Latest / Technically Sound Strategy / Episode 49 - The AI-Augmented Law Firm With Eric Romano

Transcript

Episode 49 - The AI-Augmented Law Firm With Eric Romano · Technically Sound Strategy

0:02
Welcome back to Legal Marketing Radio.
0:04
I'm your host, Chip LaFleur.
0:07
Today I'm joined by Eric Romano, a board certified trial attorney with Romano
0:11
Law Group in West Palm Beach, Florida.
0:14
Eric has decades of courtroom experience and has tried more than 125
0:17
civil and criminal cases to verdict.
0:21
In this episode, we're talking about how AI is being used in legal practice
0:25
today, what Eric has seen, firsthand is AI adoption accelerates in the
0:30
legal world and how law firms are approaching AI powered marketing tools.
0:35
Eric . It is great having you on here.
0:36
I always appreciate our conversations.
0:38
I know that you and your firm, you've been really active in the AI space.
0:42
You run a very effective practice.
0:44
And so I wanted to just take a couple minutes and hear
0:47
from you what you're seeing.
0:49
What you're implementing, what has your confidence.
0:52
And I'll give The floor over to you but tell me what your experience has been
0:54
now in these last six or eight months.
0:57
I'd tell you, I'll back up a little further than that.
0:59
I have always taken a keen interest in technology and as a firm, we've
1:03
always been forward thinking because I've learned lessons the hard way.
1:08
If you're not looking.
1:10
Past the horizon or around the corner to what's coming next, you're gonna
1:13
get left behind pretty quickly.
1:15
And I am by no means an expert in any technology.
1:19
I am a trial lawyer but as a trial lawyer and really regardless of what field
1:25
you're in you have to embrace technology.
1:27
And as we were talking here before we started just the rapid advancements in
1:32
technology that we've seen just in the last 10 to 20 years is fascinating.
1:37
And in the last one to two years now with ai it, it is mind blowing to me how
1:44
quickly that is evolving and adapting and improving, and it's now touching on.
1:52
Every aspect of our lives.
1:54
And 10 years ago, as a matter of fact, in 2016 I was asked to give
1:58
a talk to a group of trial lawyers.
1:59
And the topic I was given was, at that time, 2026 and beyond 10 ways your law
2:07
practice will change in the next 10 years.
2:10
And it was a really fun exercise for me to sit down and put
2:15
that presentation together.
2:17
'cause I really did have to kinda look 10 years into the future specific to the
2:21
practice of personal injury law and try to identify what appeared to be coming around
2:28
the corner and what I thought we needed to be ready for and what might be coming.
2:33
And I had an opportunity to revisit that presentation just about two months ago.
2:39
And here we are 10 years later.
2:41
It was fun to look back and see how right or wrong I was about those 10 predictions.
2:48
And I'll just touch on one.
2:49
At the time I had read a couple articles about the IBM Watson project,
2:55
and you know how that was the.
2:58
The birth of AI as we know it today, and I know there were many
3:01
other projects going on, but that seemed to be the most well known.
3:05
And I read an article at the time about how IBM had teamed up with a law firm in
3:09
Chicago to test what they, at the time called digital attorneys, where you could
3:14
go in and basically type some prompts into a computer asking the computer to.
3:20
Draft some initial documents.
3:23
And so that was one thing that really stood out to me at the time where I
3:27
said, look, one of the things we're gonna see over the next 10 years
3:30
is rapid advancements in digital attorneys that might start taking a
3:36
lot of that day-to-day to day research, drafting, writing, summarizing.
3:42
Away from lawyers and put it in the hands of computers . Yeah.
3:48
And so we've seen rapid advancements in that just, not only in the last
3:53
10 years, but the last 10 months.
3:56
It's crazy.
3:57
I mean, That timeline's wild because I have a not dissimilar story
4:01
that also has a landmark in 2016.
4:04
2016 was my 20 year high school reunion.
