Latest / Coopers' Code / Who Controls Speech When AI Moderates Everything?: This Month in Legal AI

Transcript

Who Controls Speech When AI Moderates Everything?: This Month in Legal AI · Coopers' Code

• 0:00
  These companies are now using AI to moderate all of this content.
• 0:05
  It'll be interesting to see how that plays out from a censorship
• 0:08
  standpoint, because before it was how many people can Facebook
• 0:12
  or TikTok employee to moderate what's going on.
• 0:14
  Now you have very sophisticated AI algorithms that are going to
• 0:18
  be unleashed on the uploads of the various content and how
• 0:25
  that's going to affect the big picture in terms of
• 0:27
  communication and what people are sharing.
• 0:29
  And what actually ends up in your feet, depending on what
• 0:33
  they think you're leaning is from a political or social
• 0:36
  standpoint. Howdy, and welcome to Cooper's
• 0:39
  Code. This is the AI issue and we're
• 0:43
  joined with Sherrod Milanfar and today, unfortunately, Marshall
• 0:47
  Cole's in trial. And for those of you who are
• 0:49
  keeping the score in terms of Marshall Cole being in trial,
• 0:51
  he's got a pattern. So he's either like ADA or
• 0:54
  somebody who doesn't settle out much.
• 0:55
  So if you're trying to stalk him and learn about him from from
• 0:58
  the handle in the podcast, know that you you may have a trial on
• 1:01
  your hands. So we wish him the best of luck
• 1:03
  there and look forward to joining when we return.
• 1:05
  So, Sherrod, how are you doing? I'm doing well, thanks.
• 1:09
  How are you? I'm, I'm, I'm doing well.
• 1:11
  It's nice to return to this after having a little bit of a
• 1:13
  of a holiday break. I wanted to kind of jump right
• 1:15
  into it. You had flagged something that
• 1:18
  it sounds like it could be very interesting to watch.
• 1:21
  And and that is February 9th, those who want to see Mark
• 1:26
  Zuckerberg take the stand in trial, get the opportunity.
• 1:30
  You you want to go into a little bit of of this story.
• 1:33
  Yeah. So basically the Daily Journal's
• 1:35
  report and other app that's have reported that their trials that
• 1:39
  are the social media addiction trials are getting ready to go
• 1:43
  down in Los Angeles. And Mark Zuckerberg is slated to
• 1:46
  testify in that case. And the case on the plaintiff
• 1:51
  side, it's more of the bellwether cases is being
• 1:53
  handled by the famed trial lawyer Mark Lanier.
• 1:57
  So it, you know, just as a human being and as an attorney, it
• 2:01
  would be great to watch that interaction.
• 2:04
  Apparently plaintiffs have asked the court to allow the
• 2:08
  proceedings to be recorded and the defense is indicating that
• 2:12
  they're going to file motions to essentially exclude any kind of
• 2:16
  cameras or recording in the courtroom.
• 2:19
  So it'll be interesting to see what the court does with that.
• 2:22
  Apparently, what app and TikTok or a Byte dance, I guess it is
• 2:29
  on one of the subsidiaries of TikTok have settled out, but
• 2:33
  Meta and Google are still in the case.
• 2:37
  So we'll see how that pans out. But Bytedance was the one that
• 2:42
  actually settled, apparently either today or last night,
• 2:44
  because the news came out this morning about that.
• 2:48
  That should be really interesting and and we don't
• 2:51
  generally see the CE OS of companies take taking the stand
• 2:56
  in cases like this. For those who are are following
• 2:59
  along, this is in LA Superior Court.
• 3:03
  I I don't know that it will be live streamed or if it's going
• 3:07
  to be on court TVI don't know if you know the answer.
• 3:10
  I don't, I mean sometimes CVN Courtroom View Network is
• 3:13
  involved, but they, I don't believe the court has ruled on
• 3:17
  it yet. So hopefully it's on CVN and you
• 3:20
  know, people can watch it there or on court TV because I think
• 3:25
  it'll be of interest to just the general public and the legal
• 3:28
  community as a whole. Yeah.
• 3:30
  The information that may come out in terms of what these large
• 3:38
  entities are doing, in terms of gathering data, in terms of
• 3:41
  tailoring content and and in terms of getting us addicted to
• 3:46
  the school, whether it be children or adults.
