What’s Next With AI for Law Firms?

Karin (00:01.534)
In 2022, Jessica Ares came on this show and we talked about what was next in digital marketing. And it’s kind of funny because four years ago we were talking about such different stuff. We were talking about the metaverse. We were talking about TikTok a lot. and it is a different world today. And so Jessica is one of our new Council cast collaborators, and we are going to talk about AI, but this is gonna be a different one.

Because I know you’ve heard a million podcasts about AI. And Jessica founded by Ares after a decade in legal marketing, business development, and digital strategy. She’s also a former lawyer. So she brings that nice kind of lens to all of this. And she’s on the board of the Legal Marketing Associates, Legal Marketing Association’s International Board, and trains all these legal marketing teams on the practical use of AI.

And this is her third time on the show. There was a whole other show I didn’t even mention. but what we want to talk about and focus on today is firms are spending all this money, time, effort on AI tools, AI courses, AI employees, but nobody can tell them which of them will still matter a year from now. So she knows she sees the inside from the firm’s side, from her side, from the marketing side. She knows where they go wrong on AI and what holds up when the tools are.

change because we all have seen they are changing like every five minutes. So we’re going to talk about what to build next, where things are going wrong. And before we jump in, if your firm has been buying AI tools and training and you’re not sure any of if any of it will matter next year, this is the kind of stuff we solve at Conroy Creative Council. So Jessica, welcome to Councilcast again. Thank you for being here. I love these conversations and I know you are in the thick of it.

Jessica Aries (01:48.861)
Yeah.

Jessica Aries (01:54.012)
Yes. Thank you for having me. Yes, I I am not just in the thick of it. I feel like I’m drowning in it. Like just in terms and I feel like everyone is right now with AI. Like I have not talked. I literally have a private text change with CMOs at some of the largest global firms in the world who are like, I feel behind. And I’m like, girlfriend, so do I. Boyfriend, so do I. Like we all feel that way right now. We are all our cuppeth or what our cup, our cup of like floweth over whatever that term is, right? Whatever. I’m totally butchering it right now.

Karin (01:59.628)
Yes. Yeah.

Karin (02:11.638)
Yes. Yes. Yes. Yes, yes.

Karin (02:21.44)
Yeah runneth runneth yeah just keep adding adding the T H at the end of all the words. Yeah. Yeah. Cupeth runneth over it. Yes. You know Yeah, but you know what? I think I I saw this recently and I think this is so true that what makes me frustrated about the fact that we it is running over, but it’s not running over in like a I feel so blessed and grateful. It’s like I’m

Jessica Aries (02:23.687)
Yeah, runeth over. Yeah, just Yeah. Just all the thus, right? It’s just everywhere, right? Like every yeah. I’m so yeah, so good at yeah, not not great at quoting. Famous quotes are just

Karin (02:50.89)
Stressed and behind is the word I keep hearing too. And I brought this up on another show recently. that I saw that AI has this marketing problem and it originally had kind of presented itself way back when we were little as the Jetsons, right? Like our lives were supposed to be the Jetsons right now, and they’re not. No, yes, right.

Jessica Aries (03:09.469)
Mm-hmm.

Jessica Aries (03:15.155)
It’s it’s still not doing my laundry. Even as much as I prompt Claude, it just doesn’t do it. It is doing my grocery shopping now, but not my laundry, and that is very infuriating.

Karin (03:22.104)
Same. Yes. Meal planning, grocery shopping, check. Yes. but no laundry and my car’s not flying yet. and I don’t have Rosie like to do all the stuff, right? Yes. And the other one that like angers me is the getting ready in the morning. You know, remember they had like that tube that they would stand in and then they would just spray all the stuff. I’m like,

Jessica Aries (03:25.933)
Ha ha ha.

Jessica Aries (03:31.89)
No.

Jessica Aries (03:35.783)
Y the housework?

Jessica Aries (03:50.545)
Yeah. Doesn’t exist. I still swi stare in my closet like, what am I wearing today? And it takes way too long.

Karin (03:51.234)
Where’s my tube? Yeah. So

Right. And why am I using my brain power for this? Like that, come on. Like someone else needs to do this for me. So this is where I feel like at the end of the day, where our frustration is lying with AI, we’re overwhelmed and it’s supposed to it was supposed to relieve all of this. Right. and so what so what what are we gonna do about all of this and how can we make this better?

Jessica Aries (03:59.796)
Mm-hmm, mm-hmm.

Yes.

Jessica Aries (04:14.035)
Mm-hmm.

Jessica Aries (04:23.795)
Well, I think the mistake we all made very early on was that we framed it from this idea of time saving. When the reality is measuring the ROI on AI is much more difficult than any of us anticipated. Measuring how much time it’s supposed to be saving us is very difficult when all of our people aren’t up to speed, all of our context layers a hot mess. Like if if you know anything about a context layer, like building your knowledge base, what you’re going to use AI to inform.

Karin (04:29.718)
Yes,

Karin (04:36.493)
Yeah.

Jessica Aries (04:53.159)
the AI to give you better outputs is a is a disaster. and we’re realizing we’ve we sold the wrong message, basically, that that really AI, I think, is about taking good work and making it better, right? It’s about taking our A team and giving them an A plus capabilities by utilizing

Karin (04:53.824)
Everything? Yeah. Right.

Karin (05:01.525)
Yes, yeah.

Jessica Aries (05:15.367)
you know, AI skills, better context layer, but you have to build the foundation first. And many people don’t want to build the foundation or didn’t realize they needed to build a foundation and aren’t prepared to build the foundation. So you know, I’ve I’ve been to countless conferences. I’m literally coming back from the Marketing AI Institute conference. I went to Iltacon in August and I sat there and listened to all these brilliant, smart people be like, there is no ROI on AI. Like, like and that’s a hard thing to hear when your firm just invested if you’re like a Kirkland

Karin (05:39.737)
my gosh. Yes. Yeah. Right. Right.

Jessica Aries (05:44.284)
Right. Five hundred million dollars in it, right? That that finding that number, figuring out what that is, it’s it’s not an easy task and that it’s not something that we’re gonna find tomorrow. And so the way we need to think about AI needs to shift from speed to leveling up, I think, to leveling up what we do and and making us better marketers.

Karin (06:03.488)
Yeah. Yeah. Well, and so this is where I get frustrated with people trying to measure the ROI of any marketing because there are so many things you just can’t measure. And nobody likes to hear that. And so I literally just saw a post on LinkedIn where they were talking about all, you know, once again, it had a chart of like all the marketing activities that they s saw were measurable. So it had like Reddit,

you know, whatever Google ad, you’re like all these things. And I’m like, okay, first of all, this is only one tiny slice of a marketing, you know, activity, like a full marketing campaign and everything that you should plan for marketing. And a huge chunk of it you really can’t measure. How are you gonna measure an event? And like, yeah, maybe you get, you know, a case or whatever, but maybe five years from now,

Jessica Aries (06:40.413)
Yeah. Mm-hmm.

Jessica Aries (06:57.777)
Yes.

Karin (06:57.868)
That person thinks from you about you and sends you something. Are you gonna go back to that event five years before and somehow tie the like sixty dollar entrance fee to to what? Like these numbers don’t make sense. It’s like numbers and colors. Like you’re trying to line them up, right?

Jessica Aries (07:16.797)
Well, and I think on top of that, there’s a big demand from at least with the lawyers I’m working with who are, you know, tend to be B2B firms. their clients are looking for the lawyers to be using AI to be more efficient because they’re still being billed by the hour, right? And the clients are saying, We have a legal budget of X and we would like to reduce that. And now we’re all using AI. Are you using AI, right? So it’s also

Karin (07:32.716)
Right, right.

Karin (07:37.826)
Yes.

Jessica Aries (07:41.852)
It’s not just ROI on marketing, but ROI on legal services in general, right? And then yeah, on your actual work product. Like what are you what are you producing? And and I think you’re absolutely right on the attribution aspect of it. Cause I mean, I still to this day will get someone who’s like, I found your your YouTube video about LinkedIn and I filled that video six years ago, right? Like, and and and people are still referencing it in like discovery calls with me, right? So

Karin (07:46.368)
On your work, yeah. Yeah. Yes.

