March 19, 2025 • 16 mins
Sales teams know AI can automate tasks, streamline workflows, and generate insights but most aren’t even scratching the surface of its full potential.
Reps and leaders keep hearing “use AI to sell smarter!” but what does that actually look like in practice? This episode is the tactical breakdown you’ve been waiting for.
Advanced prompting – How using high-quality AI models with better prompts can 10x productivity
Role-based AI agents – The game-changing way to eliminate manual tasks and optimize your workflow
Process automation – Why scaling efficiency isn’t about replacing humans, but working smarter
Jake Dunlap dives into real-world AI applications, from writing a five-page white paper in minutes to automating discovery call prep, and tackles the big question: At what point does AI-powered selling stop being smart and start feeling like cheating?
If AI is going to be your competitive advantage, it needs to be integrated into how you work today, not some future plan.
This episode is your step-by-step guide to making AI work for you right now.
Join the AI-Powered Seller Newsletter: https://bit.ly/ai-powered-seller-newsletter
Sign up for Custom GPTs: https://link.meetjourney.ai/aff_ad?campaign_id=4859&aff_id=2751&source=Jake-KD-Podcast&hostNameId=24435
Connect with Jake: https://www.linkedin.com/in/jakedunlap/
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
This week's episode we're going to jump literally
right into it.
We're going to talk abouteveryone is trying to figure out
tactically, jake, how do Iactually implement generative AI
?
Like, I've heard 55,000different things.
I'm a frontline sales rep.
Like, what the heck do Iactually do and why should I
give a shit right.
(00:20):
And so today's episode I'mgoing to dive in and help to
answer all of those questions.
Okay, so the first thing I'mgoing to talk about is how are
companies and how should you, asa sales rep, be thinking about
generative AI?
Okay, why does this matter?
What's actually possible, etc.
So I'm going to show a littlescreen share here.
(00:41):
So if you are following along inyour car, you're following on
the podcast, don't worry aboutit.
I'll describe it to you.
What's important to think about?
If you're sitting there, you'rein your seat, you're like Jake,
why does any of this matter?
The first thing is aroundmaking people aka me more
productive, right, and there's afew different ways that we can
do that, right.
(01:01):
There's a few different waysthat we think about ways that
organizations can do that, right.
There's a few different waysthat we think about ways that
organizations can do that.
Specific to these conceptsaround advanced prompting and
role-based agents, right?
So advanced prompting, okay,what the hell does that mean?
Well, what that means is thatyou're using the latest and
greatest AI, right?
A lot of you have kind ofstumbled through how to prompt
and, by the way, if you'restruggling, we actually have a
(01:23):
huge prompt library.
I will link to it in thepodcast.
So if you're somebody who'slike Jake I don't know how to
prompt man, I'm not quite sureno problem, like we got you
there.
But certain roles, if you wantto be more productive, you just
need to know how to prompt.
Let me show you guys this thisis the new ChatGPT01 Pro model.
Okay, and for those of youfollowing along, if you're just
(01:43):
listening, do not worry, I willdescribe it, hopefully well, but
I want you to look at thisexample.
This is something that Icreated with a client, literally
in about five minutes, and thisclient is in like the risk and
exposure space in medical device, like in in Europe, and so it's
like okay, with the O1 promodel.
So again, this is what I talkabout with advanced prompting.
(02:05):
You have to understand thedifferences between the models.
You might just think, oh, allgenerative AI is the same.
It's not Okay, the O1 Pro.
I'm telling you I have triedall the models and I'm not
saying for my sellers out there,you need to get the O1 Pro.
There's a brand new versioncalled the 4.5 model too.
(02:26):
That came out literally like aweek ago so you can use the, the
4.5.
The 01 pro is the best, but the01 pro is 200 bucks, so start
with the paid version.
You got to have the paidversion.
That's my other big piece ofadvice.
If you want to get good atadvanced prompting, you have to
have that.
So check this out.
Write a five-page white paperon product compliance in
relations to regulatory exposure.
