LongCut logo

AI Tools for Law Firms [Practical Steps to Learn AI from a Legal Tech Expert]

By The Law Firm Owner Podcast by Velocity Work

Summary

## Key takeaways - **AI Like Cognitive Prosthetics**: AI is like the cognitive version of the $6 Million Man, replacing limbs with super-powered prosthetics; it amplifies high-level thinking but low-level work will disappear. [00:16], [31:41] - **Join AI Workshops for Reps**: Learn AI like a new language by getting reps in conversations with other lawyers using it; Ernie's twice-weekly ChatGPT lab has at least 10 lawyers discussing live, with recordings and AI summaries. [04:07], [04:38] - **Mandate 15-20 Min Daily ChatGPT**: Richard Susskind advised a company to have everyone use ChatGPT 15-20 minutes daily for a month, even personally; they quickly turned a corner with breakthroughs as skills translated to business. [06:32], [07:05] - **Granola Delivers Instant Note Payoff**: Granola records meetings, generates powerful summaries incorporating your own notes, shares without audio for privacy, and doesn't join as a bot; everyone in the group adopted it after questions answered. [09:42], [10:05] - **Whisper Flow Fixes Dictation Friction**: Whisper Flow transcribes what you mean, not verbatim mess-ups, with smart formatting for bullets and lists; hold button for short bursts everywhere, making typing obsolete and AI prompts richer with context. [21:04], [23:50] - **Custom GPTs as AI SOP Agents**: Custom GPTs are AI's SOP equivalent: feed examples like past LinkedIn posts, it extracts essence for specialized tasks; name by task, @-invoke quickly, evolving to automatic agents amplifying firm workflows. [25:47], [28:00]

Topics Covered

  • Learn AI Through Lawyer Conversations
  • Granola Delivers Instant Note Payoff
  • Whisper Flow Eliminates Typing Friction
  • Custom GPTs Become Agentic SOPs
  • AI Replaces Cognitive Limbs

Full Transcript

They increased capabilities. There was a TV show back when I was a kid called the $6 million man. And you know, it was like he could do all this amazing stuff because he'd been in an accident. They

replaced like his arms and legs.

Basically, it was like Superman, but this is all because his limbs had been replaced. I think of AI as like the

replaced. I think of AI as like the cognitive version of that.

>> Welcome to the Law Firm Owner podcast powered by Velocity Work. This is the work that creates velocity.

Okay, Ernie Spinson, you're back for round two. Thanks. Thanks for being

round two. Thanks. Thanks for being here.

>> I'm ready. Let's Let's do it.

>> Okay. Well, last episode we talked about the LinkedIn post that you wrote that I loved and I wanted to have you on to talk more about it. And really the the gist of it is is that there needs to be

an identity shift for I mean I guess we're speaking specifically to law firm owners, but there needs to be an identity shift for law firm owners about how they're going to work and what

they're going to get paid for in the future and how their business needs to evolve. Um just to touch on a quick

evolve. Um just to touch on a quick couple things, the winners won't be traditional lawyers. They'll be

traditional lawyers. They'll be knowledge engineers, data scientists, and no coders. those are tomorrow's lawyers. That was something we chatted

lawyers. That was something we chatted about for a bit. And then also the three phases that that Richard Suskin talked about when it comes to technology in the firm and AI in the firm. So automation

was phase one, innovation is phase two, which is delivering outcomes in new ways and then elimination is phase three, just preventing the problem that people have uh before before it becomes a

problem. So, we we talked about that and

problem. So, we we talked about that and and I loved our conversation about that.

And then the natural followup is, okay, if we're all on the same page about that, we all believe this is true. How

do you take the steps you need to take?

What does that look like? And you know, before we record, we pushed record on this episode, you were saying, "I don't know how someone would do it alone. I

don't think it's that that doesn't feel like a a realistic way to approach this.