Karin (08:01.228)
Yep. Yes, exactly. Exactly.

Yeah. Yeah, same thing. I I have people who no, exactly. And like, how would you even do that? Because I have people who don’t say the same thing. I’ve got I’ve been getting your email newsletter for the last six or seven years. And I’m like, okay. So I I will put that as the email newsletter, but then they probably also follow me on LinkedIn. They probably also this, that, and the other thing. So it’s the whole pie. And to try to say that you’ve got a one-for-one attribution.

Jessica Aries (08:08.781)
yeah. And I’m not going in there and attributing that. Like I’m just

Karin (08:36.882)
on like this particular ad to the closing of a case, it’s just not that’s not accurate.

Jessica Aries (08:44.371)
Well, and I think it’s not a holistic way that people build relationships either. Like it’s not looking at the whole way relationships happen, how people interpret marketing, how they ultimately decide to choose you over someone else, right? Like there’s so much more that goes into that. And yeah, I just

Karin (08:52.374)
Exactly.

Karin (08:56.194)
Right. Exactly. Exactly. It’s not just a numbers thing. It’s not just clicking and, you know, robotic. So let’s talk about what what you’re seeing. And because I know you go to all these conferences and there’s stuff that even when I have these shows and I have everybody on, there’s stuff we’re all saying over and over and over, and that’s getting old. And some of this stuff is

Jessica Aries (09:18.224)
Mm-hmm.

Karin (09:22.762)
Is actually getting old and kind of outdated. And where is it going? And what do you kind of see down the road? Let’s just say 12 months. I don’t think we can go much more farther out if you can even go that far out. Like what’s on the horizon? Where are you seeing things heading and then also heading away from?

Jessica Aries (09:26.44)
Mm-hmm.

Jessica Aries (09:45.044)
So AI transformation within firms is huge right now, making sure like so many of the firms have now bought the co pilot, bought the Claude, bought the Chat GPT, but the actual adoption, like real usage of the AI tool in a strategic way, super low, super poor. Like you’re most of your people are likely using AI to write better emails. As I’ve seen on your podcast, people say, right? Like they’re not you people are not using it as envisioned by those, you know, the lawyers when they

Karin (09:57.623)
Is overwhelming. Yep. Yeah.

Karin (10:07.598)
Exactly.

Jessica Aries (10:13.191)
They purchase the product, right? They sign that check, right? so so much

Karin (10:15.434)
Exactly, exactly. And people are listening right now and they’re like, But my emails sound really good. Like what am I supposed to do?

Jessica Aries (10:20.945)
yeah.

You’re supposed to build systems. Like you’re supposed to redesign how you work. And I think so few of us have so few firms have had the opportunity to have like a forward deployed engineer, like marketing technology engineer. that’s what they’re calling them in some of these other industries, for forward deployed marketing engineer who comes in and helps build the systems, builds the context layer, builds the whole entire workflow with you, the expert, the subject matter expert.

Karin (10:26.636)
Yeah. Yeah.

Karin (10:51.724)
Yeah. So what does that look like? Because honestly, I feel like we are possibly speaking over the heads of a lot of people. I heard a podcast yesterday where they were talk still talking about hallucinations. And I’m like, come on. Like, yes, we all know that happens, and that’s like a hurdle you have to overcome, and blah, blah, blah, and talking about issues and and errors and all of that stuff. But let’s move on from that. What this forward deployed.

Jessica Aries (10:53.555)
to ultimately design it.

Karin (11:20.684)
What what was the title that you said a minute ago?

Jessica Aries (11:21.917)
for a deployed marketing technology engineer or for a deployed engineer and then they sit in marketing for a bit and build out the workflows or they and in in legal we call them legal engineers and they’re people who usually come from knowledge management like on the legal side, legal practice side, and they come in and they help to they’re JDs usually or just like former pair legals who have a lot of experience and they come and they help design the workflow with the lawyer that then utilizes AI. And so in marketing they’re called for deployed marketing engineers.

Karin (11:47.981)
Okay.

engineer mark marketing engineer. Okay. So I’m picturing like I’m sure there’s a million other applications, but let’s talk about like someone who’s gonna deploy a marketing, some kind of simple marketing system. And then someone who would also deploy some sort of a legal the work side of it, not just the marketing and the emails and the kind of whatever, but like something that would actually go and into your daily deliverables to a client.

Jessica Aries (11:51.719)
Mm-hmm.

Jessica Aries (12:09.523)
Mm.

Karin (12:20.627)
and legitimately affect your ability to perform at a higher level.

Jessica Aries (12:26.035)
Mm-hmm. W are you asking the difference between the two or what what are their roles?

Karin (12:28.46)
What would it look yeah, what would it look like? Because honestly, I feel like people can’t even imagine some what some of these workflows, what the result is. Like what is that even at the end of the day, what is this marketing? Cause I obviously know what this stuff looks like, but what what would this kind of marketing system look like in in play?

Jessica Aries (12:35.603)
Yeah.

Jessica Aries (12:48.477)
So I’ll give you this example. I’ve built an entire intelligence briefing dashboard that is connected to a full agentic system that has, I think, 12 agents attached to it that all have specific roles within our marketing workflow. So one is our project management agent whose whole goal is to manage ClickUp, our project management tool. Another agent, their whole role is to basically gather research from our competitors, gather information from newsletters we’re reading, and put together basically a

Karin (13:01.517)
Okay.

Jessica Aries (13:16.839)
database of different topics we could potentially talk about or write about. And then a third engine or third agent in that workflow, its whole job is to help put together video outlines. So short form video outlines. then the next engine or the next agent in that, I keep calling it engineer. Next agent in that that workflow’s job is to help write optimized SEO blogs, right? So doing key root research, pulling that together once we’ve identified a topic. And then the next one in that is a

Karin (13:34.974)
That’s right.

Jessica Aries (13:45.864)
Video editor agent. Its whole job is to use Claude Motion to redesign our short form videos and edit them. And it actually uses a vibe-coded version of frame.io that we developed that allows us to give feedback on the videos like we would with a traditional editor, right? And then it makes those revisions and send them back to us. So that we have that. And then we have a final like output and upload AI agent that’s whole job is to just like

Karin (14:00.756)
Nice. Yeah.

Jessica Aries (14:10.675)
quality check everything against AI slop and our standards and things like that and then decide is it ready to go to me for review or ready to go to whoever’s next in the the so in clickup we have like a very mapped out workflow of who reviews what. Yes. So we have a a quality control that’s a human who actually reviews the AI written blog post and things like that.

Karin (14:12.459)
Yes.

Karin (14:22.902)
Review process. Yes, yes.

Karin (14:29.486)
But it but there’s a AI quality control as well. And yes, okay, so there’s two levels of quality control, which I’m focusing on for a minute because this is key. And this is the difference between that conversation I was talking about that I saw yesterday, where they’re still talking about hallucinations and how AI can get things wrong, and the actual like using it and using it in like a whole other level. So

Jessica Aries (14:32.893)
Correct. We have two.

Karin (14:56.056)
Give me example. Let’s say like you go to an event or some or how what’s the trigger for so this these are 12 agents inside one project, right?

Jessica Aries (15:06.163)
Yes, within we we did it in Claude Code. So yeah, it’s within one our our our Biaries OS and our content marketing system that we built. It’s a second brain, essentially. If you’ve heard that terminology, we have a second brain filled with all of our past marketing writing, everything we’ve written on blogs, every video transcript we’ve ever made. So it’s it’s huge, right? It’s massive. And it pulls from all of that.

Karin (15:09.154)
Okay.

Karin (15:25.528)
Tons of stuff. Yeah. Yeah. Okay. So it’s got your language, it’s got your tone, it’s got the things that you like to say, also the things that you actually have said and you know, quotes and all that stuff. and there’s 12 agents in there. And what would be a trigger for some piece of work that you needed from that? Like an event that you’re going to and that you need like a couple videos or something like that, or you or would you trigger it and it does a whole campaign?