4:07
And ran Into one of my buddies from high school there that I
4:10
haven't really kept in touch with.
4:11
But started Chatting with his wife and she worked for Microsoft and she said.
4:15
Hey, you really have to start looking into predictive analytics
4:18
and machine learning and ai.
4:21
And I left there and I thought this stuff's, I'm a pretty tech
4:25
savvy person and I love this stuff.
4:27
And I started to dig in then and we ended up building a predictive
4:31
model around a PI practice.
4:33
We took about 40,000 records.
4:35
Of intakes.
4:36
And then we built a model at that time that said, thumbs up, thumbs down.
4:40
This would be a pain case.
4:41
And that Was 16, 17, 18, somewhere around those couple years.
4:44
We built it after having that conversation.
4:47
And then I got so busy that I just shelved it and I didn't do anything with
4:50
it, and then life went on and my main business got really busy and we started
4:55
to take on some clients and then I look at some of those systems that we built
4:58
at the time and we were getting a 95% accuracy rate . If it had a level of
5:01
confidence with the first kind of intake, questions that the person would answer.
5:04
But I thought, how Is this gonna be practical?
5:06
How do you put this into a format where.
5:08
Intake staff can see it, it can give 'em that green light.
5:11
And the Hoops were plentiful, right?
5:13
because it was dependent on the case management system.
5:16
You'd have to plug in and install that.
5:18
And now you look today and you think, man, like that is so easily accessible
5:23
because all the parts that you'd build to put it together are accessible.
5:27
And that's never been the case before.
5:28
So I do think it's wild to think 10 years back.
5:31
And what predictions, so what were the predictions that were right and
5:34
what were the ones that were wrong?
5:36
I, I don't think any of them were necessarily mind blowing because they
5:40
were at the time, we were at the, I think the very beginning stages of
5:45
some of these things that we've seen really develop over the last 10 years.
5:48
But I do recall.
5:49
I had predicted that we would see significant improvements in
5:54
self-driving or autonomous vehicles.
5:56
That of course would change the the way we look at and assess cases.
6:00
Because one the data shows that with the enhancements in autonomous
6:07
driving, self-driving features, there are fewer accidents.
6:13
Because a lot of this technology built into the cars is helping prevent
6:16
accidents, but the accidents that do happen tend to be much more severe.
6:20
And it is turning traditional negligence cases oftentimes
6:25
into product liability cases.
6:27
Because if you take, for example, Waymo, a driverless car, if a Waymo
6:32
car runs over a pedestrian, there's no driver behind the wheel to blame.
6:38
It's a company that built and designed the car and the software that is to blame.
6:42
So whereas in the past, that would've been a traditional negligence
6:47
case against an individual.
6:48
Now it's a product liability case against a company or.
6:52
A number of companies.
6:54
So that's one big shift we've seen and that's gonna continue
6:59
heading in that direction.
7:00
I, I really do think, and I joke with my kids about this, I say,
7:03
look, in 10 years, cars aren't gonna come with steering wheels anymore.
7:06
Now I, whether 10 years is too soon, I don't know, but
7:12
I expect that in my lifetime.
7:15
Most cars will not have steering wheels and they're going to drive themselves.
7:21
That I think is going to be a big shift, not just in terms of the
7:25
practice of law and liability claims.
7:27
But it's gonna affect, every aspect of our lives.
7:30
AI touched on that a little bit.
7:32
At the time it was referred to as a digital attorney that was just being
7:36
tested in one law firm and at the time, I think in a couple of small medical
7:41
practices and now it's everywhere.
7:44
I'm trying to think now, some of the others.
7:46
Oh.
7:47
This concept of the internet of things, right?
7:50
Everything's connected.
7:51
That now, and we're seeing it in all of our cases, there are discovery
7:55
requests now for your Apple Watch data, your Fitbit data, your aura ring data.
8:00
To help prove or disprove injury claims to help prove or disprove.
8:06
CBI brain injury claims.