• 3:49
  At the same time having Montessori schools in their own
• 3:53
  homes that do not allow the use of any electronics by their own
• 3:58
  children. So that's an interesting
• 4:01
  juxtaposition in terms of, hey, you know, all these apps are
• 4:05
  great for your kids, but not mine, even though the amount of
• 4:10
  revenue that comes in is directly tied to all these
• 4:14
  eyeballs that are stuck on these apps for hours a day.
• 4:18
  As we're talking about that, there's also the consolidation
• 4:25
  within the media landscape overall.
• 4:28
  And, and here we're we're referring to TikTok and, and
• 4:31
  what's being done with TikTok as far as its ownership is
• 4:35
  concerned. Yeah.
• 4:36
  I mean, you look at the TikTok now, it's been taken over by the
• 4:42
  Allison LED investor group in the United States.
• 4:45
  So the Chinese interests presumably have been divested,
• 4:49
  and they're very aggressively moderating the content now.
• 4:54
  And a lot of creators and social media influencers are
• 4:57
  complaining that if they say something that's not consistent
• 5:01
  with the philosophies of Ellison and his cohort, that those
• 5:07
  pieces of content are being thwarted and they're not getting
• 5:10
  any views on it. So it'll be interesting to see
• 5:12
  how that plays out from a censorship standpoint and how
• 5:16
  these companies are now using AI to moderate all of this content.
• 5:21
  Because before it was how many people can Facebook or TikTok
• 5:25
  employee to moderate what's going on.
• 5:26
  Now you have very sophisticated the AI algorithms that are going
• 5:31
  to be unleashed on the uploads of the various content and how
• 5:37
  that's going to affect the big picture in terms of
• 5:39
  communication and what people are sharing and what actually
• 5:43
  ends up in your feed depending on what they think you're
• 5:47
  leaning is from a political or social standpoint.
• 5:50
  One of the things that I'm I'm hearing from you as a potential
• 5:52
  take away as far as how it relates to AI and the law is
• 5:57
  that as this consolidation continues, as this AI tool for
• 6:04
  moderation continues, we, we run the risk of having tremendous
• 6:09
  power held by very few over what was being seen by by everyone,
• 6:16
  including potential jurors. Absolutely.
• 6:20
  I mean it. It really is a matter of who
• 6:23
  gets to have a voice and the fact that there is basically the
• 6:29
  federal government is trying to tell the states you cannot
• 6:32
  legislate and put any safeguards on any of these of AI platforms.
• 6:39
  So what we don't understand is OK, what what are they creating
• 6:44
  right. Then somebody described this.
• 6:46
  I think we talked about that there in the ADC conference is a
• 6:49
  black box and it's being fed a tremendous amount of power.
• 6:53
  But yet they really don't understand what it can do.
• 6:57
  But on top of that, they're giving it now instead of letting
• 7:00
  it train and kind of be on a neutral plane, they're giving it
• 7:04
  essentially a very biased training where if I say
• 7:07
  something or you say something that doesn't necessarily line up
• 7:10
  with the owner's views, you're going to be called or, you know,
• 7:16
  labeled as an enemy of some sort.
• 7:19
  And that's very disturbing just from a place of living in this
• 7:24
  country that's very unique in its approach to having a voice
• 7:28
  and the 1st Amendment. Within the concerns of of having
• 7:33
  a voice privacy, your own awareness, have you done
• 7:37
  anything lately in terms of Google and and updated privacy
• 7:40
  policies around making sure that you're doing the best you can to
• 7:45
  protect yourself as we move forward with more and more
• 7:47
  consolidation? That's interesting because as
• 7:52
  you know, and anybody who's tried to read a privacy policy
• 7:57
  in online finds out that these are incredibly complex
• 8:01
  documents. And I think when it came to
• 8:03
  these online forms, I think it was Judge John Roberts who is
• 8:08
  quoted as saying that he doesn't even read these things when he
• 8:11
  clicks on them, right? So it's a problem because you're
• 8:15
  clicking on things, and that then may have consequences down
• 8:19
  the road. So Google sent out an e-mail
• 8:22
  saying we've changed our privacy policies, and I clicked on it
• 8:25
  just to look at it. It's this incredibly long
• 8:27
  document. So I'm like, all right, how can
• 8:29
  I digest this a little bit easier and figure out how I can
• 8:32
  delete my information if I choose to do so?
• 8:35
  So I took that document or link and I put it on notebook LM and
• 8:39
  I said give me a synopsis and I asked the questions and it can
• 8:44
  give you information about this is how you go about deleting
• 8:48
  your personal information. It's a five step process that it
• 8:51
  creates. So there are creative ways to
• 8:54
  use it to make it efficient for you to essentially counteract
• 8:58
  some of the things that we're seeing.
