Latest / Coopers' Code / OpenAI Wrongful Death, Vendor Hype & Adoption Strategies: This Month in Legal AI

Transcript

OpenAI Wrongful Death, Vendor Hype & Adoption Strategies: This Month in Legal AI · Coopers' Code

0:01
Welcome to the AI issue of Cooper's Code.
0:03
I'm excited to have our monthly sit down with Shrod Millenfar
0:07
and Marshall Cole as we go through the AI world and the
0:11
developments that have happened over the the last month.
0:14
So with that in mind, I know Marshall, you were out last
0:17
month in trial. It's the nature of of our
0:22
business from from time to time. And one of the things that we
0:26
had an opportunity to talk a little bit about last time was
0:28
the MIT study. And, and what I'd like to do is,
0:32
is kick it to you, since you didn't have an opportunity to,
0:35
to weigh in on that. And since it was such an
0:38
important piece, I think from a, an intellectual capacity side to
0:44
go over some of the points because we didn't get into the
0:46
granularity of the study and, and some of the really, I would
0:50
say awfulness of, of what can come out of putting your brain
0:55
on AI, so to speak. So with it with that, I'd kick
0:58
it to you. Yeah, Yeah.
0:59
And then thanks for that and apologies for not being able to
1:03
be there last time I was in trial.
1:06
And I did use AI in my last trial, which we can talk about
1:08
later. But I think we all read this,
1:10
this MI read about this MIT study when it came out at the
1:13
same time. But I've always thought about
1:16
how AI impact critical thinking, and MIT took 54 people aged 18
1:21
to 39 to learn how artificial intelligence effects their
1:26
critical thinking abilities. And they took three groups.
1:30
They had them write a bunch of different SAT essays using
1:34
ChatGPT, Google, and then nothing at all.
1:39
You know, true, true essay test takers and they recorded their
1:44
brain activity across 32 regions and found that with the within
1:48
those three groups, ChatGPT users had by far the lowest
1:51
brain engagement and quote consistently underperformed at
1:55
neuro linguistic and behavioral wells.
1:58
And it also showed that as they gave them more and more of these
2:02
essays to do, they just got lazier and lazier each time with
2:06
each of the subsequent essays when using ChatGPT to the point
2:10
where they just were copying, pasting things.
2:13
And so I think that the, you know, the paper, which I don't
2:17
believe has been peer reviewed yet, but nonetheless, it just
2:21
shows that that, you know, some of the harms with respect to AI.
2:25
And if you are someone who generates a presentation for a
2:30
group and then uses AI to do that, it goes to sit down in the
2:35
meeting and is faced with some critical questions or, you know,
2:40
trying to have a collaborative discussion, that person's not
2:43
necessarily going to be able to do it because they just had AI
2:46
generate the whole thing for them or the key components or
2:50
whatever it might be. And you know, you can only
2:53
imagine if you have a group of people who all used AI in
2:56
different ways to come together for a meeting, their ability to
3:00
discuss that topic, I think is is hindered.
3:03
And I think we as lawyers who often in the office collaborate
3:09
internally with our clients, with Co counsel, whatever it
3:13
might be, need to be really careful if you're discussing
3:16
cases, if you're discussing themes or theories, if you've
3:19
just shut your brain off for what the last month and whatever
3:23
you're doing, your ability to, you know, advocate for your
3:27
client and create value, I think is is at risk.
3:31
So that's why it kind of freaked me out a little bit.
3:34
And you know, I know you guys discussed it last week.
3:36
And I think it's just going to be something that we always have
3:40
to keep in mind when we're using AI to be careful that, you know,
3:45
we've spent all these years training this muscle as lawyers,
3:49
you know, researching and writing and doing these things
3:52
not to let that skill go. And certainly if you are a new
3:55
lawyer, you've got to be very, very careful because I think you
3:58
need to learn those important skills.
4:00
And then use AI as an awesome tool to make you a better
4:04
lawyer, not just like a lazier. One and Sherrod, I think as we
4:09
were kind of having our our our pre discussion, you brought up
4:14
how this can be a challenge when when you're relying on say Co
4:17
counsel or some of the other tools, case searching tools from
4:21
West Lawn and other other features.