8:08
So the notion that everything is connected, everything is constantly
8:14
collecting data, and that data now is being used by us or against us,
8:21
right
8:22
in, in litigation that's gonna keep trending in the same direction.
8:27
You can't avoid the data anymore.
8:29
I, maybe it was a little slow on this.
8:32
I predicted at the time that we would start seeing a lot more drone liability.
8:36
10 years ago.
8:37
Drones were still relatively new.
8:40
Yeah.
8:40
They're more commonplace now.
8:41
But I again think that in the very near future we're gonna start seeing a
8:46
lot more drones in the sky, not just.
8:49
The video Drones like we're seeing now.
8:51
But delivery drones and very soon I think we're gonna start
8:56
seeing, passenger drones.
8:58
Yeah.
8:58
There's a company that recently announced here, they're hoping within the next
9:01
two years to be offering passenger drone service between West Palm Beach and Miami.
9:08
That's crazy.
9:10
Yeah.
9:11
And you Start putting drones in the skies.
9:14
Particularly overpopulated Areas.
9:18
You don't have to be a genius to understand that is going to lead
9:22
to some accidents and problems.
9:25
And so I think a whole new field of drone liability type claims.
9:32
No, that makes sense, man.
9:33
That's interesting.
9:34
I think that the those emerging claims are al have always been an area of
9:38
interest to me just because, obviously because of the work that I do.
9:42
On the marketing side in the past, the barrier to entry to go
9:46
into those spaces was always a little bit prohibitive, right?
9:49
And I've Worked on an, I've worked on all sorts of different things, but I
9:52
had a client of mine who wanted to go out to the Volkswagen emissions cases.
9:57
Oh yeah.
9:57
And we ended up going after 'em.
10:00
We put up a landing page because that's what you did at the time.
10:03
'cause it took work to get it built.
10:05
And then we ended up using.
10:07
A car dealership, basically mailing lists that were designed for car dealerships.
10:12
You could rent them and you could reach out to very specific, make model owners.
10:15
and so we went after that and we were successful.
10:17
We signed up a bunch of cases, but it was a, what's the word?
10:21
It was a very unique way of chasing that very specific user, right?
10:26
And so we always try to think like, how do we reach, how do we reach that person?
10:29
How do we reach that audience?
10:32
But now that's changed so much.
10:33
And now we can spin up a website in a couple hours, and so when I hear
10:38
you talking about drone liability.
10:41
The first thing I would do is I would spin up a, I'd go buy a URL
10:45
and I'd spin up a site and let it get older, because I think that age
10:48
is still gonna count for something.
10:50
But when you can do that for a couple dollars in a couple hours,
10:54
that's a whole different, that's a whole different ball game.
10:56
We used to have to build a whole WordPress site, take a couple
10:59
months, cost a lot of money.
11:01
And then launch it.
11:03
And then that's when the clock starts ticking.
11:05
Because, a big part of that is how old does Google see your domain is
11:09
and how long has it had content on it?
11:10
Not even actively engaged, but just content.
11:14
That's been a huge change in and of itself.
11:15
So if you're prediction's right, now's the time to go and lock down
11:19
those drone liability URLs and start pushing some content out there.
11:25
I will confess, we bought up a couple of URLs 10 years ago, just in the off chance
11:30
we were right about some of these things.
11:32
Oh, that's awesome.
11:33
That's awesome.
11:34
The next thing to do then is to quick spin up a site.
11:36
A lot of times when we spin something up fast, if we want to, what we'll
11:40
do instead of all the, instead of all the traditional ways, is.
11:44
We'll do basically what you and I are doing right here.
11:47
We'll have a conversation on teams or on Zoom, and we'll record it with my
11:52
little recording bot here that I use on Circle Back or just the teams iteration.
11:57
And I'll take that transcript and I'll put that into cursor.
12:01
And I'll tell it to build me a site with that purpose and
12:04
it will build the whole thing.