• 9:00
  Because presumably if you are able to delete your personal
• 9:03
  data out of Google, then that lessons some of the information
• 9:07
  that they have on you. And also, California just passed
• 9:10
  the law where you can go in and tell data brokers to delete your
• 9:15
  information, which is also a great thing in my opinion,
• 9:18
  because you should have a choice as to what information about you
• 9:22
  is being sold out there and how that's being used to market to
• 9:25
  you. And, you know, try to figure out
• 9:29
  how they can essentially get you to buy things.
• 9:32
  And for those who are interested in, in what Sharad just
• 9:34
  referenced, that is, is the drop DROP program through the state
• 9:40
  of California. And, and you can go on there
• 9:42
  and, and fill out the form to help reduce your overall profile
• 9:46
  and, and tell data brokers that that this is not something that
• 9:50
  they can have. Is that fair?
• 9:52
  Yes, that's that's exactly what it is.
• 9:54
  And also the other thing is, I think Apple just instituted a
• 9:58
  function on the on your iPhone where you can essentially have
• 10:02
  the person say who they are before your phone rings so that
• 10:06
  it would help reduce on some of them spam calls that we get on
• 10:12
  our phone at least 2-3 times a day minimum.
• 10:16
  That would be nice, wouldn't it? Yes, of course.
• 10:19
  Then there's that flip side. And I, I, I mean, this is not
• 10:22
  the AI episode, but this is something that I always, always
• 10:25
  worry about as a service oriented personal injury lawyer,
• 10:30
  which is it's that number that I don't recognize.
• 10:34
  It doesn't say spam. Is this somebody who's trying to
• 10:37
  trying to get a hold of me? And if I don't pick it up, what
• 10:40
  happens? Right.
• 10:41
  Yes, absolutely. And I think your family and my
• 10:44
  family can have a robust discussion about why we take
• 10:47
  calls at times where we should not be paying attention to our
• 10:52
  phone. But that is a real fear that we
• 10:54
  all have as personal entry lawyers.
• 10:56
  Yes, it'd be a different abbreviation than FOMO.
• 10:59
  You know, phobia, fear of personal injury, case loss.
• 11:04
  I don't, I don't know. But in the realm of of AI, we
• 11:11
  talked a little bit again before we got on about some of the the
• 11:15
  bigger moves that are going on in the legal AI marketplace.
• 11:19
  A lot of news this week about Harvey's acquisition of a
• 11:23
  another company called Hexis. For those who are, are not
• 11:26
  familiar, Harvey is, is one of the larger AI platforms in the
• 11:31
  legal space, primarily marketing to the larger entities.
• 11:35
  And I, I know that there are some personal energy firms that
• 11:37
  are using it on the whole, though they, they tend to be
• 11:40
  more of the, the big firm platform.
• 11:45
  And did you have any thoughts on on their acquisition on Hexis in
• 11:49
  terms of what it what it says in terms of Hexis being product
• 11:54
  demonstrations and video guides to help people learn how to use
• 11:57
  products? My only observation is that
• 12:00
  basically consolidation seems to be the consistent pattern with
• 12:06
  companies and private equity, right?
• 12:08
  It's a you, we're seeing it in the space of, you know, Netflix
• 12:12
  wanting to buy other companies and Allison wanting to buy all
• 12:15
  these media companies. And the question is at some
• 12:19
  point is there going to be the return on investment?
• 12:24
  Is it really going to be there? Because the amounts of money
• 12:27
  that's that are being put on the table for some of these things,
• 12:30
  you just kind of scratch your head and say, wow, like, how did
• 12:33
  you calculate the valuation on these things?
• 12:37
  And that then leaves to leads into the whole issue about the,
• 12:41
  you know, AI bubble and whether it's real or whether it's not.
• 12:45
  But again, hopefully somebody at Harvey had a plan and understood
• 12:50
  exactly what they were acquiring and why they were acquiring it
• 12:53
  because again, you know, having to go to my boss and tell them
• 12:57
  we're spending this much money, there better be a good why and a
• 13:01
  strategic plan on the return on that investment.
• 13:04
  I think some of the things that, that, that caught my attention
• 13:07
  in, in this particular deal was, was Harvey's current $8 billion
• 13:12
  valuation in terms of its last funding round.