Jessica Aries (15:34.439)
Mm-hmm. Correct.

Jessica Aries (15:51.988)
Does a whole campaign by a trigger. We have a control center that it operates from. So every morning it sends me an intel brief and I decide do I want to talk on that topic or do I not? it usually gives me five topics to choose from because I’m the type that I’m never satisfied with one topic. So it gives me like five different topics. I decide do I want to create that into a campaign? And then we have different like workflows. So we have a campaign that would include a blog article and a short form video series, but we can choose in that moment when we turn it on.

Karin (16:03.659)
Yes. Yep.

Jessica Aries (16:18.237)
These are the deliverables I want, like check boxes, right? And then hit submit. It’s kind of like an order form, right? Our control center is. And it just sends it all off and starts working in the background. And then it slacks us as it completes things with links from click up to the task that it’s working on because it makes all the subtasks as well for us. So we can watch him and follow along.

Karin (16:22.902)
Yes. Yes.

Karin (16:34.572)
And meanwhile, you’re like eating breakfast, right? And like totally yes, exactly. You are like not at all involved in the actual work part of this. Okay. Yes.

Jessica Aries (16:38.238)
yeah, I’m working on something else. I’m working on proposal or something. Yeah.

Jessica Aries (16:48.483)
Review, yes. We do keep track too of how much it costs each agent to run its workflow. So we have it also track the money so that then when we do our profitability analysis, we can say, what did the it cost us to do this? How much in token you know, usage and tokens and things like that. So it gives us visibility on the ROI. Now is it perfect ROI? No, but it gives us some visibility on it.

Karin (16:53.986)
Nice.

Karin (17:02.164)
Nice. Nice.

Karin (17:07.838)
No, but it’s better because that’s amazing because that is a a thing I see regularly that it’s really hard to monetize the tokens because first of all, it’s it’s arbitrary. Nobody has any idea like how how much time does a fable five point one token like what does it take me to get through that? Like and it’s like, I don’t know, just keep throwing money at it. Like I have no idea. It’s just like totally a mystery.

Jessica Aries (17:30.841)
Mm-hmm, mm hmm. Mm-hmm.

Karin (17:37.986)
So that’s amazing to be able to track that.

Jessica Aries (17:38.312)
Yeah. We that’s why we ask we asked it to base it on the API cost, right? If we were building this outside in like Hermes or some other sort of agent, because that was the way my brain needed to work. I need to have some number. And you know, when you purchase a Claude account, right? Or you purchase a Chat GPT account, the hundred dollar a month account, it’s like substitut, it’s not the f real amount. So it makes it really hard. And I wanted to also be able to track spend over time. Like, does is it getting more expensive?

Karin (17:43.872)
Yes.

Karin (17:50.357)
Yeah.

Karin (18:04.322)
Yes. That’s amazing.

Jessica Aries (18:05.595)
Are those workflows changing? Do we still want this workflow to exist in this way? And I think if you don’t have visibility on if you’re not documenting and benchmarking marking along the way, you won’t know. And one day token prices, let’s be real, are gonna go up because all these companies are gonna go public and they’re gonna wanna make more money. Yeah. So

Karin (18:09.677)
Right.

Karin (18:16.908)
Like probably tomorrow. Exactly. Yeah. Okay. So I want to hear like an example of like the other side of it. So the legal work side. And then I want to talk about like how it has changed and what you know you kind of it cause it sounds like you’ve got it set up for a lot of potential like flexibility in the future. like both of these systems where it’s like, okay, if

Jessica Aries (18:26.579)
Mm-hmm.

Karin (18:45.418)
If this changes, this’ll still roll with it. Okay.

Jessica Aries (18:48.433)
Yes, correct. And it has to be sc it has to be able to be modified and it also has to be able so I’m thinking about two the system that I built. It’s currently built within like our our my drive. I want to scale that out into like a GitHub repo and allow my team to be able to access it and run it. But right now, so much of it is based on my thought leadership, right? It’s not needed to scale yet, right? But thinking about as you’re designing a system, what does scaling look like and am I designing it and giving

Karin (19:09.942)
Right. Yeah.

Jessica Aries (19:17.263)
AI a framework for how this might then be pushed into the cloud, right? And what does that look like? And who has permission? Who has a governance issues? Which which is a whole nother topic. But yeah, these are the things you need to think about when you’re designing these systems. How does it scale? How does it not just live in my system? Or if my computer is off, does it still run? Right. Like where, how will those

Karin (19:21.293)
Yes.

That’s a lot. Yeah. Yeah. Yeah, exactly. Totally.

Karin (19:34.861)
Right.

Right. Right.

Jessica Aries (19:41.093)
it be available to other people on the team or if I break my leg and I need someone to write something in my voice, right? Can they do it while I’m in the hospital getting a cast put on?

Karin (19:46.359)
Yeah. Right. Yeah. Can they get in there and will it be will it be good too? Like will it just be kind of a mess or will it actually like will it roll without me? Okay. So what’s an example of like the the legal work deliverable side that’s because I feel like so many people focus on these marketing things because they seem that that feels sort of like superficial and easy to substitute and it’s like, it’s a LinkedIn caption, you know, it doesn’t have to

Jessica Aries (19:52.371)
Mm.

Karin (20:14.786)
I’m not saying anything legal. This isn’t something somebody’s paying me to do. I’ll just slap it together and, you know, you know, whatever, bang, throw it up. But when it comes to the actual work product and things that someone is paying you to do, what does that look like that when you get in there with firms and help them set that kind of that kind of system up?

Jessica Aries (20:37.555)
So that would be like more like the legal engineering side. And when we’ve seen that, like we we worked with one firm where they wanted their trademark practice, wanted to have a like a database of sorts of past trademarks that were accepted in any comments and feedback that came back and and basically build almost a a way for them to consistently know like these are the standards and practices for getting something through on a trademark, right? Like what it and so they built a whole workflow connecting to the API of USPTO.

Karin (20:58.574)
interesting.

Jessica Aries (21:04.871)
To be able to pull that information in and then ultimately map a workflow where those standards were codified into skills that then became part of their workflow of every time we work on a trademark, run through these skills. Like that’s all workflow, right? But it’s one that that improves the work of the lawyer, right? So the lawyer itself could, yes, use the knowledge they have, but if they utilize past histories of things they’ve submitted, comments and feedback they’ve gotten back USBTO, right? Incorporated that into

Karin (21:11.545)
fascinating.

Karin (21:18.574)
That’s yeah.

Jessica Aries (21:32.721)
Developing sk skills that then can apply across the entire practice, then everyone’s operating at a higher standard. And so it’s, it’s,

Karin (21:39.437)
Right. It’s like the idea of, you know, this information sits inside this one partner’s head. And how do we this is literally a technical way of pulling that out of his head where it would have just been like, you know, in the past, X, Y, and Z did not work. And I only know that because I have it filed away in my brain. and now we’re pulling out that information, and you could easily pull that into

Jessica Aries (21:57.395)
Mm-hmm.

Karin (22:07.564)
all of your proposal process and the the the entire like application process now so that you don’t even waste time with these, you know, words and and phrases and whatever that didn’t work in in the past.

Jessica Aries (22:20.243)
Well, now you have marketing language too to say we designed a whole system that ensures that our our work product gets better results, is less likely to get rejected, is less like you create a value statement that you didn’t have before because you designed a system to audit and level up as a team, right? And you have something no one else designed, right? And you can say that and I mean now others could probably take that idea, right, and do something similar, but it’s this this idea of

Karin (22:31.552)
Yes.

Karin (22:35.842)
That’s amazing. Yes. Yeah.

Karin (22:43.692)
I hope so. Yeah.

Jessica Aries (22:44.859)
Yeah, yeah, creating something and then using that to add value and talk about that as why work with us first. It’s a differentiator, right? Versus another firm.