12:06
And then I just need to connect it to domain and launch it and
12:08
it goes live , which is a big change to how we've done that.
12:11
So if you want to if you wanna spin something up quick, just let me know.
12:14
We've got enough right here to do it.
12:15
Yeah.
12:16
You May recall this, but when you and I spoke a month or two back.
12:20
You taught me a couple of quick tips about using AI and creating your
12:25
own GPT to do some of these things.
12:28
Not just from a marketing standpoint, but for research and trying to help
12:32
with everyday tasks in the practice.
12:35
And it was great stuff.
12:36
I am still a novice at ai, but trying to learn every day because
12:41
it's here and if we learn to use it the right way, it can be a.
12:46
Game changing tool for the better to help us deliver better results for our clients.
12:51
Yeah.
12:52
So How are you using it?
12:53
What are the ways that you're seeing the best results and what
12:56
are you looking to explore next?
12:58
I, using It in some of the.
13:00
More common ways, I think with initial document drafting, trying to get those
13:04
rough drafts, which saves you maybe the first few hours of starting from scratch.
13:09
So you get a good working draft to, to build from.
13:13
That's been a big time saver.
13:15
I was working with it earlier this week.
13:16
We're preparing to take a number of depositions in the camp Lejeune
13:21
litigation and some other product cases we have coming up where we need to.
13:27
Summarize and review.
13:29
Voluminous amounts of reports, depositions, and other documents.
13:34
And using AI to run queries of those documents, for example, help us identify
13:41
inconsistencies in this expert's prior testimony, identify inconsistencies
13:47
in their reports identify, strengths and weaknesses in their opinions
13:52
or gaps in their methodology.
13:55
And it's great for summarizing, querying large batches of documents like that.
14:01
But what I have found is that to me, the greatest tool is, it's, is for
14:06
brainstorming and idea generation.
14:08
As I've said, to me, AI is kinda like.
14:12
It's almost like having an ex access to an expert in anything you want.
14:16
And you basically just tell ai, here is what I'm trying to accomplish.
14:21
Here's the role I want you to play.
14:23
And then I will often just do some back and forth for, half hour or so on a
14:28
particular issue or topic and just start getting ideas and feedback to build from.
14:38
And it's very helpful in that regard.
14:40
Oh, that's cool.
14:41
That's cool.
14:42
Have you played with, have you played with Claude's Cowork yet?
14:46
I have not.
14:47
So you Might find that one interesting.
14:49
So I think when we talked before, I know we covered a couple of different
14:52
things and we I demoed a couple things.
14:54
I think I might have demoed one of our rag implementations, right?
14:56
Where we put a bunch of files into a spot and then you could query those files.
15:01
And it sounds like you, you're doing you're doing that in some different ways
15:03
with some depos and stuff like that.
15:05
Claude, so Claude Code is writing a ton of code now, so developers are
15:11
not writing code for the most part.
15:14
Developers have moved from writing code to using Claude Code or something like that.
15:17
I, personally, I use.
15:18
Some cloud models.
15:19
I usually use Cursor , but I'll switch back and forth
15:21
between that and Cloud code.
15:22
But cloud Code is designed and marketed towards developers so that they can write
15:27
and create stuff, but more and more it's bleeding over because there had been
15:32
bleeding over because, I'm not a developer necessarily, but I can use cloud code to,
15:36
for example, take the notes from this call and I turn them into a website that has
15:41
great information hierarchy where we can.
15:43
See where we were going and we can group topic, we talk about things,
15:46
we jump all over the place, right?
15:48
But if you wanna pull together and good AI note takers do this too, right?
15:51
They pull the ideas together and give you, here's the summary.
15:55
But what I like about about this approach is not only do I get
16:00
the text, but I get a design.
16:02
I get a design that's has a better hierarchy, right?
16:05
Text.
16:05
We have H ones and H twos and, different heading sizes and.
16:09
That helps us navigate a little bit, but it doesn't have an index necessarily.