So maybe just high level say more about that and what you feel is important before we kind of get more into the nuts and bolts. Yeah. So uh the thing about

and bolts. Yeah. So uh the thing about learning AI or learning how to make use of it is you have to deal with the

reality that while like other technologies it's rapidly evolving, it's evolving much more rapidly than other technologies and it's radically

different in how it works and the the output the benefits are so massive. So I

analogize it to like it's like learning another [clears throat] language. You

have to learn another language and the best way to learn another language is to go speak it with other people who either already know it or are learning it like you are. And you just get those reps and

you are. And you just get those reps and you learn things constantly. And you

have to be learning this constantly because it's evolving constantly and very quickly. So it's more about the

very quickly. So it's more about the conversations that you have with other lawyers who are using it that will help you feel confident about you security

and privacy and methodology [snorts] and trying certain tools or not trying certain tools that supposedly hold great promise but in reality are

not for solar and small firm lawyers or not sufficiently evolved for you yet. I

mean, you have to allocate your attention and manage that resource, your attention. And the best way to know that

attention. And the best way to know that you're doing that well is to be talking to other people who are doing the same thing.

>> I mean, you're the inner circle seems like an obvious answer to me that that is a a space for those kinds of conversations and to immerse yourself in

the learning the new language. your

community and the way that it's set up.

How often are these conversations happening asynchronously? You did

happening asynchronously? You did mention through posts and and comments and things like that in the in the community. Yeah. I'd love to hear just a

community. Yeah. I'd love to hear just a little bit about what's what's the activity like or the engagement level.

How much time do people spend in here?

>> Yeah. So for the AI part in January I started what I called the workshop and it's an ongoing workshop and we meet twice a week on Tuesdays and Fridays for

an hour and some people come to one or the other some people come to both. Some

people come to a few for a while and drop out but there's an ongoing conversation. There's always at least 10

conversation. There's always at least 10 people there live discussing what's going on. Sometimes there's more but

going on. Sometimes there's more but usually it's 10. We record them. People

can watch the recording, read the transcript, or read the AIdriven summary. So, we we've been learning a

summary. So, we we've been learning a lot of the same things over time, but then we've also been learning new things. And the way that I started the

things. And the way that I started the the workshop is I call it the Chat GBT lab.

And I decided to focus on chatbt because I know that when you're trying to learn something, if you spread your effort across multiple places, you dissipate,

you know, the capability of learning. So

I said, look, if you want to learn Claude and that's similar enough to chat GBT, you can come in, you can analogize, but that's on you. We're going to focus on one tool so we can get everybody on

the same page as much as possible. And

that has worked really well. And then

that fact is what's recommended by you know the gurus who help people at the highest level which are Ethan Mhollik who's a professor warden who's written a

book about AI and he he's somebody you should follow on LinkedIn Alli K Miller same thing she's a Warden grad and she recorded a master class with Ethan

Mollik so they're on the same page although Ali K Miller is more about nuts and bolts of actual use by people in companies Mollik's more the researcher, the academic, but they're both using it.

They're both saying you have to use it and they give you guidelines and and ways of thinking about it. But they also said, you know, use one of them a lot

for everything. And it's funny at the

for everything. And it's funny at the talk that Suskan give, I I forget, I think this was in response to a question about, you know, how would we learn to use this? And he first of all said he

use this? And he first of all said he himself practices law and uses it in his practice. But he said in advising one of

practice. But he said in advising one of these companies he told one of them look forget about all the fancy tools just tell everybody

in your organization I want you to use chat tpt for at least 15 to 20 minutes a day for you know a month or six weeks and we don't even care if you use it for

law related stuff. We want you to use it. So if it's all personal, that's fine

it. So if it's all personal, that's fine because it really does translate. It's

not like if you learn to use it in a personal way that that's those skills don't translate to business. They do. So

it's the use. And he said that when this company did this, they very quickly turned a corner and had a big breakthrough because everyone was

learning to use it at a certain level and that gave them the leg up that they needed. So yeah, that's what people

needed. So yeah, that's what people need.

>> Yeah. Yeah, absolutely. Okay, there's

two things happening in my head right now. When I imagine law firm owners

now. When I imagine law firm owners getting better at implementing and leaning in to AI and where it's going, there's two things that strike me. One

is they need to spend time on it.