Karin (22:51.926)
Right, right. So where are where are firms getting this wrong? And what are they kind of missing the boat on in terms of like that kind of I feel like as soon as we start talking about actually doing your legal work and like setting up an AI system, I can feel like people’s hairs on the back of their necks stand up and they’re like, whoa, whoa, whoa, whoa, whoa. Like, let’s not go that far.

But you know, like what you just described is genius. And but also when like when you think it through, like it seems pretty straightforward and and low risk. So where where are people going too far? and you know, where should they be kind of taking a look at things that are are similar, like kind of low risk, easy options that kind of automate systems like that?

Jessica Aries (23:25.171)
Mm-hmm.

Jessica Aries (23:41.672)
Think using things you already have in your system, right? Not necessarily active client work in the moment, right? Like in this example, they were literally taking from their database of their prior submitted trademark applications, right? And seeing what feedback and comments they got. That’s as you mentioned, low risk. Higher risk, a lawyer going off in vibe coding or connecting all of you know your

Karin (23:47.97)
Yeah.

Jessica Aries (24:03.857)
your document management system to AI without understanding are they on an enterprise level account? Are they all of those kind of security things, right? Slash, are they allowing the agent potentially to delete files that they shouldn’t, right? And who where’s that the going rogue on their own without telling anyone what they’re doing? And I also I actually understand that as someone who is an early adopter of technology in general, the idea that I want to play an experiment, but I think do it in a safe environment, right? Protect protect your

Karin (24:16.695)
Yeah, yeah.

Jessica Aries (24:33.713)
your key d client documents and maybe play in a special folder that you set up, right? Or experiment somewhere on a a different machine, right? Than your your work machine. So I think I think there’s that too.

Karin (24:38.743)
Right.

Karin (24:43.384)
Well, I think this goes back to your double layer of approval process. So do you have kind of standards around how you like where the approvals are and you know what does and doesn’t get where your AI does and doesn’t have access to and do you have any kind of framework or thoughts or like standards for how to think about all of that?

Jessica Aries (24:47.624)
Mm-hmm.

Jessica Aries (25:07.943)
Yeah, I mean, so much of what we initially build is in like a a separate drive, right? And then we deploy it out, right? And like we create plugins and things like that into our own clawed code that then the whole team has access to. But then there’s the responsibility of maintaining it, which falls on me at this point because we haven’t hired anyone to be the full-time person in that in our agency. but so much of it is do you have a plan for that? Do you have some way of of determining who owns, you know, maintaining the skills, the context layer?

what goes in it, what doesn’t go in it, when what your contracts say regarding client information and what you put and don’t put into, especially around marketing, right? do your clients know how you’re using AI? so so much of that is part of just a structure we built. We also have an AI responsible use policy. And in our agency, we do a workshop every two weeks on AI because it changes so quickly that we’re constantly teaching new things and

Karin (25:41.474)
Yeah. Yeah.

Karin (26:00.78)
Yeah. Changing. Yeah. Yeah.

Jessica Aries (26:03.707)
Yeah, and then requiring our team to build something with that new skill or show us that they understand the difference between a soft hallucination and a hallucination. You know, we talk a lot about hallucinations. We don’t talk enough about soft hallucinations. And I know you so a soft hallucination is when AI confidently tells you something that sounds real because it’s the way a lawyer would talk. Like across the industry, we’re seeing X, right?

Karin (26:14.701)
What’s that?

Karin (26:24.898)
Yeah.

Jessica Aries (26:26.087)
But then gives you no citation. And it’s the number one thing that gets missed on the AI assessment we give marketers is they miss that question. Like hands down, because it’s the way lawyers actually speak. They’ll say things like, across the industry, we’re seeing X, right? And so when AI says it, they’re that day, they don’t think to redline that or to say where’s the source for it, right? They think it comes from the conversation of AI, but AI is not human. AI’s not having conversations with your clients, right? So those are soft hallucinations and it’s like the number one thing legal marketers miss.

Karin (26:40.595)
Yes, yeah.

Yeah. Yeah.

Karin (26:49.484)
Yeah. Yeah.

Jessica Aries (26:55.267)
and it’s the thing that I often tell them, if it if it’s very confidently saying something, you need to go back and look for the source. Yeah, look and and I mean, we we notice the ones that are like stats without a source. We don’t notice those types of statements as much because we’re so used to seeing lawyers write them.

Karin (26:59.96)
Probably not. Yeah. Yeah.

Karin (27:06.52)
Yes.

Karin (27:10.744)
Totally. Yeah. Yeah. I see that all the time. And and also just in like personal stuff, I had Claude very confidently give me a recommendation for this cafe that I wanted. I was at one and I was like, okay, we’re gonna be in this other area of a different city. Find a cafe like that for me. Found this one that sounded perfect. It did not exist, but it was so confident about it. I was like, Claude, actually, no, I think this was chat.

But either way, when there I feel like that’s a red flag. When it’s too confident and then it comes back and it’s like, my bad. I’m so sorry. Yeah. You’re right to question me on that.

Jessica Aries (27:45.159)
Mm-hmm.

Yeah, you’re right. I’m so sorry. Yeah. Yeah. Like Yeah, because I’m an actual human living in this physical world, standing in front of this address that isn’t what you said it is, right? Like

Karin (27:55.751)
Yeah. Yeah, I am.

Karin (28:02.574)
Right. Exactly. Yeah. Okay. All right. I want to talk about the difference between women and men for a minute in AI. Because when when we first started talking about like what was next, I sh I shared this newsletter of this amazing woman in Australia named Nikki Sweeney, and she focuses very much on the difference women bring to the table when especially when it comes to AI.

Jessica Aries (28:11.763)
Mm.

Karin (28:30.71)
And I feel like you probably have some opinions on this. But when I talk to women and men, it’s different conversations about AI. And I feel like we have philosophically at our core have different thoughts about why it’s cool or how we can use it and how it kind of slots into our lives just because of like the differences in gender roles and you know all of that. What do you how do you feel?

And where do you br think women bring kind of a unique approach and perspective when it comes to just use of AI?

Jessica Aries (29:04.957)
So it’s funny you say this. I was literally having a call, I guess an hour before this call with one of the lawyers I’m working with who’s actually a very prominent, well-known family law attorney. And she’s been doing a lot of work around the difference between men and women and just di and I was asking her a question and she was talking about women and how our brains work. because she’s done a lot of research on the science behind it and how we look at things from the community spec perspective often. We’re looking from

Karin (29:29.933)
Yes.

Jessica Aries (29:31.026)
the weakest to the strongest, right? Of the members, the eldest to the youngest. we’re looking holistically a lot of the times. Our brains operate in that way because we, for whatever reason, biologically, were created to look at the community as a whole, because community was important for our survival and for our futures and yeah, our baby survivals, right? We’d have babies and then we’d go, you know, work in the field and then, you know, someone else was caring for that baby. Right. And so having good, strong community was important.

Karin (29:46.573)
Yep. Right. Are babies survivals? Yeah. Yeah. Yep. Yeah. Exactly. Exactly.

Jessica Aries (30:00.295)
That is not the way men’s brains work. much more singular focused. Nothing wrong with that. Just different brains, different brains. And so, yes, competitive, thinking much more about winning the race. Whereas I I mean, often the people, the first people who raise their hands when I’m giving an AI workshop to bring up considerations, to bring up concerns are women and they’re bringing them up about how do I teach my child critical thinking if this is happening? What is gonna happen? Are they gonna have jobs in the future? Or

Karin (30:02.005)
No.

Competitive. Yep. Yep.

Karin (30:13.397)
Yes.

Karin (30:26.316)
Yes.

Jessica Aries (30:29.543)
How do I ensure we’re still protecting the earth? You know, and I’m like, I don’t have the answers to all the questions about AI, but I will do my best to answer them, you know, as ethically as I can. I do think it’s important to think about these things, but you know, often I’m brought in to and I have sixty minutes to tackle all of these things. And yes. I was like, I wish, I was like, when I run for president, which I never will, so but it it kind of feels that way, you know?