It's around medical device andsustainability and the
(02:48):
differences between the UK andEurope.
Okay, how many of you have anyidea?
I do not have any idea whatthat means, right?
But if you're in that space,you do, and you can see here for
those of you who are followingalong on video In seven minutes
it cited 36 sources and wrote afive-page white paper.
(03:08):
So when I talk about advancedprompting, what I mean is that
this would have taken a typicalmarketing team I don't even know
five a month maybe.
So that's what I mean byadvanced prompting is in your
role.
You need to think about howmuch faster can you be if you
(03:29):
had a better model or betterprompts, okay.
The next way to think aboutdeploying AI is agents.
Okay, and again, most peoplewere not even not even there yet
.
Okay, but agents are.
If you look up here, you know.
Again, I've got a chat GPTpulled up.
You can see these little customGPTs on the side.
(03:50):
So when I refer to productivityagents, this is what I'm
talking about.
By the way, you can build theseout If your company uses Gemini
or Copilot or Perplexity.
This is possible there, but theconcept is something that
prompts you, okay.
So if you think about how todeploy AI, if you're sitting
there, you're a rep, step oneman.
(04:11):
If I used a slightly bettermodel, is this going to make me
exponentially more productive?
Because it's real.
This is real, real stuff as apart of this.
So think of things likediscovery call prep, which is
like the next phase, which islike, okay, I could sit there
and I can prompt.
Or, with discovery call prep,check this out, right, I can go
and it says great, hey, tell meabout you, I'm ready to get
(04:33):
started.
And again, instead of youhaving to prompt, it's done, all
the prompting on the back end.
So if you're wondering howthese custom agents work, all
the prompting is written intothe instructions.
So what it knows to do is tosay, oh, my job is to help Jake
with discovery call prep, so I'mtrying to prep for a meeting
with their VP of sales.
So, again, instead of me havingto learn how to prompt, it will
(05:15):
just go and do these things forme.
Again, it always wants theirLinkedIn profile.
So, okay, let's go ahead.
And hey, you know what you want, it, we'll get it.
Vp of sales on phase.
Okay.
So let's say we and hey, youknow what you want, it, we'll
get it.
Vp of sales on phase.
Okay.
So let's say we want to reachout to Colleen, for example,
here.
Right, so what you can do, justso you guys know how to do this
if you're, if you're curious,is all you really have to do is
(05:36):
just go to print page.
You just print page as PDF,okay.
So here's Colleen, she's the VPof Corporate Development.
So then I'll go back here andwe'll go to here you go, we'll
upload that.
Upload from computer.
We've got the LinkedIn profileas a part of it and I wanna
(05:59):
focus on.
Let's say, how we can helpoptimize sales strategies.
Okay, so what I want you tothink about is that's the beauty
in agents, right?
Imagine, like, how much lessthings I am actually typing and
instead all I'm doing is havingan agent prompts me.
So that's how all of you haveto think about this, the
beautiful part about this andthen check this out, check this
(06:22):
out Again.
It's gonna start to go throughall of this for you.
So there are these use casebased agents.
Think like annual report,summarizer, account plan builder
.
So the next thing that youshould think about is like, okay
, I could do advanced promptingand I have the best model and I
can create this really coolstuff.
Well then the next thing islike, well, is there an agent
(06:42):
for that that could just promptme, and then I don't need to
know how to prompt at all.
So again, it's telling youabout the role, how scaled can
help, right, and I'm going tosay you know, I want more on the
company, right?
So you can kind of prompt it alittle bit further and it will
just keep going and going andgoing.
So instead of having to writethese like absolutely amazing
(07:03):
prompts, I literally can justhave something that will prompt
me right.
And again, it's going tocontinue to go deeper and deeper
and deeper and you can tell itlike, hey, I like more discovery
call prep around this.
But it knows its whole goal isto help you do discovery prep.
So think of that.
It's like it's your use case.
It's like, hey, I'm gettingready to do this part of my job,
(07:24):
push a button right, and that'swhere these agents come in.