They're not going to get attention. They

need to spend attention on it, which in in most cases means they're going to need to spend some time on it. And they

can decide how they do that. or you know maybe they pay to have some assistance with that thinking and you know that's that's all all great but I do think that

people don't know what to spend their time on which being a part of a community can be helpful it can give you a bit of a guide to where to start but

it does seem like the busyiness of the dayto-day is easy to win out against spending time because this is even though it's probably not true. But what

people are perceiving is that this is this is important, not urgent. So they

may spend think that they need to spend time on it. They want to spend time on it and they don't make the time for it.

They don't carve out the space for it.

And so they're going to get left behind as as time go moves on. And this is true. I deal with this with whatever the

true. I deal with this with whatever the priorities are or what should be your attention has to be given there. There

has to be space in your calendar for those things. And so people number one,

those things. And so people number one, I would imagine you would agree, they need to make it a priority if they're going to have a shot at staying ahead of the game. This is almost like not um you

the game. This is almost like not um you can't read a book on how to swim. You

have to get in the pool. So that's one barrier that comes up that it almost seems like not worth talking about because of course, but I do think a lot of that is one of the main reasons

people will be left behind.

>> Oh no, it's worth talking about. But my

approach to this, because I've been watching lawyers deal with technology for a long time, was to say, okay, to me, I know cuz I use it, it has a

huge payoff. So, I took stock of like,

huge payoff. So, I took stock of like, well, where is that huge payoff going to be immediate and likely to be applicable to as many people as possible, namely

practicing lawyers? Because when people

practicing lawyers? Because when people get the payoff immediately, >> they they don't go, "Well, gee, convince me." Like they're convinced they're

me." Like they're convinced they're getting a payoff.

>> Yeah.

>> The areas where there's an immediate payoff are >> note takingaking. And there's a bunch of tools out there like Otter and Fireflies and Fathom.

>> But there's a newcomer that a lot of people don't know about now in our discussion group. Everybody knows about

discussion group. Everybody knows about this and everybody has adopted it because they all their questions have been answered. their fears, concerns,

been answered. their fears, concerns, whatever, and they go, "Oh, wow." And

then when they hear other lawyers saying, "Oh, yeah, now I'm using this and it's amazing." The tool's called granola.

And you don't even have to pay for it right now cuz they're doing beta testing for solo people. You do have to sign up with a Google account initially, but after that, that doesn't even matter.

So, some people got stuck with that hiccup. But once they get over the hump

hiccup. But once they get over the hump and they realize that wait, this thing can attend any meeting I come to, like I'm using it right now to record this because anytime I'm speaking into a mic

or listening, I can fire it up and it'll record. And it's useful because I can

record. And it's useful because I can see the transcript in case while I'm talking I forgot what we covered. But

more importantly, at the end of it, it will immediately create a summary. And

that summary is really powerful because I can I didn't have to strain to create notes.

>> The summary is there. Now, lawyers would say, "Okay, because I've worked through all these objections or questions."

Well, but wait, I want to take my own notes. Okay, well, go ahead. It's got a

notes. Okay, well, go ahead. It's got a pad. You can take your own notes in

pad. You can take your own notes in there, and those will be preserved as well. But more importantly, they'll be

well. But more importantly, they'll be incorporated into how the AI generates its notes because if it detects you're more interested in something than something else, it'll work that in. The

notes are immediately sharable with other people. It doesn't have the audio

other people. It doesn't have the audio recording, so that makes lawyers happy because it's not doing that. and you can opt out of training and all that good stuff. It's taking the world by storm.

stuff. It's taking the world by storm.

Everybody who starts using it goes, "Oh my god, it's the most amazing thing in the world." And it doesn't join a

the world." And it doesn't join a meeting as a bot. So that's an whole annoyance that's happening now with a lot of people. It just listens to you talk and here's what you hear. So

Granola super amazing and currently apparently still free.