Karin (30:32.044)
Yeah.

Karin (30:38.05)
Gosh, yeah. Yeah.

Karin (30:46.082)
To solve the world’s problems with AI.

Yes. I think that’s so true. Yeah. A hundred percent. One great analogy that I I keep coming back to is that most women that I am familiar with and I am reading and following approach AI like I need a wife. I need someone who carries the mental load.

Jessica Aries (31:17.533)
Mm-hmm. Correct.

Karin (31:17.644)
has all that information and it’s like this administrative stuff that like we don’t really want to hold it. It’s just that nobody else is holding it, right? Like I know the shoe size. and I, you know, all of the grocery stuff that we were talking about and the laundry and all of that stuff. I need that covered. And that’s where I’m starting from. Whereas men are like,

Jessica Aries (31:22.663)
Yes, one hundred percent.

Karin (31:42.53)
How can I conquer the world and this industry and whatever and win and put a flag in it? Right.

Jessica Aries (31:52.66)
And I think that makes sense biologically. Like when you look at the br you know, how we are approaching things. And I mean, I’m a woman business owner. I am married. I do have a six year old, right? The f one of the first things I offloaded was grocery shopping, right? Like that was how do I get my AI agent to make healthy meals that my kid will eat, right? And grocery shop for me, right? Because that is a mental lift I don’t wanna do anymore, right? And my husband will do it, but he’ll always forget something important, right? Not to not to but it’s just there’s something that

Karin (31:55.03)
Yes. Yeah.

Karin (32:04.59)
A hundred percent. Yep.

Karin (32:09.602)
Yep. Exactly. I’m so sick of it. Yeah. Mm-hmm. Yeah. But it’s the truth, yeah. Yeah.

Jessica Aries (32:20.293)
It lives in my head and I d didn’t come out of it, right? Like in time to make it on the list, right? so yeah.

Karin (32:23.072)
Exactly. Right. Yeah. And I feel like I just wanted to bring it up because I feel like it’s not being talked about enough. and I feel like there are not enough AI women leaders, which is why I as much as we’ve had these AI conversations, I for sure wanted to really focus you specifically as a woman. And ’cause a lot of these conversations I’ve been having with other men, and we don’t talk about it in this in this kind of lens and talk about

Jessica Aries (32:31.603)
Mm-hmm.

Karin (32:52.226)
The difference in making sure that in your firms, you’re sitting down with the women and saying, hey, really, what how could this really work for you? Like what do you really need? And like we’ve got all these ideas around what we think AI should do, but you know, is that what you really want? And and what ideas do you have that we’re not thinking about?

Jessica Aries (33:02.642)
Mm.

Jessica Aries (33:18.021)
Yeah. Well, and and I think that’s what’s so powerful about when you do sit like we run AI labs where we bring, you know, lawyers into they work in small groups and they design a system, something, a workflow, something together. And it’s so powerful when you get them together because it is a diverse group. You’ll have the person there who’s like the senior lawyer and the like whipper snapper new associate, right? And they’re they’re on totally different wavelength about AI, but what they build together.

Karin (33:33.613)
Mm-hmm.

Karin (33:39.672)
Yes.

Jessica Aries (33:44.496)
is so much more powerful than what they could build on their own. because the the young, you know, new associate is bringing in a totally different perspective to the law. They’re fresh green eye green to the whole idea of how do we build this and they’re excited about AI and they’re thinking about it differently. The maybe more okay.

Karin (33:47.97)
That’s amazing.

Karin (33:56.407)
Yeah, yeah, yeah.

Karin (34:02.53)
And they just are I have a sorry, I was just gonna say they just have a different lived experience, right? So they’re just coming to the table with different thoughts, different needs, all of that. this was one of the things that the big takeaways from my MBA program is it’s very much group work. Like I think it was like eighty to ninety percent group work, where

Jessica Aries (34:07.442)
Yes.

Karin (34:24.452)
this is like you gotta figure out how to work with groups and these are very strong personalities that all come from different backgrounds. And I’ve talked a lot in the past about my group was two women and three men. One was Asian, like you know, full English was not his first language, one was Greek and like totally different culture, and then a white guy. And we

Jessica Aries (34:45.835)
wow, okay.

Karin (34:53.208)
Could not have had different backgrounds and cultural expectations of how we were going to work together and what felt good. So in our first conversation, we’re sitting there and we’re trying to like, you know, be friends. The Greek guy is being very kind of extroverted and chatty and kind of cutting into people’s sentences and ch you know, and the Asian guy is so offended and he’s instantly like, I cannot stand you.

Jessica Aries (35:00.829)
Mm-hmm.

Jessica Aries (35:14.555)
Mm.

Karin (35:22.81)
And this was day one. And so then we had three years of that, and it went down from there. And so to ignore, I know. So it was it was a journey. But I bring this up because I feel like it’s so important to think about these underlying channels and flows of the culture and your the things everybody’s bringing to the table, whether it’s

Jessica Aries (35:24.435)
Mm-hmm.

Three years.

Karin (35:48.387)
that their their culture, whether they’re male or female, what they d ate for dinner last night, whatever it is, it’s gonna affect the everything going on in your firm and the kinds of decisions that are made and whether those decisions are really the right ones to be made for the whole firm. So I feel like that was kind of a tangent.

Jessica Aries (36:03.643)
Mm.

Jessica Aries (36:10.801)
To that, when the models first came out, if you would ask it to produce like an image of a lawyer, it used to produce a white male in a suit, right? Like that even the models themselves are coded by a certain demographic of people who are going to have a certain perception of what it means to provide you an image of whatever, right? Or to provide you feedback on whatever. And so I think that’s where it’s so important to encode your own judgment into the system and know how to do that effectively through skills. Like what does great look like? What are our standards? How are we

Karin (36:15.894)
Yeah. Yeah. Right.

Karin (36:22.274)
Yes.

Karin (36:27.288)
Yes.

Karin (36:36.354)
Yeah, yes.

Jessica Aries (36:40.633)
codifying those in a way where we’re talking about what great looks like and then feeding that back into the system and telling it this is how we want you to grade yourself. Don’t just grade by the universal standards of the internet or whatever someone coded into you, right? Like this is how we we identify greatness. And that is where you really start to level up.

Karin (36:50.52)
Yeah.

Karin (36:56.034)
Right.

Karin (37:01.208)
Well, exactly. So that was something you mentioned a a while back that I wanted to come back to and it reminded me that when you were talking about these core w so we call them kind of brand files. What did you call like these core files that you have that are all file you know loaded into your Claude code and everything? What do you call the kind of your brand? all the notes and you know, all that stuff that you have in there? Second brain. Yeah.

Jessica Aries (37:24.135)
Yeah. The s the second brain, the context layer. So we have a operating some people call it an operating system. Some people call it a second brain. Some people call it a context layer. we call it a context layer most of the time, but it’s essentially where all of that knowledge lives that’s going to inform AI to give it better perspective on how it gets should give you an output. And that can be everything from your messaging guidelines, your ICP, right? Like all of that information that’s collected, your voice, right? Codified into a skill. that all lives there.

Karin (37:31.798)
Yeah. Right.

Karin (37:52.441)
This is not different. Yeah. So it all lives inside inside Cloud, where in the past it would have just been in somebody’s head. So this is not new. This is just technically different. So once again, I feel like I say this 10 times a day, but this is not, we’re not reinventing marketing here. We this is marketing strategy and defining your clients and putting all that stuff and then drop it in a dropping it into, you know, you and I use Cloud or whatever your platform of choice is. But

Jessica Aries (38:09.299)
No.

Karin (38:21.458)
We would have 10 years ago, we would have been having the same conversation and it would have been talking about marketing strategy and you know, defining your clients and all of that stuff. And that would have just been some Google Docs or, you know, maybe whatever version, it still would have existed and it wouldn’t have been that different than what you’re describing. It just would have gone from human to human instead of like bringing in the AI thing. And the reason I brought that up in terms of like this, you know, let’s think about.