And again, it's like this was atype of prep that would have
taken me 20 minutes.
I now did it in what?
Two minutes here.
So that's what I mean when I sayrole-based agents.
That's what I mean.
So again, think about you know,rep, or if you're a company,
(07:51):
you're trying to figure out howto implement generative AI.
Step one for each role are wedo we have the right model?
And maybe each different rolesare going to need different
things.
Okay, great.
Step two, around peopleproductivity Do we have
role-based agents that we candeploy?
So, hey, my account managementteam.
We created one called like getup to speed on new account, and
it's like give a link and italready knows like your buyer,
personas and all those things.
You can then give it multiplecall transcripts and it's like
(08:11):
here's everything you need toknow about the account.
So, again, these agents justrequire a ton less prompting and
a ton less you know like needto customize as a part of it.
Last but certainly not least, isprocess automation Okay, and if
you're a rep out there, thereare there, believe it or not.
There are actually some thingsthat you can build on your own.
I'd highly suggest this is foreverybody.
(08:34):
I don't care if you're a VP ofsales, a CEO, a sales rep
listening to this.
You need to just know what makean N8NR.
These are the tools that we useto really build a lot of our
you know kind of AI automationsaround.
You know we focus on thesetypes of tools.
(08:55):
So imagine a world where thenext variation is I see a new
meeting with Colleen and it saysonphasecom.
It hits my Gcal with Colleenand it says onphasecom.
It hits my Gcal.
I then create a web hook thatsees that domain runs the
discovery call prep agent, gpt,and then automatically puts that
into Salesforce.
Right, so that is the future.
(09:18):
Like we're building one rightnow called proposal.
Let's call it proposal SOWbuilder and what it does?
What it does is it has awebhook to Zoom.
So Zoom integrates, it, pullsthe recording, then pulls the
transcript from the recording,then it runs it through Proposal
Builder agent.
It then slacks you two or threefollow-up questions to make it
(09:42):
better.
Once you answer the Slack, itimmediately will email or Slack
you over a completed statementof work based on the call and
your additional insights.
So the next next level is well,the agents are kind of running
behind the scenes.
We're looking for all thedifferent things and then we're
just automating.
So if you're a rep out there oryou're a company and you're
(10:03):
thinking about how to deploythis is how you have to think
about it.
Right, stop trying to thinkabout what's our generative AI
strategy?
What should we do?
Like, oh my gosh, it's a roleand process specific.
So whenever we're working withcompanies, this is kind of our
checklist is again and, by theway, all of you can do this for
(10:24):
yourself right, audit yourselfand say, okay, what am I doing
right now?
Okay, well, I've been doingthis.
Okay, I want to get better.
I want to make sure I canprompt as a part of this, or I
might want to make sure, like,I'm using the best model and the
reason that you have to useadvanced prompting and you can't
just go to agents is some ofthe best models aren't available
(10:44):
through the API to createagents.
That's really what it is,whereas.
So the agents are really good,but again, like the white paper
example, well, I can't use thatO1 pro model in an agent, so it
actually is probably better forme to just be good at prompting
for now.
So just think about it likethat it's like man, is this
(11:06):
possible?
Like, let me get really good atdoing this advanced prompting,
let me try some of thesedifferent models and see what
works for me.
Okay, great.
And then it's hey, I'm doingthis in my day to day, you know.
Again, step one is do your audit.
Is there an agent for this?
You know like, is somebodybuilding something to where?
Like, is there an agent forthis?
You know like, is somebodybuilding something to where,
like, I don't have to do this.
(11:27):
And then, last but certainlynot least, is just getting
familiar with what processautomations look like.
You know, for me.
I mean, everybody is going tobe doing this.
This isn't a matter of you knowwhen you're going to do, or you
know if you're going to dosomething.
Here.
It's a win, 100% win.
So again, if you're a frontlineseller, this is a or or a
(11:47):
company.
This is how you get started,right.
You get started by gettingbetter at prompting.
Like I said, we'll put a linkto the prompt library that we
have down below.