16:12
So I've been using Cloud code and I've been using cursor to build those out.
16:15
But now Cloud came out with cowork and Cowork is like getting rid of all
16:21
of the more technical side of all of that and using cloud code to build
16:26
things for more of a business case.
16:30
So I just started using that and what you can do with that is you
16:33
can give it a folder and say, Hey.
16:36
I'm trying to do this with this stuff.
16:38
And it will start to do it for you and it will start to, to give you the next steps.
16:43
It'll, for example I had something I was working on today and I said,
16:47
Hey, I wanna write a response to this, I wanna send it in the mail.
16:51
And it created a Word doc for me.
16:53
It put the response in the word doc, finalized the letter.
16:57
All I had to do was open it and print it.
17:00
So it's
17:00
incredible.
17:01
It's wild to see.
17:02
So it Might be something that you might want to check out because it
17:05
just, it takes what you're doing and then it takes it, and that's where AI
17:10
continually blows my mind is when it's moving to the next stage age instead of
17:16
just giving it a task, task response.
17:18
Task response.
17:20
Now it's taking it to.
17:22
What are you actually trying to do here, and how do I get closer to that
17:25
deliverable, which is a big change,
17:29
right?
17:29
When I first started experimenting, it was exactly that, yeah.
17:33
Prompt response.
17:34
Prompt response.
17:35
And now some of the more advanced tools that I've been seeing and I've not spent
17:40
a lot of time with Claude, I need to.
17:42
Because I've heard good things.
17:43
But some of the other tools I've worked with, it will give you a
17:46
response and then it is suggesting.
17:48
Follow up prompts, which are far better than anything I
17:51
could come up with on my own.
17:53
Yeah.
17:53
But it's really phenomenal technology.
17:57
We just gotta learn to use it the right way.
18:00
So this is the thing that I think every business owner
18:03
leadership always struggles with.
18:05
How are you?
18:06
How are you helping the rest of your team use it?
18:08
How are you doing that in a way that's responsible with some guardrails?
18:14
First of all, any change is going to face some natural resistance.
18:18
We saw it years ago when we switched our case management system after 15 years.
18:23
And we've seen it more recently when we started to implement
18:26
AI into our workflows.
18:28
There are some people that just naturally.
18:32
Like things the way they are.
18:34
And as I've explained it to our team, AI can be a huge time saving tool.
18:44
And a tool to help us improve the quality of our work and deliver
18:48
better results to our clients.
18:50
So we all need to learn how to use it, and we are going
18:53
to work it into our processes.
18:56
And for those, and I've heard this point blank from some people, not just
19:00
in my office, but others, it just, it takes too much time to learn and I.
19:08
The way I've been doing things for 25 years has worked just fine.
19:11
I don't need to change.
19:13
And to me the, complaint that it takes too much time to learn AI to use it.
19:22
It's kinda saying, I don't have time to put gas in my car.
19:27
I'm just gonna walk where I need to go.
19:29
Yeah, you putting gas in your car takes a little bit of time, but it
19:33
will ultimately get what you're, where you're going much faster.
19:37
I think AI is the same way.
19:39
Yeah.
19:39
Put in the time on the front end.
19:41
Learn it.
19:43
And it will get where you need to go much faster and probably much better
19:48
because it, it's a second set of eyes.
19:50
We've heard the horror stories here about lawyers being lazy, just printing
19:54
out what ai spits out and filing it with the court without checking anything.
19:59
That was never a good practice anyway.
20:03
And now whether.
20:06
AI is the starting point, and then I review AI's work to edit
20:11
it, improve it, and finalize it.
20:13
Or the other way around.
20:15
Sometimes I'll draft something from scratch and then put it into AI
20:20
and say, help me make this better.
20:22
Yeah.
20:23
So it's, it's like having a an assistant, an expert, a teammate
20:27
right there with you to help.
20:30
Yeah.
20:31
Yeah.
20:31
No, it makes sense.
20:32
And I Like where you're going with it.