Who your people are in your team and whether they’re men or women is because if you have not done a good job of defining this, whether you’re using AI or whether you defined this in Google, it doesn’t matter. You’re gonna have a garbage output and then you’re just gonna decide that marketing doesn’t work. And if you have decided that you’re wrong, you just probably have some garbage somewhere that you haven’t done a good job of, you know, doing whatever, wherever that brain or the

brand files or whatever you’re calling them, they’re probably garbage.

Jessica Aries (39:21.853)
Well, and I think to add on to that, are you c making it iterative, right? Because the more you input data, connectors, things like that into AI, you start having these sales conversations where the transcripts are now, you know, put into AI, you start to realize maybe your ideal ICP is even more nuanced than you thought. And that if you refine those, if you continually update those based on what you’re learn like AAI is learning about you because you’re connecting all this data to it, you’re gonna get even more refined output. You’re gonna get an even

Karin (39:26.08)
Yes.

Karin (39:37.462)
Yeah, yes.

Karin (39:48.268)
Yes. Yep. Yep.

Jessica Aries (39:49.01)
better ICP. You’re going to get an even better brand statement. but that’s an iterative process. And that’s also where I think having a good understanding of the structure of those context files and how to build an orchestration layer that is self improving is really important. I don’t I don’t know if we’re talking enough. We haven’t talked a lot about that, I think, in general in marketing. Some people are, but there is a self improving l like aspect to it of having a second brain that is continually making those adjustments and making sure you’re you’re

Karin (39:57.719)
Yes.

Karin (40:07.392)
Yeah, yeah.

Jessica Aries (40:17.939)
creating the right types of content. You’re targeting on the right channels, right? You’re doing that kind of analysis from a strategic perspective. You also have to have a marketer who knows to look for that, right? Like and

Karin (40:25.974)
Yeah. Well, and this kind of comes back to the earlier question about like where are we going twelve months from now? And I I I see it being a very refined, like we’re not talking about like a whole new AI plot platform. We’re talking about it’s just getting better and refined and faster. And you know, what are you seeing that would be different? Cause I I I feel like that’s that’s what you’re saying is that having it learn, having it get better, having it improve.

And then going from there. What what else are you kind of imagining that’s kind of down the road?

Jessica Aries (41:00.755)
Well, I think to the thing to remember is the more you refine that, what what is that where does that output live? Like you can write really great blog posts, but if you’re only like we’re the future of marketing is really gonna be a a AI to AI in the sense that our agents are going to be the ones reading the content on the blog and then serving it back to us in our chats and things like that. So what are you doing beyond that to create human connection? Right? How are you?

Karin (41:24.812)
Yeah, yeah. Yes. Yeah. Yeah.

Jessica Aries (41:26.173)
How are you going to the conferences and having the networking discussions and talking to people and I those parts of it are the time is still gonna be invested just in different places. You’re no longer gonna be writing maybe the whole blog article. You’re now gonna be breaking bread with someone across the table and having a real conversation and asking them how their kids are and you know, doing all those the human things, right? That AI can’t do for you. and so not to lose that in your marketing. Yes, and AI is doing the follow up, AI is keep keeping it in track. AI’s doing that, honestly, that that.

Karin (41:35.543)
Right.

Karin (41:40.321)
Exactly.

Karin (41:45.068)
Right. Exactly. A and then AI’s doing the follow up. Yeah, yeah.

Jessica Aries (41:54.418)
The burden we were talking about earlier, but but from like a BD perspective, a marketing per se, right? Like doing all of that part of it. And you know, someone has to be overseeing that, right? And double checking it, making sure it’s not breaking. But that’s that’s the part too. And I know you’ve had some people on the podcast talk about that, but I think we can’t miss that as well, that our time is gonna be hopefully yeah, hopefully spent there.

Karin (41:55.587)
Yeah. Exactly. Exactly. Yeah.

Right.

Karin (42:13.016)
The human part. Yes. Yeah. Yeah. Okay, so it’s time for the book review. I almost forgot about this. We’re getting there. so I know you have a good one to tie in. So Jessica, what’s the book that you’re gonna add to the Thought Leaders Library in this episode?

Jessica Aries (42:18.867)
Mm-hmm. No worries.

Jessica Aries (42:30.961)
Yeah, so The AI Driven Leader by GF Woods. If you have not read that, it’s a great book on using AI as a thought partner. and really thinking about are you using AI to be like a coexecutive in your business, right? And helping you rethink, you know, find those gaps, those holes, those those moments when you’re trapped in your blue box box of law firm, right? And you’re not thinking or innovating or looking for new opportunities.

Karin (42:51.031)
Yeah.

Jessica Aries (42:55.719)
that using AI to bounce ideas, to challenge you to operate and even sit like as your ICP, right? And like give you feedback on things, right? That can be really telling and really helpful in using AI in that way versus just content production or just video editing, right?

Karin (43:10.784)
Yeah. Right. Trying to get out of those low level things, like writing emails. Yeah. Just the basic stuff where you’re just asking Chat GPT. you know, just very super Google Google Google things you could Google. Yeah. yeah, trying to get up and like level up, like you were saying earlier, kind of elevate what it can do for you and like.

Jessica Aries (43:15.623)
Yes, tactical. Mm-hmm.

Jessica Aries (43:26.211)
Yeah. Things you go Google. Yeah.

Karin (43:38.527)
really expand. There’s so much that we are just barely like tip of the iceberg with all of this stuff. And even tip of the iceberg is amazing. Like it is it’s crazy to see the amount of things it can do. okay so the book was AI Driven Leader. We’ll we’ll obviously link to that on the show page. so you guys can download that and it’ll kinda I I haven’t read this one, but I’m assuming it kind of walks you through how to like how to think bigger. Okay.

Jessica Aries (43:49.971)
It is.

It is hot.

Jessica Aries (44:06.929)
Yes, correct. And and there’s example prompts and all kinds of things in there. And there’s actually, I believe Jeff Woods has launched even like a whole program. Like if you wanted to do that, but I I saw him speak at a conference a few years ago. I was just so impressed with the way he was elevating his executive brain, basically by rethinking, like using AI to challenge assumptions and the things that we often don’t. We don’t ask it to push back on us. We don’t ask it to argue different points and different perspectives in front of us. And so it was really cool to watch live.

Karin (44:10.392)
Perfect.

Karin (44:22.667)
Yeah.

Karin (44:34.978)
That’s amazing. So you mentioned prompts a second ago. And so this is kind of one of the last things I want to kind of talk and spend a minute on. Do you feel like like there was a time which wasn’t that long ago? I I don’t want to guess because it feels like it was a year ago, but maybe it was last week, where you would get these sort of copy and paste prompts and they would sometimes be, you know, paragraphs long, day pages long of whatever, some sort of task. I am seeing that less and less.

And it seems like it’s more of a conversation where you’re just sort of telling it like this is the sort of thing I’m and then walk me through, talk to me about like kind of where we’re at with prompts and what you anticipate that to kind of look like over, you know, in the near to long term future.

Jessica Aries (45:23.025)
Yeah, well, I mean, so I I’ve recently said the prompt library is dead because I I believe like with the context layer, the way we’re adding so much information about our clients, our marketing strategies, the things we’re trying to build, your prompts can become much simpler. In fact, most of my prompting is voice prompting now. Like I and I talk to it like I’m talking to my husband, like, do this, right? Like whatever. My poor husband. Yeah, like why are you why did you do that? What what were you thinking?

Karin (45:28.022)
Yeah.

Karin (45:44.878)
Stop arguing with me.

Yeah, exactly. I my gosh, I tell it all the time. I’m like, this is horrible. What were you even thinking? You didn’t read yeah. I also have found that some of the prompts that I used to use are limiting. Like they’re they’re telling it to do stuff. Yeah, exactly.

Jessica Aries (45:55.812)
Ha ha

Jessica Aries (46:02.033)
Yes. Well because they’re construct

They’re constructed in such a formulae way, you know, that they they’re following an old methodology and the LLMs have improved so much that you don’t you they can create their own plan, right? From and they they you don’t even have to choose the mode anymore in some of them, right? Like it’ll choose like the reasoning mode for you by default, right? So before you used have to tell it, I want reasoning mode on so that you think deeper and like give me a better plan than this. Sorry, I don’t know why it sounds like that in my head when it’s AI, but it does.