Definitely check it out.
We'll also link to some of theagents that we built and other
people that are doing agents.
You know any of you can go andtry to build your own custom GPT
.
Maybe we'll do a whole episodeon that where we'll kind of talk
(12:12):
through how to do your buildyour own custom GPT and you know
what it's supposed to do, etcetera.
Um, and then process automationis the future.
It's not.
Well, the funny part is Ialways say this I'm like it's
actually not the future, likeit's here right now, um, but I
think for a lot of people it'slike whoa, whoa, whoa, like we
need to start back here.
So for everybody out there,it's a little quick hitter
episode this week.
If you're trying to figure outhow to get better or where to
(12:35):
start when it comes togenerative AI, this is how you
need to think about it.
Okay, prompting is V1, findingagents to replace the prompting
and then finding automations toreplace the agents, and that's
the future.
It's gonna be exciting, youknow.
I think a lot of people, a lotof leaders I talk to her, you
know nervous or scared about.
I was on a call yesterday and aleader said this feels like
(13:01):
cheating.
She's like I don't wanna tellmy team to go do advanced
prompting or use these agentsbecause then they won't learn
how to think.
Let me tell you this right nowthe only skills that are gonna
be left for humans in the futureis creativity and critical
thinking.
The key is to understand youneed to use AI to get to V1 in
25 minutes instead of a week.
(13:23):
Right, I showed you that whitepaper example.
Guys, you can build a five-pagewhite paper with 36 cited
sources in seven minutes.
Like it's not cheating, becausethen I'm going to spend four
hours making it amazing.
Ah, that actually doesn't fit.
This works better here.
I know the audience.
I move this thing here.
Amazing.
Ah, that actually doesn't fit.
This works better here.
I know the audience, I movethis thing here.
(13:43):
So the future is the ability toleverage AI and then turn on
your strategic thinking, turn onyour creativity.
You know we're training teams.
Look, you're not.
You're not copying and pastingthe AI.
Instead, what you're doing isyou're you know you're
leveraging it to get to V1 andV2 and come up with a couple of
new ideas.
You know, is it cheating?
Let me ask you this Is itcheating to get to a better
(14:05):
answer if you ask your coworkerwhat they you know their opinion
on an idea?
Is that cheating?
Is it cheating to usegenerative AI to come up with a
couple of ideas and then youmake them amazing?
Nobody's saying you have to usegenerative AI.
Nobody's saying you have to usegenerative AI.
So I want you to think.
Anybody out there who's worriedabout is this cheating?
The future is you have to be.
This stuff is moving so quicklyRight now.
(14:25):
I think the best estimate iswhat generative AI is capable of
is already two years ahead ofwhere our brain can comprehend
what it can do.
And so if you're not startingwith advanced prompting right
now and understanding, if you'renot starting to look for role
agents, you are gonna be doing amountain of work that nobody
else is gonna do because they'relike yeah, I just go ahead and
(14:47):
use this and then I get aheadreally quickly and then I spend
less time and my version isactually better than yours
because I've had more time torefine it, and so this is what
everybody out there needs to bethinking about is I have to
deploy this stuff now?
Right, what I'm giving you isthe 101, the 201 and the 301.
What I the interesting part isthe 301 is already here.
(15:09):
You can these things that we'retalking about.
We're building this stuff rightnow, and you can build it too.
This stuff does not require youknow advanced computer science
as a part of this.
So that's this week's episode.
A little quick hitter here, notas long as our typical ones, but
I wanted to get somethingreally tactical out for
everybody out there who's likeJake.
(15:29):
Why is this stuff important?
What are the nuances about it,how should I think about this,
or frame why I should use it,when I should use it and look in
every episode.
My goal is to give a ton andton of tactical advice, so make
sure to drop a comment onYouTube if you're watching this,
or on your favorite podcastplatform.
(15:51):
Make sure to DM me on LinkedInor any of these platforms, and
happy to tackle whatever topicsare top of mind for you, but for
me, this should be thefoundational episode, so if
you're trying to think, where doI start?