• 13:15
  And, and I had forgotten, but Andreessen Horowitz is, is one
• 13:19
  of the VCs backing it. And, and that's a, that's AVC
• 13:22
  backer who's known for making wise decisions in terms of what
• 13:25
  they've invested in. So when you, when you see that
• 13:29
  valuation, when you see the backing for me, leads me to
• 13:33
  believe that this is potentially going somewhere.
• 13:35
  The the flip side of that is that's a huge valuation and and
• 13:41
  while there is a fair amount of money in the legal marketplace,
• 13:46
  is the amount of money that's being generated from law firms
• 13:51
  likely to justify that that kind of an expenditure?
• 13:55
  The question really is going to be whether this is being looked
• 13:59
  at as something that's essentially an investment that's
• 14:02
  going to go directly to law firms.
• 14:04
  Or is this private equity playing their card in setting up
• 14:09
  systems that's going to allow them in markets such as Arizona,
• 14:13
  where they're allowed to own law firms without attorney owners,
• 14:17
  to undercut the pricing model of the large firms that exist in
• 14:23
  those markets. And we've seen, you know, 30
• 14:26
  attorney groups leaving their traditional firms to go to firms
• 14:30
  like KPMG and establish these, you know, pseudo law firms that
• 14:36
  operate in, you know, Arizona and other places.
• 14:39
  But the question they are going to have to ask themselves is at
• 14:42
  some point once you set this thing up, are they going to say,
• 14:45
  well, you know what, we don't really need you.
• 14:47
  We can pay some junior associates and Harvey's going to
• 14:50
  run the show. And if that's the play, then
• 14:53
  there's potentially a significant upswing that, you
• 14:56
  know they can take advantage of in terms of pulling revenue away
• 14:59
  from these traditional firms. I think you were the one who,
• 15:02
  who pointed out long, long ago to me, you know, eons ago in the
• 15:07
  world of AI, like 6 months ago, that if, if you have AAI tool
• 15:16
  that say costs $50,000 a year, but it allows you to, to treat
• 15:22
  it like a junior associate or a junior paralegal and, and, and,
• 15:27
  and be the equivalent of, of multiples of those, then you're,
• 15:30
  you're actually are looking at a cost savings overall.
• 15:32
  Right. That's the analysis that they
• 15:35
  have to make, right. And when you look at a, you
• 15:38
  know, these companies that are coming in to undercut the large
• 15:42
  defense firms or the traditional large firms, they're looking at
• 15:46
  it from the business perspective of we can do this in a way
• 15:49
  that's cheaper for you and paying for Harvey at whatever,
• 15:53
  you know, but probably, you know, 4 or $500 a month for that
• 15:57
  service. And to cut out an attorney whose
• 16:00
  salary is significantly higher than that, that's something
• 16:04
  they're going to tell their clients.
• 16:05
  And then the clients are going to look at it and say, well,
• 16:08
  this is a huge savings on defense or mergers and
• 16:12
  acquisitions or whatever. And that could have a
• 16:15
  significant ripple effect in the industry.
• 16:18
  Now, you're still going to have to have humans who are going to
• 16:21
  review things that we, as we've discussed before.
• 16:24
  But again, if that can eliminate the first draft and get you to a
• 16:30
  point where you can eliminate two or three people on your
• 16:33
  team, that's a huge savings that they're, they're going to force
• 16:37
  that, you know, firms and companies to go into.
• 16:41
  We had probably about a month and a half since we last got
• 16:44
  together. Are are there things that you've
• 16:46
  been playing around in the interim that have been making
• 16:49
  you happy, making you interested, whether that be in
• 16:51
  the field of prompting or other other tools, other choice?
• 16:56
  Mostly it it seems to me that all these AI systems are getting
• 17:00
  better and the more creative we are, the better we're going to
• 17:04
  be able to utilize them. So again, this kind of an old
• 17:08
  theme, but the, you know, keep looking at the different AI
• 17:12
  systems like notebook LM has certain utilities and they're
• 17:16
  now adding more topics to it where you can do the
• 17:21
  infographics and slide decks and the podcast feature now has it.
• 17:28
  So if let's say you and I are talking, if we could have a
• 17:31
  caller call in and ask a question and you'll adapt the
• 17:35
  podcast based on the real time question that's coming in.