Karin (46:17.389)
Right.

Karin (46:27.982)
Yeah. Right, right.

Jessica Aries (46:36.637)
It’s nerd voice thing going on there. but it was like a very I I like to say like the engineering approach to writing a prompt, and we don’t need to be engineers anymore in that sense, right? Like we need to be able to know what’s possible. We need to be able to know how to give it good instruction, good context. But spending a bunch of time, building a bunch of long prompts, not a good use of anyone’s time, in my opinion. If you have a good context layer, if you have good projects set up, and if you’re using any of the agentic systems like coworker code.

Karin (46:46.092)
Yeah, yeah.

Karin (46:59.095)
No.

Jessica Aries (47:05.271)
or codecs or I guess Chat GPT work, you don’t you don’t you can tell it like the outcome should look like X and it will build the plan and ask you, is this the right approach? Right. And so before it starts working and you can audit it and say, no, don’t do that, do this instead, right? And give it better direction. So it has gotten better at that. And so there’s no need to have these really long prompts that are limiting.

Karin (47:15.17)
Yeah, yeah, yeah.

Karin (47:21.996)
Yeah. Yeah.

Karin (47:27.49)
Yeah, and how long do you think it’s been since you were last using prompts? Do you think it was in within the last six months or so that you were still using prompts and like this has started to change, or do you think it was longer ago than that?

Jessica Aries (47:40.39)
No, it definitely was because I have and to this day still one of the most like popular things that people download from my website is our prompt toolkit, which is basically a starter prompt library. which I need to take down because I’ve been like not being very kind to prompt libraries lately. So Lately, I literally li I think my post last week was some or the week prior was like a tombstone and it said prompt library, we had a good run. Right, like

Karin (47:46.866)
yeah.

Karin (48:04.834)
Yeah, right. But you know, you know, I picture the people listening and I imagine that they are spending their time being lawyers and they’re not you and I who spend a lot of time following all of this and whatever. And so for them to hear that we’re not doing prompts anymore, it’s one of those like

Moments like I thought this was the thing and I needed to Google a prompt to help me put together an email campaign or whatever. And it’s like, nope, we’re not doing that. We’ve moved on. That’s you know, ancient history. And like this is where I think people have a lot of frustration with this AI is how fast moving it is. And it’s not just like, yeah, you could do that. It’s like, no, we absolutely don’t want to do that because like the thing that I was telling you.

Jessica Aries (48:23.559)
Mm-hmm.

Jessica Aries (48:31.963)
Mm-hmm.

Karin (48:49.898)
One hundred percent is the right answer, is now one hundred percent the wrong answer.

Jessica Aries (48:54.771)
Okay, so here’s my thought on this because we have so few lawyers who are I think using AI in that way that I’m not gonna discourage anyone from using prompts if that’s what’s working for them. You can just it’s kind of like asking, and I say this all the time, it’s kind of like asking a Michelin star chef to make you ramen, instant ramen, like what? Like, right? Like instant ramen. I mean, instant ramen’s great, but like that’s a microwaveable task, right? Like w so if that’s what gets you there.

Karin (49:12.364)
I love ramen though. Yeah.

Yes. Yes. Right.

Jessica Aries (49:21.691)
If it’s making instant ramen is what’s gonna get you familiar with using the tool, then by all means you do the instant ramen route until you can graduate to Chef Weardy or I don’t know, right? Like until you’re ready to move to the next thing, right? and if that’s what’s getting you there, great. Once you’re ready and you’re ready to learn more and you’re wanting to really kind of create these systems and things like that, I don’t think needing I know ne having a giant prompt like that is not the way to structure it.

Karin (49:22.115)
Yes.

Karin (49:31.362)
Right. Spaghetti. Right.

Jessica Aries (49:50.536)
There are instances where I’d say having strong prompting fundamentals, though, will benefit you from an orchestration level. Because if you are setting up agents and you need them all to collaborate, you need to have an orchestration layer and an orchestration prompt and potentially an orchestration agent that’s in charge of that. And that’s all done through prompting. So learning the foundational skill of prompting is not a bad thing. Like you will be better off for having learned to prompt. Do you need to save all of those prompts? Yeah, yeah. No, probably not. You can make those into skills.

Karin (49:57.261)
Yeah.

Karin (50:11.383)
Right.

Yeah, it’s not a total waste. Yeah. Yeah. Yeah.

Jessica Aries (50:19.143)
Right. You could transfer those into skills within cowork or within ChatGPT work and probably be better off for it because you don’t need to go and find it and copy and paste it every time. You just set up certain triggers and say, When I use this word, pull up the skill, right? And use the skill. And skills for those who aren’t aware are like saved prompts basically within the system. But I like to think of them more as like a saved process or a saved like codified way we approach things. And so that’s what I would recommend doing.

Karin (50:32.844)
Yeah.

Karin (50:37.612)
Yeah.

Karin (50:45.356)
Yeah. Yeah, it’s yeah, I think skills. So for one example would be like I have skills around this podcast where after you know, before like things that help me prep for an episode, and then things that help me once we’ve recorded all of that. and so it’s a lot more than just put together a LinkedIn post for me. It’s not just like one thing, it’s a whole set of different stuff that’s gonna happen based on some sort of a trigger.

And there it’s amazing. But like you said, start where you gotta start and then go out from there and and just recognize it’s a gr it’s a learning process and you know, where you are today will not be the same six months from now or next week or whatever the case might be. But try not to be frustrated with it and give up on it.

Jessica Aries (51:34.174)
Just what we’re seeing with the rise of personal agents like Muse and, you know, dots, not to get on that whole topic, but those I think will get people more familiar with AI in their day to day work. And so maybe it’ll make more sense why we don’t need long prompts if you use those or if you ever download one of those. I think I think so much of it is how soon will those things be integrated into our daily life where people will not even imagine their life without it, right? Like I can’t imagine my life before social media. I can’t imagine my life before the iPhone, right? Like

Karin (51:37.42)
Yeah.

Karin (51:44.182)
Yeah. Yeah.

Karin (51:48.941)
Yes.

The little squishy guys.

Karin (51:59.361)
Right.

Exactly. Right, right.

Jessica Aries (52:03.879)
Those are things that eventually that’s gonna it’s just gonna be ubiquito ubiquitous and we will will just be using it and not even realize it. And there’s AI and everything now, I feel like. So So you’re using it whether you want to or not. Most likely. Yes.

Karin (52:08.844)
Yes.

Karin (52:14.294)
Yeah, yeah. For good or bad. Yes, exactly. Okay, so gosh, we have covered so much. I usually at this point ask for like one thing they should do after listening to the episode, but there are a bajillion. So, like to to go back to where we were talking about in the beginning, where so many people are overwhelmed, they feel like they’re behind. What’s some like the first thing anyone

can and should do after listening to this episode to like f help them calm down and feel like they’re not quite so behind. Like they’re sort of catching up.

Jessica Aries (52:53.907)
So I’m gonna I’m gonna admit something here. I I feel behind. And I literally will listen to podcasts in the shower on AI. So like it is impossible to keep up. So stop trying to. Like my recommendation would just be make a daily practice of using AI, similar to the practice of law, like you’re currently doing with your day-to-day work. Commit to maybe 10 minutes on AI doing something new, challenging yourself to work in a new way with it. and stop trying to keep up.

Karin (52:57.431)
Yeah.

Karin (53:04.256)
Yes. Yeah.

Jessica Aries (53:21.243)
‘Cause it’s impossible. Like literally, we none of us can do it. And I I think it’s kind of like a a fool’s goal to try, right? Fool’s ear and to like try to keep up with it. So just don’t, right? Do what you can and gain AI competency, not AI fluency, which means know how to be able to fix something when it breaks in AI, right? Like know how to know like this is a bad prompt and this is sketchy because it’s clearly a a soft hallucination. Know those those fundamentals of AI competency and let’s stop trying to master it. Let’s just

Karin (53:21.581)
Yeah.