>> I just last week signed up for it. But

by proxy, these conversations happening in in your community and Jeff Lewis is the one that's saying to me, you need to know about granola. [laughter] So then I look

about granola. [laughter] So then I look into granola. Yeah, we just signed up

into granola. Yeah, we just signed up for it last week. I actually haven't seen any um I think Friday was the first day that I used it. So I haven't I haven't dug in yet to see exactly and

I'm going to set up some templates on how I want it to structure notes. I

think that's possible. I haven't done that yet. send you a link to this this

that yet. send you a link to this this this one that's going to generate >> that will be useful because you know what we talked about and then you'll see the summary and think oh okay it got most of it right

>> it got one part a little wrong but that's it's always going to not be perfect right but your notes are not going to be perfect either so it's it's an incredible tool I just think everybody should use it and when you're

using that and you say okay AI can do that now you're starting to think summaries are good I can use them in other places without granola for other things. I recognize that yes,

they're not perfect, but they're useful.

So, that's one way I convince people to pay attention to stuff because they get the payoff. They you know

the payoff. They you know >> Yeah. Yeah. You know, when we um we host

>> Yeah. Yeah. You know, when we um we host our client retreats and nowadays, you know, they're virtual often. And with

this, the hope is that because we produce a debrief, we try to produce an artifact after the the retreat that they can have and be useful for them. And

there's a lot of work that goes into that debrief. But with tools like this,

that debrief. But with tools like this, it really lightens it. It speeds up how quickly we can get the product to them.

And for instance, we always have a section for um realizations and lessons learned from the previous quarter. Well,

now they're giving them to us, but our client services manager, she can type in after we're just wrapping up that section. She can type in, please give

section. She can type in, please give me, you know, the it's sort of the live interaction with the tool. Please give

me the list of accomplishments, lessons, and lessons learned and realizations.

and we could it's a really good just like with your LinkedIn post it's a really good first draft that we can finesse to get it good enough to be in their debrief but that used to take many

more minutes than it needs to nowadays and I just it just makes me wonder and these are the conversations you're having with lawyers in their work in the legal work how can they use this to speed everything up for themselves uh

and to lighten the the load >> well my LinkedIn post began with a granola recording So when I was sitting in the audience using my iPhone because I don't think you can use Android but I

was using an iPhone. There's an app and I was recording the sessions that I thought were interesting and that was one I recorded and it generates a

transcript which the fidelity can be good or bad or you know depending on how the sound quality is. that I was sitting in a good place. It was catching everything. But I also know that because

everything. But I also know that because Richard Suskin is Scottish, some of his annunciation might have presented some difficulties. You know, I know it's not

difficulties. You know, I know it's not going to be perfect. But what I did, cuz I've learned is I took the transcript, it catchts because the speaker was Scottish. I was

in this audience. You know, gave it all the context to understand this is not, you know, straight from a mic kind of transcript. that's got some other

transcript. that's got some other considerations. And I said, I want you

considerations. And I said, I want you to create as verbatim a transcript as you can and interpret things that maybe got mistranscribed. And then it asked

got mistranscribed. And then it asked me, you know, do you want me to add headings? And I'm like, yeah, sure, add

headings? And I'm like, yeah, sure, add headings. That's a good idea. So, it

headings. That's a good idea. So, it

adds all these headings and all these topics and broke it all down. I thought,

wow, that's it's very clean, very accurate now. And then I took that and

accurate now. And then I took that and fed that into another GPT I have that helps me write LinkedIn posts. And I

said, "Write a LinkedIn post off of this." So it had the best, highest

this." So it had the best, highest quality information. It knows how I want

quality information. It knows how I want to write a LinkedIn post and it knows things that matter to me. So those were all incorporated. And then I had this

all incorporated. And then I had this LinkedIn post within a matter of, you know, less than an hour cuz it it took a few steps. And then I realized that what

few steps. And then I realized that what people who went to that conference probably want is to be reminded of what he said in a very succinct way so that we could talk about it. Would it be

helpful for me to send him a link to the granola [clears throat] thing? No,

because that's too general. This it

needed to be refined. So when it was refined, that was useful and that's why it went viral.

>> Yeah. Did it go viral?

>> Oh yeah. I I had like I 45,000 views. I had 20 or 30 like within a

views. I had 20 or 30 like within a couple of days. It was insane.

