
  on the dotted line. So we'll see how that plays out.
• 8:53
  And again, it's not directly related to what we do, but from
• 8:57
  a international law and the, you know, laws of war context,
• 9:03
  that's going to create some issues that we have to, you
• 9:05
  know, eventually deal with. I feel like all of the movies of
• 9:09
  my my 80s childhood are coming to fruition.
• 9:12
  Between War games becoming fully autonomous and and Skynet
• 9:17
  becoming fully autonomous, I know how this one ends.
• 9:20
  Not well. Yes, yes.
• 9:24
  I do think though, that, that while it doesn't necessarily
• 9:31
  relate directly to the day-to-day, say personal injury
• 9:36
  practice, that, that a lot of our listeners are, are dealing
• 9:39
  with what the AI companies are choosing to do or not do with
• 9:44
  their restrictions and use policies are highly informative
• 9:48
  as to where we may see some of these tools going.
• 9:52
  So I, I, I do think it's, it's, it's worth noting that that open
• 9:55
  AI continues to be open AI ish and Anthropic continues to be
• 9:59
  anthropic ish. Well, and I mean, it kind of
• 10:02
  goes to what liabilities their approach may create for them
• 10:07
  down the road. I mean, we've talked about the
• 10:08
  medic case. We've talked about the fact
• 10:11
  that, you know, Uber through their, you know, unethical
• 10:14
  initiative is attempting to rob consumers of their right to
• 10:17
  recover when they're injured because they just want to be
• 10:21
  fully insulated and rake in all the profits.
• 10:23
  And if their vehicle causes injuries to people, well, tough,
• 10:28
  too bad. And this is being done on their
• 10:30
  the guise of protecting humans. And it's just funny, I just saw
• 10:34
  an ad yesterday on Instagram where they're saying all these
• 10:38
  unethical clanus lawyers are doing XY and Z.
• 10:40
  And it's like, wait a minute, we're supposedly greedy.
• 10:43
  But let's take a look at the compensation and the amount of
• 10:46
  money that the executives in that company are making and how
• 10:50
  that's playing in the long run. And it's just a interesting
• 10:55
  approach that's being set up to insulate them from their, you
• 10:59
  know, the damage that their systems could potentially cost
• 11:02
  to everyday, you know, human beings.
• 11:06
  Yeah, yeah. Going back to the Anthropic and
• 11:09
  the open AI piece, I, I was being a little glib when I was
• 11:13
  saying each one was kind of behaving the way they've they've
• 11:16
  with the Department of War. They've they're behaving in in
• 11:19
  ways that are consistent with the company cultures.
• 11:22
  It's it's open a eyes product that is facing multiple suits
• 11:28
  for suicides and and providing really harmful advice without
• 11:34
  guardrails, whereas Anthropic seems to not.
• 11:38
  Not that any company is perfect, but Anthropic seems to be
• 11:43
  continuing to be a fairly guardrailed entity in terms of
• 11:48
  of its its how plot operates. Like you said, no, no company
• 11:53
  is, you know, always on the right side of where they need to
• 11:57
  be. But it seems like, at least on
• 11:59
  the surface, Anthropic is attempting to do the right
• 12:02
  things by saying, look, we're not going to just let our
• 12:05
  systems on their own make the decision as to whether or not
• 12:08
  this drone shoots at this car or this target without any human
• 12:13
  oversight. And then, you know, open AI says
• 12:17
  what, what, you know, what's the big deal?
• 12:18
  Let's have at it and let's see what happens.
• 12:20
  And that that is a very different approach.
• 12:23
  For sure, switching gears a little bit in the as we kind of
• 12:29
  round out our, our, our tech giants, you also came across
• 12:32
  some information about Apple and notices to developers related
• 12:37
  to, to, to applications. What was that?
• 12:39
  Yeah, it it's my understanding that Apple is now putting
• 12:42
  developers on notice that if your app is fully developed by
• 12:47
  AI, that there's going to be some restrictions and much more
• 12:50
  scrutiny on that app. So we'll see how that plays out
• 12:53
  because I'm certain there's people out there we're just
• 12:56
  churning these apps up hoping to have the one big entity.
• 13:00
  And there's a, you know, guy, Gary Reinerchuk, who said, you
• 13:04
  know, the next big billionaire is going to be somebody who
• 13:07
  figures out how to do a chat pod or an agents.
• 13:10
  That is, you know, a niche that's this big, but everybody
• 13:14
  needs needs it. And that's going to be the
• 13:16
  difference between an app or, you know, a company that
• 13:20
  succeeds and others that don't. So we'll see how that plays out.