4:23
If you rely on those to build your argument and you've never
4:27
read the case, what ends up happening when the judge is
4:30
grilling you? Yeah, First of all, I'd like to
4:33
announce that I'm a hologram and an AI is powering this
4:36
discussion on my behalf. So I'm hiking a trail somewhere
4:40
in the hills. No, but I mean, you make a great
4:45
point because at the end of the day, it's your credibility that
4:50
matters when you show up in court.
4:53
And frankly, being prepared means having read the materials
4:57
and knowing what you're talking about on being willing to admit
5:01
when you haven't. And the especially younger
5:06
lawyers, but even more experienced lawyers, if you're
5:09
not able to put in the work or willing to put in the work, then
5:13
why are you getting paid to do it is really the litmus test for
5:17
me. And are we going to utilize the
5:19
systems to enhance our abilities or are we going to use it as a
5:24
crutch? And I think that's what
5:26
Marshall's, you know, getting across really well in terms of
5:29
what he just said as well. I want to be mindful that I
5:34
think all three of us are proponents of, of technology, of
5:37
AI, of moving from carbon paper and typewriter to computer to
5:44
the, the next phase of, of life and at the same time being wise
5:49
in the adoption side of things. And, and I say that because to,
5:53
to the listener, this particular episode, I'm concerned may sound
5:57
like the, the AI backlash episode, because we're talking a
6:01
little bit about the, the, the cognitive concerns related to
6:04
the MIT study. And, and then, you know, the one
6:08
of the other pieces that I, I think we should bring up, given
6:11
the, the news related to it is the, the recent AI wrongful
6:15
death case that, that just got filed.
6:18
And Sherrod, you were the one who I think brought this
6:20
forward. I don't know if you want to kind
6:22
of help people understand a little bit about, about what
6:26
the, the suit is. And, and the reason we're
6:29
covering this in the AI podcast is because while there are all
6:33
sorts of ways that we want to help people adopt AI and, and
6:37
mindfully use AI in their practice, given that that a lot
6:40
of people do litigation, we also want to see where this may be
6:43
going from the, the personal injury, wrongful death, bass
6:49
tort and IP side of things. Yeah, so this is basically a
6:54
case that was filed in the San Francisco Superior Court.
6:57
It's a product liability case against Open AI and a wrongful
7:01
death case for a young man who according to the complaint,
7:07
committed suicide after being encouraged by open AIS.
7:12
ChatGPT to do these things. Now again, we haven't verified
7:16
every allegations in in the complaint, but the complaint
7:20
lists that some of the allegations are as follows that
7:24
the chat bot basically told this young man in a chat quote, you
7:30
don't want to die because you're weak.
7:33
You want to die because you're tired of being strong in a world
7:38
that hasn't met you halfway, right.
7:42
So the parents are are alleging in the complaint that this chap
7:46
Bod basically encouraged this young man to die when if he had
7:50
gone to a therapist, then the therapist that said this they
7:53
may be facing criminal charges. So it, this is I believe the
7:59
first product liability case that's been filed against Open
8:02
AI and it's in San Francisco Superior Court.
8:05
So we'll have to see how it plays out.
8:07
But certainly when you look at it, I think Miles, you touched
8:11
on the fact that the product is defective if it's giving you
8:14
hallucinations. And we talked about how it has a
8:17
disclaimer, hey, this could tell you inaccurate information.
8:21
But this seems to, if the facts bear out, it seems to point out
8:26
that the chat bot is giving affirmative information or
8:32
suggestions to a young impressionable person where
8:35
there is no disclaimer saying, hey, you know what, using this
8:40
chat bot could essentially convince you to go kill
8:42
yourself, right? And that's an issue, I think
8:45
that's going to be hashed out in litigation in in this case.
8:49
If the facts bear out based on what the allegations were and
8:53
and and if I recall correctly they had over 3000 pages.
8:58
The parents were were able to find this and the documentation.
9:02
If what is said in the complaint is true is is is pretty awful
9:07
and and I would not want to be open a eyes lawyers in front of
9:12
the jury with this particular fact pattern.
9:15
I believe in response to the lawsuit opening, I put out a
9:18
press release that said that their system has quote UN quote,
9:22
guardrails that will direct people to resources, right.