What do I need to do, how do Iget over that mindset of this is
cheating, or should I do thisnow?
Hopefully this episode is goingto get you moving.
(16:12):
So, thanks everybody,appreciate you tuning in and we
will see you all in a couple ofweeks on the next episode of AI
Powered Seller.
Thanks everybody.
This week's episode we're going to jump literally
right into it.
We're going to talk abouteveryone is trying to figure out
tactically, jake, how do Iactually implement generative AI
?
Like, I've heard 55,000different things.
I'm a frontline sales rep.
Like, what the heck do Iactually do and why should I
give a shit right.
(00:20):
And so today's episode I'mgoing to dive in and help to
answer all of those questions.
Okay, so the first thing I'mgoing to talk about is how are
companies and how should you, asa sales rep, be thinking about
generative AI?
Okay, why does this matter?
What's actually possible, etc.
So I'm going to show a littlescreen share here.
(00:41):
So if you are following along inyour car, you're following on
the podcast, don't worry aboutit.
I'll describe it to you.
What's important to think about?
If you're sitting there, you'rein your seat, you're like Jake,
why does any of this matter?
The first thing is aroundmaking people aka me more
productive, right, and there's afew different ways that we can
do that, right.
(01:01):
There's a few different waysthat we think about ways that
organizations can do that, right.
There's a few different waysthat we think about ways that
organizations can do that.
Specific to these conceptsaround advanced prompting and
role-based agents, right?
So advanced prompting, okay,what the hell does that mean?
Well, what that means is thatyou're using the latest and
greatest AI, right?
A lot of you have kind ofstumbled through how to prompt
and, by the way, if you'restruggling, we actually have a
(01:23):
huge prompt library.
I will link to it in thepodcast.
So if you're somebody who'slike Jake I don't know how to
prompt man, I'm not quite sureno problem, like we got you
there.
But certain roles, if you wantto be more productive, you just
need to know how to prompt.
Let me show you guys this thisis the new ChatGPT01 Pro model.
Okay, and for those of youfollowing along, if you're just
(01:43):
listening, do not worry, I willdescribe it, hopefully well, but
I want you to look at thisexample.
This is something that Icreated with a client, literally
in about five minutes, and thisclient is in like the risk and
exposure space in medical device, like in in Europe, and so it's
like okay, with the O1 promodel.
So again, this is what I talkabout with advanced prompting.
(02:05):
You have to understand thedifferences between the models.
You might just think, oh, allgenerative AI is the same.
It's not Okay, the O1 Pro.
I'm telling you I have triedall the models and I'm not
saying for my sellers out there,you need to get the O1 Pro.
There's a brand new versioncalled the 4.5 model too.
(02:26):
That came out literally like aweek ago so you can use the, the
4.5.
The 01 pro is the best, but the01 pro is 200 bucks, so start
with the paid version.
You got to have the paidversion.
That's my other big piece ofadvice.
If you want to get good atadvanced prompting, you have to
have that.
So check this out.
Write a five-page white paperon product compliance in
relations to regulatory exposure.
It's around medical device andsustainability and the
(02:48):
differences between the UK andEurope.
Okay, how many of you have anyidea?
I do not have any idea whatthat means, right?
But if you're in that space,you do, and you can see here for
those of you who are followingalong on video In seven minutes
it cited 36 sources and wrote afive-page white paper.
(03:08):
So when I talk about advancedprompting, what I mean is that
this would have taken a typicalmarketing team I don't even know
five a month maybe.
So that's what I mean byadvanced prompting is in your
role.
You need to think about howmuch faster can you be if you
(03:29):
had a better model or betterprompts, okay.
The next way to think aboutdeploying AI is agents.
Okay, and again, most peoplewere not even not even there yet
.
Okay, but agents are.
If you look up here, you know.
Again, I've got a chat GPTpulled up.
You can see these little customGPTs on the side.
(03:50):
So when I refer to productivityagents, this is what I'm
talking about.