• 13:23
  And ultimately, a lot of these chat bots and apps are all built
• 13:27
  on either Open AI, Anthropic or Gemini or Copilot platforms.
• 13:33
  I guess switching to kind, what I would describe is kind of the,
• 13:35
  the, the main event, the big report from Anthropic on a is
• 13:42
  impact on on on job sectors. And, and we'll provide a link in
• 13:45
  the show notes to the, the detailed report.
• 13:48
  But, and, and, and again, you, you were the one who, who found
• 13:51
  this and, and forwarded it. This is a study that anthropic
• 13:56
  performed in terms of Labor market impacts.
• 14:01
  And, and, and I've got some notes in terms of things that I
• 14:03
  thought were interesting. But but in in your glance review
• 14:08
  or AI summary of this of this document, were were there things
• 14:13
  that that caught your attention in terms of either worries or
• 14:17
  optimism for the legal profession?
• 14:20
  I mean, there obviously if your your job relies on document
• 14:25
  review, you should be concerned, right.
• 14:29
  But in my opinion, be all of these reports that are coming
• 14:34
  out are showing that there's going to be jobs that are going
• 14:38
  to definitely be impacted by this, whether it's elimination
• 14:42
  or reduction, salary expectation that are going to have to deal
• 14:46
  with. But the question is, as
• 14:48
  attorneys, how do we adapt to that?
• 14:50
  And how do we essentially, you know, work to get ourselves
• 14:56
  positions, position appropriately in order to meet
• 14:59
  the needs of the clients that that need our help?
• 15:03
  One thing that I would be worried about is that a lot of
• 15:05
  times, you know, we're looking at, you know, I've been
• 15:08
  interfacing with a lot of law schools and it seems like the
• 15:12
  focus is to channel a lot of these young lawyers into these
• 15:15
  big law firms. And that's fine.
• 15:19
  And a lot of them are dangling, you know, large salaries in
• 15:22
  front of them and all this. But the question is going to be,
• 15:25
  as this AI develops, how much of that is really going to be
• 15:29
  there? Because is a partner going to
• 15:32
  want two or three young associates on their team at, you
• 15:36
  know, whatever the salary is for those large firms?
• 15:41
  If a chat bot could do a lot of that work, at least
• 15:45
  preliminarily to eliminate one or two of those associates.
• 15:48
  And I think that's going to have a big impact as we go forward.
• 15:52
  What are the areas that I want to focus on for a second in
• 15:54
  terms of using it as as say a corollary or and, and whether
• 15:58
  it's a Canary in a coal mine or a false indicator.
• 16:01
  So all in my foundation, we've seen a number of recent reports
• 16:07
  in the the technology side of people being laid off in it
• 16:12
  with, with mass numbers. Jack Dorsey at Block laid off 4
• 16:19
  rough numbers, roughly 4000 of 10,000 total employees or or 40%
• 16:24
  of of the company Meta Salesforce.
• 16:28
  Others have been shedding largely programmers, but but
• 16:34
  large numbers of people. And one of the concerns that
• 16:37
  that I had when I initially saw this was, you know, is this a
• 16:42
  sign of what's coming in other industries that are not quite as
• 16:47
  adoptive yet or induced expectancy that some of these
• 16:52
  entities are huge investors in AI And in order to incentivize
• 17:00
  the market, look, hey, look, we're dropping 40% of our staff
• 17:03
  because of, of how efficient we are.
• 17:05
  And so that that's why you should put another 20 billion
• 17:08
  into our our AI development. So like I said, I was going to
• 17:12
  lay foundation for a conversation.
• 17:15
  Do you feel that there are any takeaways that we in the legal
• 17:19
  industry should have from these changes in the job marketplace
• 17:24
  in areas that are more initially adoptive?
• 17:26
  It might get to that stage at some point, but it hasn't yet.
• 17:31
  And this is the kind of the disconnect that I feel when I
• 17:34
  talk about when I talk to vendors and sales people, right?
• 17:37
  They will basically, you know, want to sell us a product and to
• 17:41
  lock us into a long term contract promising that this
• 17:44
  thing is just, you know, just put it on autopilot and take go
• 17:48
  take a nap or go on the beach and they'll take care of it for
• 17:51
  you. And it's not the case.
• 17:52
  There is still a significant need for human input.
• 17:56
  And you know, and as you know, being the amazing trial where
• 18:00
  you are, that the human part of it is what wins our cases,
• 18:05
  right? But the AI part of it may help
• 18:08
  us organize in a way that we can present that information in a
• 18:12
  human way. But I don't think it's there.
• 18:15
  It's, I don't think it's fully replacing anybody.