>> That is amazing. Okay. I didn't I actually didn't know that, but now I know I wasn't the only one that loved it as much as I loved [laughter] it.

That's remarkable. Okay. Also, we're

thinking about how and steps people can take. I what a good point that there's

take. I what a good point that there's certain things where the payoff is immediate and it'll keep you engaged so that the things that are a little longer game you can make some headway on but

the immediate payoffs are are powerful.

That makes a ton of sense to me. Do you

think or or do you know of consultants or coaches out there that

will work with you for you on the thinking and planning of sort of the wave of how this can be implemented into your firm and or do you do any of that

work? Any one-on-one work if someone

work? Any one-on-one work if someone asks you? I do. But the way

asks you? I do. But the way >> I do it, because this is better, is it if there were two paths, me meeting with somebody oneon-one, >> once a week, a couple times a week, and

they got no exposure to the other lawyers, that would not be as good as exposure to the lawyers. The exposure to the other lawyers is the most important part. At some points, and I can see this

part. At some points, and I can see this with some people, they'll say, "Oh, you know, I'm having trouble with this."

I'll say, "Oh, wait. Let's get on a Zoom call, just you and me, and work through that." So, it gives me the opportunity

that." So, it gives me the opportunity to see who needs help and when they need help and what they need help with. And

that's a lot easier for me and them. And

I think it has to come off of this ongoing discussion with other lawyers because a couple of lawyers have said very emphatically that, you know, they they'll like not

participate in the discussions as much and I'll say, "Well, do you not want to share, you know?" Oh, no. Well, I just love hearing what other lawyers are saying. That's what helps me the most.

saying. That's what helps me the most.

And I thought, okay, yeah, that would help me, too, to hear them verbalizing why they're concerned about the privacy issue of something or other and how to deal with it. And if they get the answer they want and know that there's

consensus about it and that all the questions have been addressed, then they're free to go off and implement based off of what they heard. So, you

don't necessarily even have to participate. It's just hearing what

participate. It's just hearing what people are talking about and making sure that you're not, you know, off on some weird trail. Even if you're getting

weird trail. Even if you're getting great results, you you're still wondering like, am I about to step off a cliff?

>> Mhm.

Yeah. I

there's there's another question that isn't fully formed in my head, but about implementation.

I think you probably remember from way back we talked about turtle steps.

turtle steps for the win and it's just one foot in front of the other and you just commit to doing that one little thing every day for example. Um and how

it feels so seemingly insignificant in the moment but the stack of that adds up um to to something quite remarkable. So

for anyone listening, there is a propensity for most people to jump in and need eight hours that they think they need a weekend or they think they

need, you know, a retreat to just focus on these things. And not that that wouldn't be helpful, but I I'm just wondering if you can encourage if it feels true to you that it isn't that's

not the way forward. The way forward is consistent turtle steps. [laughter]

>> Yeah. There's there's a thing I saw on LinkedIn the other day and I screen captured it because every week I sent out an email and one of them is this is worth pondering and I usually have some

quote or something. This one's going to be used next week and I thought of you when I got it. It said, "If you feel like giving up because you're not seeing results, remember this. The last thing

to grow on a fruit tree is the fruit."

And I think that that's part of it. But

then again, I if if people can get a result or feel like that ongoing thing, they're seeing some fruit and not waiting for

it, that's good. So the granola example is one that's fruit that bears, you know immediately.

>> Mhm.

>> But then there's another one which I don't know if I've mentioned this one to you before, you've heard about it, but it's been adopted like wildfire in our group, which is called Whisper Flow. And

Whisper Flow is, it's easy to misunderstand this and think, "Yeah, no, I know what that is." Because it's voice to text. You dictate. It captures your

to text. You dictate. It captures your voice. It turns it into text. And people

voice. It turns it into text. And people

would say, "Well, I can dictate and attach EBT using a little microphone button." And that's good enough. And

button." And that's good enough. And

it's like, well, if you haven't tried something better, maybe it is good enough. But as soon as you try whisper

enough. But as soon as you try whisper flow where you hold down the button and dictate release it and it doesn't transcribe exactly what you say, it

transcribes what you meant to say. So if

you mess up in the middle and go, "No, I meant 10:00, not 11:00." It doesn't record all that verbatim cuz that's stupid. So it understands what you're

stupid. So it understands what you're trying to say and then it generates it.