• 17:38
  I mean, these are things that have happened in the last six
• 17:41
  months. So as imperfect as these systems
• 17:45
  are, where you should be very vigilant and to, you know, make
• 17:49
  sure that you're not sending anything out the door without,
• 17:51
  you know, human eyes checking them, They're improving
• 17:55
  significantly and making things so that we can be more
• 17:59
  efficient. And as we've talked about, you
• 18:01
  know, myself having a smaller firm, we're able to stand up to
• 18:05
  much larger firms because we're able to look at the big picture
• 18:09
  and utilize various technologies that are available to streamline
• 18:12
  our our approach. In that vein, and, and, and if
• 18:16
  I've asked you this in the past on air and, and, and you've
• 18:19
  answered already, forgive me, but do you use copilot?
• 18:24
  And, and by copilot, I'm referring to Microsoft's product
• 18:28
  that it is embedded into Office outlook other places.
• 18:34
  Is that something that you felt comfortable turning on and, and,
• 18:37
  and talking back to you or is or is it something that you've left
• 18:39
  alone for now given some of the privacy concerns?
• 18:41
  I haven't utilized that in the sense of looking at specific
• 18:45
  case related information, but what I've used is, you know, we
• 18:52
  have Cleo and it has an AI component to it.
• 18:55
  So I've used that. And then Anthropics Claude is
• 18:59
  able to do what's called projects, which are very useful
• 19:03
  because you can put some information in there and then
• 19:06
  you can use that as an example. So if we have things that we
• 19:09
  want to utilize and use as templates, those are things that
• 19:15
  we can feed into it and then it can do what it needs to do.
• 19:18
  And based on my research on it is the most secure out of the
• 19:22
  other Enterprise Products that are available.
• 19:25
  So it makes it so that it's doing a better job of being
• 19:30
  consistent with what we want based on the information that
• 19:33
  we're feeding it. And it's in a closed loop, not
• 19:35
  training the larger model. Simultaneous to that,
• 19:39
  independent of whether it is closed loop or or keep your data
• 19:43
  quiet. Google has really been making a
• 19:47
  push lately with Gemini. When you open Chrome, when you
• 19:51
  use other tools, it it definitely is getting in your
• 19:54
  face in terms of trying to be there as a resource for you.
• 19:57
  Is is that something that you've found is useful in a legal
• 20:01
  space? Has that impacted you at all?
• 20:03
  I know that what Google is doing is basically trying to make
• 20:07
  their, you know, search engine an AI search engine, right?
• 20:10
  I mean, that's, I mean, they would be dumb not to do it from
• 20:15
  a legal standpoint. I haven't really gotten into it
• 20:19
  because obviously there's some concerns like we're not going to
• 20:22
  put client information in in there.
• 20:25
  But it does have utility in the sense that, you know, you have a
• 20:29
  Casey has a publicly traded company.
• 20:31
  You can ask them, Gemini, to give you a deep dive research
• 20:36
  paper on that company's finances.
• 20:39
  And you know, I, we, I did that 2 days ago on a case that we
• 20:43
  have that's a wrongful death case and the companies publicly
• 20:46
  trade it. So when we ran the company, it
• 20:49
  was clear that despite their assertions that, oh, you know,
• 20:53
  we don't have that much money, they have 300 million in cash on
• 20:56
  hand. And they're, they're, they're
• 20:59
  more than able to, you know, essentially cover the case and
• 21:02
  their insurance becomes irrelevant.
• 21:04
  So for us as legal consumers, knowing which particular and,
• 21:12
  you know, AI engine to use I think is important because it'll
• 21:16
  provide you with better information where, I'm not sure
• 21:20
  if I went to ChatGPT, I would have gotten the same sort of
• 21:24
  research on it, but maybe, you know, somebody else has done it.
• 21:27
  And I figured Gemini was the most powerful engine when it
• 21:32
  comes to searching publicly available information about a
• 21:36
  publicly traded company. That makes a lot of sense as we
• 21:39
  talk about some of the larger companies grabbing your crystal
• 21:43
  ball. Do you have any prediction in,
• 21:45
  in the legal world as to whether the Elon Musk versus open AI
• 21:49
  suit will actually get to trial? I, I know it, it passed the,
• 21:52
  the, the motion to dismiss hurdle and and so they get they,
• 21:56
  they get the opportunity of trials.
• 21:58
  You think we'll actually get there?
• 22:00
  You know, I, I, I think I have to put my mediator hat on that
• 22:03
  one on for that one. But based on the animosity that
• 22:08
  I've seen between Elon Musk and Sam Altman, if, if it goes to
• 22:14
  trial and I think it's very likely it's going to be in a
• 22:17
  very entertaining trial. So I would love to see a
• 22:20
  televised so we can see what these guys have to say and how
• 22:24
  they're going to prove their case against each other.