By the way, you can build theseout If your company uses Gemini
or Copilot or Perplexity.
This is possible there, but theconcept is something that
prompts you, okay.
So if you think about how todeploy AI, if you're sitting
there, you're a rep, step oneman.
(04:11):
If I used a slightly bettermodel, is this going to make me
exponentially more productive?
Because it's real.
This is real, real stuff as apart of this.
So think of things likediscovery call prep, which is
like the next phase, which islike, okay, I could sit there
and I can prompt.
Or, with discovery call prep,check this out, right, I can go
and it says great, hey, tell meabout you, I'm ready to get
(04:33):
started.
And again, instead of youhaving to prompt, it's done, all
the prompting on the back end.
So if you're wondering howthese custom agents work, all
the prompting is written intothe instructions.
So what it knows to do is tosay, oh, my job is to help Jake
with discovery call prep, so I'mtrying to prep for a meeting
with their VP of sales.
So, again, instead of me havingto learn how to prompt, it will
(05:15):
just go and do these things forme.
Again, it always wants theirLinkedIn profile.
So, okay, let's go ahead.
And hey, you know what you want, it, we'll get it.
Vp of sales on phase.
Okay.
So let's say we and hey, youknow what you want, it, we'll
get it.
Vp of sales on phase.
Okay.
So let's say we want to reachout to Colleen, for example,
here.
Right, so what you can do, justso you guys know how to do this
if you're, if you're curious,is all you really have to do is
(05:36):
just go to print page.
You just print page as PDF,okay.
So here's Colleen, she's the VPof Corporate Development.
So then I'll go back here andwe'll go to here you go, we'll
upload that.
Upload from computer.
We've got the LinkedIn profileas a part of it and I wanna
(05:59):
focus on.
Let's say, how we can helpoptimize sales strategies.
Okay, so what I want you tothink about is that's the beauty
in agents, right?
Imagine, like, how much lessthings I am actually typing and
instead all I'm doing is havingan agent prompts me.
So that's how all of you haveto think about this, the
beautiful part about this andthen check this out, check this
(06:22):
out Again.
It's gonna start to go throughall of this for you.
So there are these use casebased agents.
Think like annual report,summarizer, account plan builder
.
So the next thing that youshould think about is like, okay
, I could do advanced promptingand I have the best model and I
can create this really coolstuff.
Well then the next thing islike, well, is there an agent
(06:42):
for that that could just promptme, and then I don't need to
know how to prompt at all.
So again, it's telling youabout the role, how scaled can
help, right, and I'm going tosay you know, I want more on the
company, right?
So you can kind of prompt it alittle bit further and it will
just keep going and going andgoing.
So instead of having to writethese like absolutely amazing
(07:03):
prompts, I literally can justhave something that will prompt
me right.
And again, it's going tocontinue to go deeper and deeper
and deeper and you can tell itlike, hey, I like more discovery
call prep around this.
But it knows its whole goal isto help you do discovery prep.
So think of that.
It's like it's your use case.
It's like, hey, I'm gettingready to do this part of my job,
(07:24):
push a button right, and that'swhere these agents come in.
And again, it's like this was atype of prep that would have
taken me 20 minutes.
I now did it in what?
Two minutes here.
So that's what I mean when I sayrole-based agents.
That's what I mean.
So again, think about you know,rep, or if you're a company,
(07:51):
you're trying to figure out howto implement generative AI.
Step one for each role are wedo we have the right model?
And maybe each different rolesare going to need different
things.
Okay, great.
Step two, around peopleproductivity Do we have
role-based agents that we candeploy?
So, hey, my account managementteam.
We created one called like getup to speed on new account, and
it's like give a link and italready knows like your buyer,
personas and all those things.
You can then give it multiplecall transcripts and it's like
(08:11):
here's everything you need toknow about the account.
So, again, these agents justrequire a ton less prompting and
a ton less you know like needto customize as a part of it.
Last but certainly not least, isprocess automation Okay, and if
you're a rep out there, thereare there, believe it or not.