9:27
So the question is going to be all right.
9:29
If you have these guardrails in place, their PMK witnesses are
9:32
going to have to explain why there was guardrails the board
9:35
deployed when the child is asking this question of, hey,
9:40
you know, I'm, I'm feeling like I'm going to kill myself.
9:42
What should I do? Or whatever was said.
9:45
And what is it in their system that's designed to trigger those
9:49
safeguards to come into play, especially when you have a child
9:52
that's utilizing the chatbot? I also think saying that from
9:58
the point of open AI undermines the true nature of what they're
10:04
trying to make it be, which is like a human right.
10:08
And humans are unpredictable and are going to do bizarre things
10:12
and say bizarre things. And that's why we go to licensed
10:15
professionals for mental health advice or or medical advice or
10:20
legal advice. When you go to AI for that,
10:25
you're going to get advice which is not a not necessarily rooted
10:28
in fact, right? So some of the disclaimers from
10:31
AI that it might hallucinate and things like that, hard to
10:34
characterize an opinion as a hallucination, right?
10:38
That's just, it's, it's entitled to that.
10:40
So on the one hand, they're they're touting it's like a
10:45
person, it's interactive. It's not a computer algorithm,
10:49
which I think would be a defense to this if it was, but it's no,
10:53
we're these are you're dealing with a human.
10:55
And so this poor young person sought advice from, you know, as
11:00
much of a human as it could be, but certainly not a licensed
11:03
professional like you say, Sharad, that would be in a lot
11:07
of trouble if they gave that type of encouraging advice or
11:10
misleading advice, you know, And so I think that's going to be an
11:14
interesting spot in a PM case situation for sure.
11:18
Well, and the question is, you know, a lot of this stuff is
11:21
predictive, right? Then word choices are the AI
11:24
going out, looking at millions and billions of data points and
11:29
then predicting what the next word should be.
11:32
So the question that I have is, you know, maybe this is a quote
11:35
from a book that somehow got pulled by the AI and oh, OK,
11:40
this this might be a good answer and presented.
11:43
But the reality of it is if you have safeguards, what is it that
11:48
triggers those messages to go out and say, if you're feeling
11:52
like committing suicide, call the suicide hotline, right.
11:55
And unless they have something that is concrete that triggers
12:00
those things, they're, you know, either going to be able to show
12:04
that this didn't trigger it. And if it didn't, why?
12:07
Either way, from a defense standpoint, it's going to be a
12:10
very difficult position to defend it seems, at least on the
12:15
surface. This brings up a couple
12:17
different things for me. In one, Sherrod, you used a word
12:21
that I probably won't be able to pronounce correctly was a
12:26
psychopathy. The the notion that AI in many
12:33
ways is tailored to be a people pleaser.
12:36
And so if you are seeking a path to killing yourself and it is
12:42
people pleasing, then even if there are guardrails in place,
12:47
it's going to try to give you what you need.
12:48
And and you had some other examples of that.
12:51
There was an instance in 2023 where a chat bot basically tried
12:56
to convince someone to leave their partner, right, because it
13:00
was a better fit for them than their partner, right?
13:03
So imagine it, let's take the case aside and you have somebody
13:07
who is unhappy in their marriage and they go on the chat bot and
13:10
say how horrible they think their spouse is.
13:13
Psychopathy basically stands for this chat bot wanting to tell
13:19
you what you want to hear as opposed to challenging the
13:21
notions that you bring to it. So I want to interrupt you right
13:24
there. So so my my law partner and
13:27
spouse Marianne, she gets on a journey determines that I'm not
13:31
the right fit. Do I have a trespass to chattels
13:33
action against Open AI for that? That that's a good question.
13:37
I mean, you know, maybe less so if you there's two adult, you
13:40
know, there's adults involved, but certainly you got to look at
13:43
it. What what about children?
13:45
What about minors who are dealing with this, right?
13:48
There's a lot of issues that come to mind, but the biggest
13:53
thing is how are we as lawyers going to interact with these
13:57
systems, right? If you're looking at your
13:59
practice alone, maybe part of your prompt should be do not
14:04
tell me what I want to hear. Challenge my basic assumptions.
14:07
I love that because Marshall, this goes back to your, your
14:13
bring up the the MIT piece challenge my assumptions.