And once you get in the habit, and it's a habit of learning to do it, and it works everywhere on your computer, then you realize typing is friction. You want

to stop typing, and you want to use this tool. And when you're using it with an

tool. And when you're using it with an AI like chat GBT or Claude, and you're having conversations very quickly back and forth, and you don't feel restricted in the ability to give context because

it's annoying to have to type all the things, you know that you want to say, so you just type some of them. The AI

works better when you give it more context. If you have total freedom to do

context. If you have total freedom to do that because you just hold down a button and blabber, it's better. And so, yeah, Whisper Flow has caught on. There's

other ones that are similar, but this one seems to be the one that has has caught, you know, fire in a greater way.

And I know why because it just works better. I tried the other one. It didn't

better. I tried the other one. It didn't

work as well. And this one is growing and it's I forget what it cost $12 a month. It works on your phone and I

month. It works on your phone and I think it works on Android too, but it definitely works on iPhones. And that's

where I've really had to like retrain my brain and I really hate typing on a phone. But as much as that's true, I

phone. But as much as that's true, I tended to fault and behave in the automatic robotic way. But once I've kind of trained myself not to do that because I'm like, look, if you just take

a few more steps and do this, you'll be able to get a better result. So yeah,

part of getting the fruit to grow on the tree is doing things differently, but you can get immediate benefits too. And

Whisper Flow, that's the tool to use.

>> Cool. Is it also for your uh laptop or No, it's just for >> Yeah. Yeah. On your laptop, you just

>> Yeah. Yeah. On your laptop, you just hold down the function key and talk or you can hit it twice and it locks it in and then you hit it again to release it.

But it's not for like I'm going to dictate, you know, 12 hours or I'm going to record Richard Suskin's speech. It's

not for that. It's for short burst. You

have a thought. You have something you want to say. You want an email you want to dictate. And the other thing I should

to dictate. And the other thing I should mention is it's smart about formatting.

So you don't have to do all this, you know, new paragraph, new line. It just

knows what to do and it'll sense when to create a bullet point or numbered list.

It's not perfect, but it's good enough and it's fast and it's yeah >> frictionless.

>> When you say short bursts, I mean, I'm curious. There's times where I have an

curious. There's times where I have an idea I need to flush out and so I will open up that little mic that I thought was good enough. Now I'm definitely

going to not think that it's good enough and just talk. So sometimes

that could be 10 minutes. Is that too long [laughter] >> or whisper? It might not be too long.

The thing is when you finish dictating and release the button, it then it's going to up to the cloud and maybe doing that while you're talking, but it takes

time to process it and it processes in one chunk based on when you release the thing. So, it's going to use the

thing. So, it's going to use the opportunity to understand the whole context and figure that out. So, it

takes it a little longer when you do it that way. So, what I find myself doing

that way. So, what I find myself doing again is I had to retrain my brain because it's like, oh, this is easy.

I'll just hold it down and blabber. I'll

find I get to a point where go, okay, I finished a good thought there. I'm going

to stop, let it generate, see what that looks like. Okay, fine. And then it's

looks like. Okay, fine. And then it's easy to pick back up from where I was.

So, it's a little more choppy, but it actually makes more sense to do it that way.

>> Okay. Yeah. No, that's that's great.

Okay. So, you've given us two two immediate payoff tools, which is Granola and Whisper Flow. If anything else comes to mind that's immediate payoff, definitely share.

>> Attach GPT or a client [laughter] for immediate payoff.

>> Fair. And then I think the other thing is what's top of mind for you that's that's not immediate payoff >> but worth at at some point you'll probably start

heading there once you've gotten some of the quick wins.

>> Yeah. So with ChatGpt there's a thing.