Karin (53:28.088)
Totally.

Right. Yeah.

Karin (53:35.415)
Yeah.

Jessica Aries (53:50.609)
Do the best we can with what we’ve got.

Karin (53:52.559)
Do you read every single every single thing that is gonna go out to the public? Do you have some sort of like kind of universal rule in terms of approval where you will not put anything out without, you know, you or your partner reading it or what do you have in in mind in terms of like making sure that you’ve got that human in the loop?

Jessica Aries (54:16.497)
It’s just built into our workflow, right? Like someone has to review it. I d I don’t think there’s like an easy way to really manage that other than just make that a priority. But we are also an agency. We have always operated on a two quality control checks before anything goes to a client. So it’s just built already into what we do, like meaning someone was gonna proof it, someone was gonna so we’re just adding the agentics system into the loop and now two two people are gonna look at it, right? Like that’s just the now the the next phase of that.

Karin (54:18.478)
Okay.

Karin (54:34.328)
Okay.

Karin (54:42.284)
Yeah, yeah. Right. But you’re not removing the person like that was part of your process before and that doesn’t go away. It’s just easier and faster and you know. So yeah, I think

Jessica Aries (54:49.54)
No.

Jessica Aries (54:53.107)
Yeah. Less less time spent on the to tactical, right? And more time and then and then also that person who’s reviewing it, if they see something that consistently is wrong, we want them to go in and actually fix the skill or fix the agent that’s malfunctioning, right? Like don’t just let it continue to give you crap outputs, right? Like like go in and fix it, right? And I I think that’s also some of it too, like, having that workflow developed. but I think too

Karin (54:56.428)
Yes.

Karin (55:05.538)
Yeah. Yeah. The bad thing. Right. Yeah.

Jessica Aries (55:19.003)
I I don’t think you can really operate without humans and legal because we have such a fiduciary duty to our clients to make sure what we’re putting out there when we’re stamping our name on it, that it’s been reviewed, right? That I I think it’s very, it would be very unethical of me to say, like, don’t worry about it, just send right like, and so I think you have to have that in there. And you have to also realize the value of those people who are reviewing the work. They have deep subject matter expertise, and that’s why you want them reviewing the work versus somebody else, right? Because my coordinator.

Karin (55:33.57)
Yeah. Yeah.

Karin (55:43.042)
Yeah. Yeah. Yeah.

Jessica Aries (55:46.566)
is gonna think a social post looks great with emojis in it and with legal books and a gavel, right? Whereas I’m gonna be like, what the heck is that? Right? Like, why would you send that to the client? So I think that judgment layer, we didn’t talk a lot about that, but that’s another aspect of it, having that judgment and building that up and knowing what quality looks like.

Karin (55:52.206)
Right. Yeah. Yuck. Yeah, right. Exactly.

Karin (56:00.035)
Yeah. A hundred percent.

Well, and I think at the end of the day, this is really like when you look at the core of what we’re talking about as a service business, whether it’s marketing or legal services or you know, whatever it is, it is that judgment and expertise. And so you cannot remove that element of it. And anyone who really understands the value of what we are all selling as a service based business, you know, that question about like, is AI coming for all of our jobs?

I just find it slightly ridiculous. As much as I’m into AI and love all of this, there are things that I am optimizing and things that I am, you know, eliminating from my daily need to actually be a part of. But I’m still in there doing all this stuff. And like, you know, so the idea I I find that you know, it’s that’s something I’m even having conversations about or worried about in terms of like removing that human, like somebody coming from.

Jessica Aries (56:49.57)
yeah.

Karin (57:01.986)
Stop.

Jessica Aries (57:02.557)
Well, and who’s gonna tra who are who’s gonna be the next generation if we get rid of all the people who like I still need someone who’s eventually gonna sit in the seat and become our next client manager, right? And in order for them to be able to do that, they need to know how to talk to our clients. They to know what great looks like, right? They’re gonna there still has to be like a the conference I went to just last week was talking about the apprenticeship model, right? Of who’s that next generation that’s gonna lead? And the same is true of lawyers, right? Like if you get rid of all the associates, who’s gonna be the next generation of partners, right? Like

Karin (57:12.374)
Exactly. Exactly.

Karin (57:23.392)
Yeah. Yeah.

Karin (57:30.432)
Right. Exactly. Exactly.

Jessica Aries (57:32.281)
You won’t have a firm anymore if you get rid of all your associates, right? So that there’s a secession planning aspect to it.

Karin (57:35.841)
Exactly.

Exactly. All right. Well, I feel like we have covered so much. We’ve covered what you should be kind of doing next. Sit down, take a breath. It it is overwhelming and just kind of know that, know that all of us are overwhelmed, feel behind. But for me, I tend to kind of just say, Okay, I’m gonna stay in my lane. I’m not gonna try to learn how to do Claude for something that doesn’t relate to something that I need to have done. So I kind of stay in.

like in and I also have my platform of choice, which is Claude. And so I’m not gonna really I’ve kind of given up learning stuff on Chat GPT because it’s it’s just not I’m not I don’t have time for that. And so and then I do kind of what you were talking about. Good take 10, 15 minutes every day, week, every so often. I find a lot of really good stuff on YouTube where like good quick videos that just show a skill or something that I hadn’t really thought about before, some new thing everybody’s talking about.

Jessica Aries (58:27.773)
Yeah, me too.

Karin (58:36.07)
and so just kind of get your feet in there and be familiar with it and then develop some sort of a comfort level with it so that you can get in there and use it in in some kind of interesting different way.

Jessica Aries (58:48.423)
Yeah, I actually love that you mentioned picking one model, like one tool. Like I see too many people try to be the master of all of them and my gosh, right? Like what? Well, and every five days the UI changes on each of them. So like you lose every you can’t find the projects today because they’re now in a different codex is now here. And like, you know, so just yeah, pick one. Just you don’t need all of them. Trust me. It’s not like shoes.

Karin (58:51.512)
Yeah. Yeah.

Karin (58:57.326)
Yeah. It’s a mess. Yeah.

Karin (59:03.265)
Yeah, yeah.

No. Yeah. Pick one. Yeah. Yeah. Yeah. Yeah. I can’t imagine trying to keep that stuff organized either. Cause like I have like systems and stuff, and I it that’s just too much to organize. But anyway, all of that being said, Jessica Aries is the founder of By Ares, which is such an awesome, I feel like over the years of these podcast episodes, it it by Aries itself has evolved.

Jessica Aries (59:21.053)
Mm-hmm. Yeah.

Karin (59:34.854)
And it’s so fascinating to see that over the years, but it also kind of aligns with everything that’s been going on in legal marketing. So it’s it’s been so fun to have you on. And this is gonna be such a good episode. And I know people are gonna just like replay it because there’s so much good stuff in here that we covered. it was like five of your speech your talks that you do. but thank you so much for being here. Yeah.

Jessica Aries (59:34.93)
Yes.

Jessica Aries (01:00:03.451)
And I apologize to all the podcast listeners. That was like a that was like five workshops in one. So you there you go. Just listen to this podcast episode.

Karin (01:00:08.066)
There you go. There you go. Amazing. Yeah, there you go. Just send everybody here. We’ll get a ton of downloads. Yeah. Amazing. Thank you so much for being here. We will have you back at some point, maybe six to you know, sometime in the next year, and we’ll see what happens then. And who knows? Who knows? Maybe we’ll have the flying cars.

Jessica Aries (01:00:14.067)
Yeah.

Jessica Aries (01:00:26.299)
Yeah. We’re gonna have that that that that AI that does my laundry. That’s what I want. Why are you gonna make that happen? We need Rosie.

Karin (01:00:30.814)
Exactly. Rosie! We need the Rosie! Awesome. Thank you again for being here. This was such a fun episode.

Jessica Aries (01:00:39.794)
All right, thank you.