• 22:27
  Yeah. Well, and, and kind of we, we
• 22:30
  talk I think a little bit at the beginning of the case of the
• 22:32
  opportunity to see certain people, you know, Mark
• 22:35
  Zuckerberg in trial. We've got the potential of, of
• 22:38
  seeing some, some big people in trial in, in the coming year
• 22:43
  should be very interesting just from the spectator side of the
• 22:46
  law and, and seeing what people have to say and, and where
• 22:49
  things go. Well, and I think that kind of
• 22:52
  dovetails into this whole issue of what's happening in AI in
• 22:56
  terms of autonomous driving, right?
• 22:58
  With these rideshare services and specifically Uber attempting
• 23:03
  to essentially strip consumers of their rights to pursue
• 23:08
  litigation when they're injured and simultaneously trying to get
• 23:13
  rid of their drivers, which means that they're trying to
• 23:16
  insulate themselves from liability caused by their
• 23:22
  imperfect driving systems that are going to be run by their
• 23:27
  AIS. And that creates product
• 23:29
  liability issues. It creates access to justice
• 23:32
  issues. And as human beings and
• 23:34
  consumers, we have to ask ourselves, how could we possibly
• 23:37
  blindly say, yeah, you know what, Uber, you're never going
• 23:40
  to have any liability if your car hits and kills a child in
• 23:44
  the in the crosswalk. And it just, it's just amazing
• 23:48
  to me that a company like that would just attempt to even pass
• 23:52
  that kind of legislation because it's just so unconscionable on
• 23:55
  its face. We're seeing some unconscionable
• 23:59
  things happening in this world right now, aren't we?
• 24:02
  Yeah, a little bit. A little bit.
• 24:04
  That's an understatement. Yeah.
• 24:06
  Well, anything else you think we should we should touch on in the
• 24:09
  in the the tech episode? I don't remember if we talked
• 24:12
  about this before, but the American Arbitration Association
• 24:15
  has now released an AI arbitrator for.
• 24:19
  Oh, this is news to me. Now this is interesting.
• 24:22
  So they've released a module that's an AI arbitrator and had
• 24:27
  believe it's marketed as an arbitrator for like construction
• 24:32
  defect type cases. And that's something that
• 24:36
  they're trying to push out there.
• 24:37
  But I suspect, you know, a lot of the carriers, they do a lot
• 24:40
  of arbitrations between themselves in behind the scenes
• 24:44
  with cases that they pay out on and stuff.
• 24:47
  So maybe that'll take off there. But to me, it's just amazing
• 24:51
  that you would just remove the human out of it and you would
• 24:54
  want an AI arbitrator to arbitrate your case.
• 24:57
  I I don't know who would actually submit to that, but
• 25:01
  that's a product that they just released a few months ago.
• 25:04
  Just it sounds like the equivalent of a digital table
• 25:07
  saw splitting the baby. The parallel piece of that seems
• 25:13
  like if that was a tool that was going to be regularly used, that
• 25:17
  you'd see if you could try to model it to see what your
• 25:19
  outcome would be so that you know what your best alternative
• 25:23
  to a negotiated outcome is. There is a paper I think that
• 25:27
  came out from Columbia University that did that.
• 25:30
  And their models clearly were wrong in their approach.
• 25:35
  And they actually, I think it was Grok, they asked and
• 25:38
  initially Grok said this case is worth like $25 million.
• 25:42
  And then later on when they asked the additional questions
• 25:46
  said, oh, on second thought it's $2,000,000.
• 25:49
  That's a pretty large swing to consider in terms of the
• 25:55
  results. And again, there is a reason we
• 25:59
  go to trial is that that human piece of it is something that's
• 26:03
  going to be very difficult for them to emulate in terms of AI
• 26:07
  analysis that could it give you some idea of what would work and
• 26:12
  what wouldn't work? Probably.
• 26:13
  But is it definitive? I doubt it.
• 26:17
  Yeah, that's that's fascinating for all you listening out there.
• 26:20
  If you are seeing anything new or cool that you want to bring
• 26:24
  to our attention, if you are using the arbitration AI and get
• 26:29
  outcomes, we'd love to hear what those are.
• 26:31
  And otherwise if you're listening and you like what
• 26:34
  we're doing, please give us a five star review on your
• 26:36
  platform of choice to all of you do and justice out there.
• 26:40
  Happy hunting.