Now, this is if you pay, which I should have mentioned, you should be paying the $20 a month. You can create what are called custom GPTs. And I like to think

of these as sort of the AI equivalent of an SOP. We still need SOPs for things

an SOP. We still need SOPs for things that are done with, you know, physical action. And [snorts]

action. And [snorts] so SOPs is documented, you know, that way is not going away. But if for example like in writing my um LinkedIn post

>> that's an SOP. There's an SOP like Ernie writes his LinkedIn post like this.

Well, if I were to try to describe how I write them and get all the, you know, essence of it boiled down to some principles or whatever, how are people supposed to execute on that? How am

[clears throat] I supposed to execute on it?

>> But if I tell the AI, look, here are five AI LinkedIn posts that I've written that I think are the ones to use to model. tell me what you think are the

model. tell me what you think are the essence of these uh posts and it'll tell you and you go great I can see that makes sense or hey no I don't want you

to do that and then when it locks it in you can go and create a custom GPT and give it that instruction set and say when I feed you content you're [snorts]

going to output a LinkedIn post using this >> methodology but it's not a template it's not a set of instructions it's it's AI's

version of knowing how to do it. And a

lot of the things that we do are more like that than following some very rigid formulaic step-by-step process.

>> Is this different than I think it is different. I just don't I haven't I

different. I just don't I haven't I haven't messed around with custom GPTs, but how is it different than a project?

You give instructions in a project >> um within chatg.

>> Yeah. So they there there's overlap. A

project can have its own instructions.

Projects are useful if you want to collect chats about that topic. Whereas

the custom GPT is a standalone thing that you use. You can use a custom GPT inside of a project or you can make the project such that when you query it and

do things, it acts like a custom GPT.

And all of this is being is evolving.

So, I'm pretty sure that what's going to be happening is these custom GBTs are sort of agents, right? They do a very specialized thing. You call them up

specialized thing. You call them up manually now, but down the road it'll be automatically and they do their job. And

then maybe that job leads to another job that a different custom GPT needs to do.

And you need to train them in for these different types of jobs separately because when you start to muddy the waters and say okay you can do this but then I also want you to do this they start to fall apart right so they need

to be told to do very specific things so a skill and this is you know on the list of Richard Suskin's you know whatever he calls them you know we can label it

knowledge worker code no code coder is thinking oh this is a place where a custom GPT or a project would be useful

to me or to my team. I shall create this and then I shall improve it or my team will create it and we will improve it.

That like thinking like that instead of thinking, oh, I'll just do it or oh, I'll create an SOP or I'll I'll let them do it and I'll worry about it. No, you

need to be thinking about where you can create these agentic AI bots to do things for you.

>> Yeah, very cool.

uh more of a um in the weeds question, but when you make a custom GPT, does it show up somewhere >> in your in your sidebar?

>> Yeah. So, you there's a collection of them in the sidebar. They'll say GPTs and you can see some of them if you've created a bunch it'll put the most

popular ones that you use but you can call any of them very quickly using at and then you start to type the name just like in Slack or any of these tools where you >> and that's like a power move like that's

a thing that I remember hearing and then I forgot about it and then when I saw it used again I thought oh that's my solution because otherwise trying to go navigate manually laboriously to go find

that GP GPT is a pain in the ass, but when you can call it up very quickly inside of whatever chat you're in and then switch gears and go to a different GPT, that's gamechanging.

>> Do you do you name your names of like people?

>> Um, some people I've heard do that.

Whatever you can remember is the thing to do. For me, it's easier to remember

to do. For me, it's easier to remember what the task is, so I have them named that way. Like I was just thinking I

that way. Like I was just thinking I don't know for some reason with your LinkedIn stuff the word Lana came to mind Lana please.

>> [laughter] >> I don't know. And and also I mean this is something that I'm learning and I'll get better at, but you read a lot about

agents and you you just you've been we've been seeing things about like they are real team members along with your team and that you can incorporate them

in in ways which sounds crazy at the moment. I have a feeling a year from now

moment. I have a feeling a year from now it'll be laughable. Well, like of course that's what people do, you know, but um that's that's where my head went was

>> having names for these these bots.