There are actually some thingsthat you can build on your own.
I'd highly suggest this is foreverybody.
(08:34):
I don't care if you're a VP ofsales, a CEO, a sales rep
listening to this.
You need to just know what makean N8NR.
These are the tools that we useto really build a lot of our
you know kind of AI automationsaround.
You know we focus on thesetypes of tools.
(08:55):
So imagine a world where thenext variation is I see a new
meeting with Colleen and it saysonphasecom.
It hits my Gcal with Colleenand it says onphasecom.
It hits my Gcal.
I then create a web hook thatsees that domain runs the
discovery call prep agent, gpt,and then automatically puts that
into Salesforce.
Right, so that is the future.
(09:18):
Like we're building one rightnow called proposal.
Let's call it proposal SOWbuilder and what it does?
What it does is it has awebhook to Zoom.
So Zoom integrates, it, pullsthe recording, then pulls the
transcript from the recording,then it runs it through Proposal
Builder agent.
It then slacks you two or threefollow-up questions to make it
(09:42):
better.
Once you answer the Slack, itimmediately will email or Slack
you over a completed statementof work based on the call and
your additional insights.
So the next next level is well,the agents are kind of running
behind the scenes.
We're looking for all thedifferent things and then we're
just automating.
So if you're a rep out there oryou're a company and you're
(10:03):
thinking about how to deploythis is how you have to think
about it.
Right, stop trying to thinkabout what's our generative AI
strategy?
What should we do?
Like, oh my gosh, it's a roleand process specific.
So whenever we're working withcompanies, this is kind of our
checklist is again and, by theway, all of you can do this for
(10:24):
yourself right, audit yourselfand say, okay, what am I doing
right now?
Okay, well, I've been doingthis.
Okay, I want to get better.
I want to make sure I canprompt as a part of this, or I
might want to make sure, like,I'm using the best model and the
reason that you have to useadvanced prompting and you can't
just go to agents is some ofthe best models aren't available
(10:44):
through the API to createagents.
That's really what it is,whereas.
So the agents are really good,but again, like the white paper
example, well, I can't use thatO1 pro model in an agent, so it
actually is probably better forme to just be good at prompting
for now.
So just think about it likethat it's like man, is this
(11:06):
possible?
Like, let me get really good atdoing this advanced prompting,
let me try some of thesedifferent models and see what
works for me.
Okay, great.
And then it's hey, I'm doingthis in my day to day, you know.
Again, step one is do your audit.
Is there an agent for this?
You know like, is somebodybuilding something to where?
Like, is there an agent forthis?
You know like, is somebodybuilding something to where,
like, I don't have to do this.
(11:27):
And then, last but certainlynot least, is just getting
familiar with what processautomations look like.
You know, for me.
I mean, everybody is going tobe doing this.
This isn't a matter of you knowwhen you're going to do, or you
know if you're going to dosomething.
Here.
It's a win, 100% win.
So again, if you're a frontlineseller, this is a or or a
(11:47):
company.
This is how you get started,right.
You get started by gettingbetter at prompting.
Like I said, we'll put a linkto the prompt library that we
have down below.
Definitely check it out.
We'll also link to some of theagents that we built and other
people that are doing agents.
You know any of you can go andtry to build your own custom GPT
.
Maybe we'll do a whole episodeon that where we'll kind of talk
(12:12):
through how to do your buildyour own custom GPT and you know
what it's supposed to do, etcetera.
Um, and then process automationis the future.
It's not.
Well, the funny part is Ialways say this I'm like it's
actually not the future, likeit's here right now, um, but I
think for a lot of people it'slike whoa, whoa, whoa, like we
need to start back here.
So for everybody out there,it's a little quick hitter
episode this week.
If you're trying to figure outhow to get better or where to
(12:35):
start when it comes togenerative AI, this is how you
need to think about it.
Okay, prompting is V1, findingagents to replace the prompting
and then finding automations toreplace the agents, and that's
the future.
It's gonna be exciting, youknow.