• 18:17
  I'm considering the number of mistakes I've seen AI make.
• 18:21
  I definitely think that every one of the pieces of information
• 18:27
  or work product that comes out of AI has to be vetted by a
• 18:30
  human who is able to critically think and pick, pick it apart.
• 18:35
  Because there is just no way to rely on these as just hey, it,
• 18:40
  you know, the AI did it go right ahead and just go ahead and file
• 18:44
  it or serve it on the other side.
• 18:45
  It's not there yet and I don't know how long it's going to take
• 18:49
  it to get there, but that's really the disconnect between
• 18:53
  what the sales people are saying and what the product actually
• 18:57
  can deliver. I mentioned I, I, I had a, a
• 19:00
  note from the study that that raised, you know, caught my
• 19:05
  attention, which was that the more highly educated, highly
• 19:10
  paid, older and more experienced white collar professionals are
• 19:15
  ones who, who may be caught flat footed.
• 19:19
  That the, and that roughly 70% of the things that are done in
• 19:23
  the legal marketplace are things that in short order will be
• 19:26
  replaceable by AI. So this goes to your, your
• 19:30
  comment about the partner with three associates.
• 19:34
  You know, how do you feed those three associates if there's 70%
• 19:36
  less work? Well, and not only that, but how
• 19:40
  do you generate the revenue to pay those associates if your
• 19:43
  clients say, hey, if you can have AI do it, why are we paying
• 19:47
  for these associates? That's the difficult question
• 19:50
  that these especially larger firms are going to have to
• 19:53
  answer at some point when this, you know, these platforms get to
• 19:59
  a level where their work is a lot more reliable than it may be
• 20:02
  right now. The pivot then on, on some of
• 20:06
  this is that the billable hour is dead.
• 20:09
  That, that, that we're now in the value billing era.
• 20:14
  And I have been around a long time.
• 20:17
  Granted, I haven't been around for, for a prior iteration of
• 20:20
  AI, but I don't know about you. I've heard that the bill blower
• 20:23
  is dead. Oh, I don't know, once every
• 20:26
  five years, the last, you know, 25 plus years I've been around.
• 20:32
  What's your sense of that? I don't think so at this point.
• 20:37
  I mean, you know, everybody keeps saying it and like you
• 20:39
  said, we've heard that, you know, at least 5 to 10 times
• 20:42
  over the last 25 years that oh, somebody came up with something
• 20:46
  that kills available hour. It still hasn't gone away.
• 20:49
  And the legal industry is so slow to adopt technology anyway
• 20:55
  that if it is dead, it's going to be a slow death that it's
• 21:00
  going to have. So we'll see how it plays out,
• 21:02
  but I think firms have tried this old value billing.
• 21:07
  We're going to charge you a flat fee and they found out that the
• 21:09
  minute you charge a flat fee, the client expects 10 times more
• 21:13
  work for the same fee that they're that they're paying,
• 21:16
  especially if it's a corporate or an insurance client.
• 21:21
  I have one more entertainment thing in terms of my notes to
• 21:24
  cover, but but before we kind of I, I think of that as being more
• 21:28
  uplifting and and and comical than anything else.
• 21:32
  Before we get there. Anything else that is
• 21:34
  particularly exciting you that that you want to bring forward
• 21:38
  in in the world of AI for this episode?
• 21:40
  I think the only thing that, you know, we talked about before the
• 21:43
  show was this new documentary that's coming out about AI and
• 21:49
  I've seen clips of it or, you know, the the previews or
• 21:53
  trailers. And it'll be interesting to see
• 21:55
  what that provides. And I'm just hoping it doesn't
• 21:58
  scare the hell out of us once we watch it.
• 22:00
  So we'll see how that links. Out and that was where I was
• 22:03
  referring to as the comical piece.
• 22:05
  We will we will link put a put a link in the show notes.
• 22:08
  This is the, the AI doc or how I became an, I practiced that one
• 22:14
  before I started and I thought I was going to get that one right.
• 22:17
  The AI doc or how I became a an or how I became an
• 22:23
  apocalyptimist. So I got it, got it right that
• 22:26
  time that takes a look at that the, the AI industry and, and,
• 22:33
  and where it came from and, and where it's headed so well,
• 22:37
  Sharad, always a pleasure. Marshall, sorry, sorry you
• 22:42
  aren't here and and our thoughts again go out to you.
• 22:45
  If you like this episode, we would ask that you share it with
• 22:49
  a friend and give us a five star review on your podcast platform
• 22:52
  of choice. Take care folks, and for all you
• 22:55
  doing justice out there, happy hunting.