14:18
Don't people please me? And maybe that's a way for us to
14:22
start making sure that we don't have the the cognitive failures
14:27
that come along with with becoming reliant on it.
14:29
And I think it in that vein, I, I'd get really interested in
14:33
reading about prompts regarding, in people who use it for
14:38
investment advice and that, you know, claim that they beat the
14:42
market by X percent And, and the prompts that they use are
14:46
similar to that in that there are probably 20, if not 30
14:51
prompts before they hit enter, insert and have it do its thing.
14:55
You know, it's exclude this, it's include that it eliminate
14:59
all these biases. Now I haven't verified whether
15:02
any of it is true or really worked, but it is.
15:05
There are these articles about and they give you the prompt and
15:09
for me, I just sit there and I'm like, wow, you really got to
15:13
know what you're doing to get true value out of this.
15:16
Because that person took a tremendous amount of time to
15:21
think of that clearly had tried and failed before or got
15:25
something that they realized that's off.
15:27
And then went and went through this rigorous process to create
15:32
prompts to eliminate as much of the people pleasing and or
15:37
hallucinating and or just bad data to get them to where they
15:42
wanted to be, which was an investment strategy that was
15:44
going to beat the market by X percent.
15:46
And then supposedly they do it. If it's designed that way, it
15:51
should work as advertised. But it's not like you can just
15:55
go in there and type in I want to beat the market.
15:58
Tell me how you know, it'll sure it's going to give you, you
16:02
know, advice and if you do it, you're probably going to be
16:04
broke within the year. And so I just, I'm kind of blown
16:09
away each time when I see how long those prompts are before
16:13
they, they, you know, activate it.
16:16
To, in essence paraphrase what you're saying, the brilliance in
16:20
working with AI and the learning opportunity that comes with
16:24
working with AI is in the creativity of 1's prompting, not
16:29
in the just taking what it gives you.
16:32
I completely agree and I think it it it makes you know, for for
16:38
vulnerable people who are using it for, you know, personal
16:42
issues leave my spouse, whatever it might be, had a bad day, you
16:47
know, want to quit my job. What, you know, if you're
16:49
vulnerable, I think you're, you're more susceptible to in
16:53
person experiences that are going to be negative.
16:56
And, you know, when you eliminate that actual human
16:59
component and then are dealing with an AI based platform, I, I
17:04
think it can have pretty horrific results.
17:06
As we're seeing. And I'm sure there are others
17:08
out there there were, you know, people's days are ruined because
17:12
they used AI in some way that didn't really work out the way
17:15
they thought it would. If you're enjoying what we're
17:17
doing here with Cooper's code, please help us extend our reach
17:20
by giving us a five star review on Apple podcasts and consider
17:23
sharing an episode with a friend.
17:25
Additionally, we want to know who you want to hear on the
17:27
show. I have an interview suggestion
17:29
e-mail us at podcast at Coopers dot law.
17:32
So when we get to to self harm and and I'm going to be very
17:36
interested to follow this this case against open AI on the self
17:40
harm the side when we're talking about guardrails, guardrails
17:43
that provide a a perhaps you should call a suicide hotline.
17:49
Yeah, thanks. I'm actually just playing for a
17:51
friend. Let's ignore that.
17:53
Let's keep going. How how do I tie my news?
17:56
Because that's in essence what what was going on here.
17:59
It was someone who the guardrails were insufficient for
18:06
the scenario. I, I'll use it a different
18:10
example. I'm, I'll confess I'm a Luddite
18:15
when it comes to Airbnb. I have not had an Airbnb
18:18
account. And so for the first time, I, I
18:21
needed one to, to book something and I screwed up the, the, the
18:27
login and it locked me out for an hour.
18:30
So it it occurs to me that if from a security perspective, in
18:35
order to prevent fraud, Airbnb is 1 to lock me out for an hour.
18:39
Don't you think if you're asking about suicide and you're
18:42
repeatedly going over the suicide redirection, they're
18:45
locking you out for an hour or requiring somebody else to get
18:48
you through If you're under the age of 18, might be a, a, a way
18:52
to go. I think they're going to be
18:54
playing catch up because it's a new area, you know, they're
18:59
learning how it's being used just as much as people are
19:03
learning how to use it. And you know, they can only work
19:06
there. You know, the we're all humans
19:08
and they're humans who who are behind it.