>> And you said something about like they they increase capability. So there was a there was a TV show back when I was a kid called the $6 million man Lee Major and you know it was like he could do all

this amazing stuff because he'd been in an accident. They replaced like his arms

an accident. They replaced like his arms and legs with these prosthetic devices that were super powerful and he could run faster and he could jump higher and he could beat people up and lift heavy

objects. Basically, it was like

objects. Basically, it was like Superman, but this was all because his limbs had been replaced.

I think of AI as like the cognitive version of that. Like you can do more with your mind using these tools, but it has to amplify something. If what you're

giving it is low level, it's there's not ampl not a lot of amplification going on. But if you've thinking through and

on. But if you've thinking through and thinking better, >> it's amplifying that. So when lawyers say, "Well, I'm afraid my my job's going to disappear." It will if it's very

to disappear." It will if it's very low-level thinking that you're doing.

But if you're thinking at a higher level, you're going to be ahead of the tsunami.

>> This is so cool. I love talking about this stuff. I was just having a

this stuff. I was just having a conversation with a friend. Her kids are middle school going into high school and so AI is a big part of the conversation

and trying to the fear for for parents especially with kids that age is that the your brain doesn't have to work the way it used to have to work. And the

language that she used is mush. Like if

it if your brain doesn't have to do this then it doesn't have the ability to do that. Then it's

that. Then it's >> there's mush there. And

You know, I can see that because that's from our current position. Yes, I can see what what they're saying. But the

brain will be used differently. It's not

that it won't be used. It'll be used differently. And it I'm so curious what

differently. And it I'm so curious what that means for our evolution as a species, but things are happening faster and faster. So

and faster. So >> it's this that's such a fascinating uh conversation and just yeah the world

is full of wonder man >> I think the brain turning to mush is more of a problem if people are saying well I'm forced to do these things with

my brain which trains it to do certain things and if I'm no longer forced to do those things my brain will turn to mush [sighs and gasps] yeah I suppose that is true but like you can that right you're Right? You can use

your brain for different things. Like if

you want to learn to speak a language, there's AI tools to learn a language, you learn it faster, you want to learn to play bridge, there's tools that like learn to play bridge. You like you can

learn a lot of things, but if one is not curious and inquisitive and doesn't want to learn, I don't know.

>> Right. Right. I mean, my son who's five and a half, I downloaded this just thinking I was ahead of the game. Just

so funny how this works. I downloaded a a typing tutor.

>> Mhm.

>> Like he's My husband was like, "He's not going to need that. He's just not going to need that." And [laughter] I thought, "Wait a minute. Is that true?"

>> So anyway, >> he's still training his brain. So that's

good.

>> Yes, that's true. Yeah. So funny, man.

Ernie, thank you so much. what a a resource you are for for law firm owners and I'm so glad your community exists.

I'm so glad you were willing to come on and talk about your viral post. Very

cool. Yeah. Again, everyone can join your inner circle.

>> Yeah. It's easy to join.

>> Be a part sit in on these workshops.

Yeah.

>> Yeah. Come come be a part of it. You

know, you're you've been in the inner circle. I think you're still in there.

circle. I think you're still in there.

And um >> yeah, it's a good group. It's a good group of people. You know, if id had been only around lawyers this nice and and fun to be around, [laughter] I might

still be practicing law.

>> That is that's true. The the people you attract are pretty remarkable. That's

for sure.

>> Same for you. Same for you. That's why

we get along.

>> Yes. Yes. I I agree. Thank you for coming on. And next time we shouldn't

coming on. And next time we shouldn't wait so long to do this. It's been a while since I've had you on. I hope we can do this more.

Before you go, be sure to hit subscribe so you don't miss an episode. And if

you'd like to hear more on this topic, let us know in the comments below.

Thank you for [music] listening to the Law Firm Owner Podcast. If you're ready to get clearer on your vision, data, [music] and mindset, then head over to velocitywork.com where you can plug in

to quarterly strategic planning [music] with accountability and coaching in between. This is the work that creates

between. This is the work that creates velocity.

Loading...

Loading video analysis...