I think a lot of people, a lotof leaders I talk to her, you
know nervous or scared about.
I was on a call yesterday and aleader said this feels like
(13:01):
cheating.
She's like I don't wanna tellmy team to go do advanced
prompting or use these agentsbecause then they won't learn
how to think.
Let me tell you this right nowthe only skills that are gonna
be left for humans in the futureis creativity and critical
thinking.
The key is to understand youneed to use AI to get to V1 in
25 minutes instead of a week.
(13:23):
Right, I showed you that whitepaper example.
Guys, you can build a five-pagewhite paper with 36 cited
sources in seven minutes.
Like it's not cheating, becausethen I'm going to spend four
hours making it amazing.
Ah, that actually doesn't fit.
This works better here.
I know the audience.
I move this thing here.
Amazing.
Ah, that actually doesn't fit.
This works better here.
I know the audience, I movethis thing here.
(13:43):
So the future is the ability toleverage AI and then turn on
your strategic thinking, turn onyour creativity.
You know we're training teams.
Look, you're not.
You're not copying and pastingthe AI.
Instead, what you're doing isyou're you know you're
leveraging it to get to V1 andV2 and come up with a couple of
new ideas.
You know, is it cheating?
Let me ask you this Is itcheating to get to a better
(14:05):
answer if you ask your coworkerwhat they you know their opinion
on an idea?
Is that cheating?
Is it cheating to usegenerative AI to come up with a
couple of ideas and then youmake them amazing?
Nobody's saying you have to usegenerative AI.
Nobody's saying you have to usegenerative AI.
So I want you to think.
Anybody out there who's worriedabout is this cheating?
The future is you have to be.
This stuff is moving so quicklyRight now.
(14:25):
I think the best estimate iswhat generative AI is capable of
is already two years ahead ofwhere our brain can comprehend
what it can do.
And so if you're not startingwith advanced prompting right
now and understanding, if you'renot starting to look for role
agents, you are gonna be doing amountain of work that nobody
else is gonna do because they'relike yeah, I just go ahead and
(14:47):
use this and then I get aheadreally quickly and then I spend
less time and my version isactually better than yours
because I've had more time torefine it, and so this is what
everybody out there needs to bethinking about is I have to
deploy this stuff now?
Right, what I'm giving you isthe 101, the 201 and the 301.
What I the interesting part isthe 301 is already here.
(15:09):
You can these things that we'retalking about.
We're building this stuff rightnow, and you can build it too.
This stuff does not require youknow advanced computer science
as a part of this.
So that's this week's episode.
A little quick hitter here, notas long as our typical ones, but
I wanted to get somethingreally tactical out for
everybody out there who's likeJake.
(15:29):
Why is this stuff important?
What are the nuances about it,how should I think about this,
or frame why I should use it,when I should use it and look in
every episode.
My goal is to give a ton andton of tactical advice, so make
sure to drop a comment onYouTube if you're watching this,
or on your favorite podcastplatform.
(15:51):
Make sure to DM me on LinkedInor any of these platforms, and
happy to tackle whatever topicsare top of mind for you, but for
me, this should be thefoundational episode, so if
you're trying to think, where doI start?
What do I need to do, how do Iget over that mindset of this is
cheating, or should I do thisnow?
Hopefully this episode is goingto get you moving.
(16:12):
So, thanks everybody,appreciate you tuning in and we
will see you all in a couple ofweeks on the next episode of AI
Powered Seller.
Thanks everybody.
Popular Podcasts
Bookmarked by Reese's Book Club
Welcome to Bookmarked by Reese’s Book Club — the podcast where great stories, bold women, and irresistible conversations collide! Hosted by award-winning journalist Danielle Robay, each week new episodes balance thoughtful literary insight with the fervor of buzzy book trends, pop culture and more. Bookmarked brings together celebrities, tastemakers, influencers and authors from Reese's Book Club and beyond to share stories that transcend the page. Pull up a chair. You’re not just listening — you’re part of the conversation.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.