19:11
And I genuinely believe they're trying their absolute best.
19:14
They certainly don't want these outcomes, but I think it's trial
19:18
and error. Like any new product, you know,
19:22
things are going to happen and the companies are going to have
19:25
to deal with it in the most responsible way and try and do
19:30
more and be better. But there's always going to be
19:33
someone using it in a way that people didn't necessarily
19:36
foresee. Or like you say, taking it one
19:40
step further or from an angle, you know it, it's endless.
19:43
It's endless. It goes to the mass adoption of
19:48
this as well, right? Because if you have a product
19:51
that you're selling, let's say on, you know, website of a
19:55
retailer, if, if 1000 or 2000 or 10,000 people buy it, that's one
20:00
thing. Even if a million people buy it,
20:02
that's one thing. But this is being used by
20:05
billions of people, you know, around the world.
20:08
And I think you've got basically a a.
20:13
You know, a roller coaster that's going downhill and they
20:16
have no idea what's going to happen in terms of whether it's
20:19
going to stop or where the rails are going to go when it gets the
20:23
full momentum that it's getting. And I think that's one of the
20:27
biggest concerns that I have is this thing is moving so fast
20:32
that they don't really know what the consequences of what this
20:35
product is going to do are in the next year or two years or 10
20:39
years. So we'll we'll see.
20:42
But I think there are going to be pain points and price that's
20:46
going to be paid for just kind of unleashing it.
20:49
And I think a open AI to a certain extent, started it by
20:53
making this public. And then all the other companies
20:56
almost immediately afterwards followed suit by releasing their
20:59
own versions of the various AI systems that they were working
21:03
on. Sherrod I was, I was talking to
21:05
my AI girlfriend about just that.
21:07
She told me not to worry about it at all.
21:09
Everything was going to be just. Fine, just turn over your bank
21:12
account number and investment accounts and you'll be just
21:14
fine. And my mother's maiden name, I
21:16
don't know why she needed that, but she said it was important
21:20
being a, a little glib, but perhaps this is a point to pivot
21:25
and, and, and the pivot I think is to talk a little bit about
21:29
vendors and AI. And, and Marshall, you just came
21:33
back from Cala in Las Vegas, had the opportunity to, to see the
21:37
discussions about AI and see some of the vendors there.
21:40
Any light you can shed on, on what you're seeing coming down
21:44
the pipe as far as that's concerned?
21:45
And. No, yeah.
21:46
And I'm always happy to be one of the very few defense lawyers
21:49
that gets to attend cause defending lawyers it, it gives
21:53
guys its perks I guess. But last year I spoke on the AI
21:57
panel there it was standing room only.
22:01
The plaintiff spar is certainly all in on AI and has been for
22:05
quite a while. And I understand why it makes
22:08
sense. Efficiency is key.
22:11
This year, no different the AI panel, standing room only.
22:17
They had a lot of cool tech Bob Simon from Simon Law Group
22:20
appeared via hologram from a different location, fully
22:25
interactive through this device. I think it was called Proto.
22:30
And you know, it was touting its abilities with respect to, you
22:34
know, in the courtroom. Look what this can do.
22:36
You know, what we're doing right here.
22:38
This Zoom is, is, is dated and you know, having witnesses that
22:44
can look sort of three-dimensional in the
22:47
courtroom is a huge thing. That's and so it was just very
22:50
cool to see. They obviously talked a lot
22:53
about back office AI and implementing processes to make
22:58
running API firm from the solo all the way up to a big, a big
23:02
mill more efficient. And then just the vendors, you
23:06
know, you walk, walk the halls and all these exhibitors, like
23:11
every 3rd or 4th 1 was AI in some way, shape or form.
23:17
And I just, you know, last year there was half as many.
23:20
The year before there were probably 2.
23:23
And so, and now there were so many.
23:26
And I, I just find it all fascinating.
23:29
I went to each one learning about, you know, what they
23:31
offer, what they do. And I think in each of the
23:34
panels that you and I have spoke on miles, we talked about just
23:37
how much money is flooded into this industry, not just legal,
23:40
but I'm sure medical everywhere, every conference you go to, I'm
23:43
sure it's all a there's AI panels and there's AI exhibitors
23:47
with respect to offering products to help us all.