20:34
It's something that I think that we all kind of struggle with, right?
20:36
Is, two, two parts.
20:38
One is I've gotta learn something new, right?
20:42
And that's hard.
20:43
If we like the way that things are, then we have to change
20:45
it and it can be difficult.
20:47
But then two, from a leadership role, how do you get people to change?
20:50
And it Sounds like you're navigating that.
20:52
It's always got some ups and downs, but it sounds like you're having
20:54
a you're making good progress.
20:56
Yeah.
20:56
I think it's a matter of we've got very good open communication here.
21:00
Yeah.
21:00
We let everybody know that this is a tool that will help them
21:03
do their jobs better and more efficiently, which I firmly believe.
21:07
Yeah.
21:08
And I think sometimes some employees just need.
21:12
To see it in action.
21:13
Before they really come around and believe in it.
21:15
I just happen to be somebody that always loves learning new things.
21:19
Yeah.
21:19
And particularly when I see a technology or a product or a service
21:23
that I think can make life better, either outside the office Yeah.
21:29
Or help me be a better lawyer.
21:32
Yeah.
21:33
And better serve my clients.
21:34
I'm all for it.
21:36
Yes.
21:36
It takes time on the front end to learn something new.
21:40
I happen to really enjoy that.
21:43
Yeah.
21:44
And I, frankly, I think most professionals who care about their
21:48
work, their business, their customers or clients also feel the same way.
21:53
You love learning new things.
21:54
And you're always looking for for ways you can improve.
21:58
Yeah.
21:59
Yeah.
22:00
I think some of it, there's a little bit of the Dunning Kruger effect too, right?
22:02
You touch it.
22:03
And it works and everything's easy and it's simple.
22:06
And you think, oh man, I love this.
22:07
This is great.
22:08
And then you drop down the hill a little bit and you realize there
22:12
is so much here to learn and it's going to take so much work.
22:15
But then that becomes the fulfilling part, right?
22:18
Like actually doing the work to learn how to do it to, it was
22:22
my, my, my response was similar.
22:24
I was resistant for I. Feels like a week, but it was probably a couple months.
22:27
And I was just like, do I need to change all of the ways that we've done things?
22:31
And then I realized that this, that's what I always do by myself.
22:34
That's what I do.
22:35
But that's wired into people differently.
22:37
I think that, leadership roles, entrepreneurial folks, they
22:41
tend to chase after that they're more willing to accept risk.
22:44
They're more willing to go put themselves out there and make themselves look a
22:48
little silly as they figure stuff out.
22:50
And then you also have people who just like things in the normal status quo.
22:53
And so I know that's, we've tried to, we've tried to be mindful of
22:56
that, but also still drive people to, Hey look, we need to use this
22:59
tool and here's how we need to do it.
23:01
So it sounds like you do a good job of that.
23:02
I certainly trying.
23:04
Yeah.
23:04
That's the best we can do.
23:05
That's the best we can do.
23:06
I appreciate your time, Eric.
23:07
Like I, I think our time here's coming to close.
23:09
Anything you wanna leave the folks you're listening off with
23:12
. No.
23:12
Well, Chip, always good to catch up with you and thanks for having me.
23:15
This is always a productive and educational conversation.
23:19
No I probably just leave with where I had just ended.
23:22
It's never stop learning and I will never be an expert in everything.
23:28
Ai but that's not gonna deter me from trying to learn as much as I can about it.
23:34
Cool, man.
23:35
Thank you so much, Eric.
23:35
I appreciate it.
23:36
It's always great to see you.
23:38
Eric, thank you again for joining us and sharing your perspective
23:41
on how AI is being used in legal practice and legal marketing.
23:44
If you'd like to learn more about Eric and his work, you can find him through Romano
23:48
Law Group or connect with him on LinkedIn.
23:51
Thanks for listening to Legal Marketing Radio.
23:53
I'm Chip LaFleur.
23:53
We'll see you next time.