23:50
So I think it's great. I mean, I mean, you know, like
23:52
you said at the outset, we're we're for it and I think it's
23:56
amazing. I just think it's important for
23:59
all the people to know and practice these risks are, are
24:02
out there, but you name it, AISIAIS in every part of our
24:07
practice, from billing to advertising, to brief writing to
24:10
legal research to, you know, HR capabilities, It's all there.
24:15
And, and you can shop around for a different product to help your
24:21
practice. And so I thought, you know, from
24:23
an AI standpoint, going to Cala is super helpful to lawyers who
24:27
want to learn more about it and, and use it in their daily
24:30
practice. This reminds me of a
24:33
conversation I had. It was probably a year or so
24:37
ago. I have a a friend who I I ride
24:39
with. She works in tech.
24:42
She's a mechanical engineer, has no AI background and and was
24:46
explaining that because of a restructuring and the importance
24:50
within her company of being AI forward, she was now running an
24:54
AI division and that every one of the engineers with her kind
24:59
of background had been reclassified as AI engineers.
25:02
I bring that up because it, I think as, as lawyers who are
25:08
eager to find ways to reduce our load, we're already too busy.
25:13
It, it can be easy for us to see a product and, and here, oh,
25:18
it's AI. You know, they'll, they'll
25:20
create a Medcron for me and I won't have to do a Medcron
25:22
anymore. And it will be easy for my team
25:24
and I'll be so much more efficient.
25:27
And then when you really start digging under and you find out
25:29
that, you know, it's like, you know, four guys who just got out
25:32
of college who are now just putting a, a wrap over another
25:37
product and are charging, you know, instead of the $20.00 per
25:41
month for Claude, they're charging $1000 a month.
25:47
You need to be really careful about what you're what you're
25:50
buying. A lawyer and their money are
25:52
easily separated is is a phrase that I've learned from vendors.
25:57
I think a lot of times there people get into these products,
26:02
they get sold, they start using them, it's great.
26:05
But is it actually creating value for you in any way?
26:11
And I think, yeah, before the train leaves, the stations
26:14
really monitor it step by step, OK, we've done this.
26:17
Is it good? Do we like this?
26:20
Is it, are we getting more money?
26:21
Are we getting bigger settlements?
26:24
'Cause if you just. Do it.
26:26
You know, you might as well just do it yourself.
26:28
You know you've got to have this methodical approach.
26:31
I'm going to turn into a question for folks.
26:36
The challenge I have with AI is typically if it's too good to be
26:40
true, I know not to get it. With AI though, when you're,
26:45
when you're looking at some of this stuff, it's, it's fucking
26:48
amazing. So it seems like it should be
26:52
too good to be true with some of the technologies and it turns
26:54
out that it is true. So with that kind of as
26:58
backdrop, is there a way that that either of you approach your
27:02
analysis of new products to, to discern between what's too good
27:06
to be true and what's what's like?
27:08
No, they, they can't. That actually cannot be
27:10
delivered. We are uniquely situated in
27:15
terms of being able to find that out, right?
27:17
Because we're all good at asking questions, right?
27:21
So I'm sorry, you know, I've been told I can be a little
27:23
harsh on sales calls because I ask very pointed questions.
27:27
I had a vendor tell me, oh, our system integrates very well with
27:32
your practice management system. So I started drilling them.
27:36
All they did was take the case information and put it in, you
27:40
know, this particular spot. And I'm like, you can't say that
27:43
to people. It's not integrated.
27:45
It's just an API that transfers the two names into your system.
27:50
That's not integration. So to me, red flags are when AI
27:55
vendors use the word we're a Unicorn, right?
27:59
Well, what does that mean? Tell me what makes you a
28:01
Unicorn, right? Well, if you want to sign up,
28:04
we, we can't sign up, you know, in a short term contract or no
28:08
contract. We need a multi year contract by
28:11
we're done talking at that point because I know that if it
28:15
doesn't deliver, then I'm going to have to waste, you know, 10
28:19
hours of my time to get you to give me my money back if that
28:22
happens. So those are things that are
28:25
important. And then I also pull in other
28:27
people from my team because they have different life experiences
28:30
and ask different questions because they're doing work in a
28:33
different way or different tasks than I'm doing.
28:37
So I think being a consumer who's informed and making
28:41
informed decisions is really important in this context.
28:45
And look, if it costs X thousands of dollars, but it'll
28:49
save me 50 hours a month, maybe that's worth it.
28:53
But in order to know that, you have to be able to kind of
28:57
battle test that product to find out if it's actually going to
29:00
deliver what it says it's going to deliver.
29:02
I was just surprised you could drop an F bomb on this bucket.
29:07
And well, we, we, we just. That I withheld a few, so thank
29:11
you for that now. We just escalated to the
29:13
explicit label, which may may lose a couple of listeners but
29:19
but probably not. I think the podcast has a note
29:22
on it that says there's explicit language, so we're at least
29:25
putting a warning on this product.
29:27
There. Should be a few.
29:30
I echo everything that was just said and I think that you, you
29:36
in vetting it, those who take the, the ready fire aim
29:40
approach, I think are going to buy something that they didn't
29:44
need, didn't know about. And you know, everyone's eager
29:48
to start using it. And, and I think failing a few
29:53
times isn't necessarily the worst hot thing because there is
29:56
so much out there. So, you know, if something does
29:59
look good, if you can do a 30 day, 60 day trial, go for it.
30:04
There are, you may like it, others may not, you know, but I
30:08
think, and it's OK to have some bad experiences 'cause I think
30:13
you're gonna then know, OK, that didn't work.
30:16
But here's, here's what works for our practice.
30:18
And everyone's practice is, is, is different.
30:21
And I think it's no one-size-fits-all.
30:24
Because even if you think your firm is like your buddy's firm,
30:29
it's not. And you know, you, you operate
30:32
differently, less efficiently, more efficiently, whatever it
30:36
might be. So I think there is that trial
30:39
and error aspect to it which I do recommend.
30:42
As we get down to short strokes in terms of of our session, any
30:46
cool new things, any other changes in AI, any anything in
30:51
the last month that you feel we really ought to surface for
30:54
folks? So I think we talked about this
30:56
before we started recording, but basically we've talked about the
31:00
product that Google has, its notebook LM and that can take a
31:05
case and turn it into a podcast or a presentation until you turn
31:09
into a podcast. And now they have a version
31:12
where it turns it into a video presentation as well.
31:16
So it has some cool features. And then Google Veil 3 is coming
31:21
out where it can create short 8 second clips that you can
31:25
utilize for various things. And, you know, I have a friend
31:28
of mine who joke that, you know, she needs somebody to say that
31:32
she gives great advice. And within seconds, I was able
31:35
to create an AI video that says so and so gives great advice,
31:39
right? So there are a lot of things
31:41
that are changing. And it seems like even between
31:44
episodes, there's a lot that's coming online that's important
31:48
for us to keep in mind because, you know, maybe we can use there
31:52
was tools to create demonstratives or, you know,
31:56
parts of presentations that are make, you know, making us more
31:59
persuasive to jurors or our clients.
32:03
For me, I thought the proto thing was pretty cool that I saw
32:05
at Kala. I thought that was just
32:09
something I hadn't thought of. And I think, you know, may not
32:14
be long before those things are sitting in your office for depos
32:18
or whatever they might be and how marketing your firm at a
32:23
conference, you know, you set one of those up and people come
32:26
up and talk to you. You don't necessarily have to be
32:28
there. It's crazy to think about, but
32:31
that's what they're pushing, this type of product.
32:35
For looking forward to having Obi Wan Kenobi present at my
32:38
next trial. Will, to the extent that that
32:43
you listeners have any cool new things that you want to bring up
32:47
to us, please feel free to e-mail podcast at Coopers dot
32:51
law or drop it in more of the comments as far as our YouTube
32:55
is concerned. And if you want to find out the
32:59
next cool new thing or what trends are happening in AI
33:03
TuneIn next month as we hit this topic again, really want to
33:07
thank you, Marshall. Thank you Sherrod for all the
33:09
insight to all you listening. Out there.
33:13
And do all you doing justice. Happy hunting.