October 31, 2024 • 15 mins
Imagine having a virtual assistant tailored specifically for your business needs, enhancing efficiency and transforming operations. This episode promises to unravel the fascinating potential of custom GPTs, showcasing how they can serve as specialized experts across various business domains. We explore two dynamic methods: using custom instructions to imbue your GPT with a unique personality for diverse interactions, and fine-tuning to transform it into a domain-specific specialist. No technical background is required—just a clear vision of your business goals and the straightforward steps to gather and upload data.
Join us as we illuminate the pathway to smarter, faster business processes with the power of AI. We'll guide you through practical steps to identify bottlenecks and organize data, ensuring that your transition to AI-driven strategies is smooth and effective. With OpenAI's platform, even medium-sized businesses can affordably implement these solutions, testing and refining their custom GPTs in a matter of hours. Discover how this technology can predict customer needs, understand industry jargon, and ultimately set the stage for unparalleled growth and customer satisfaction.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Hey everyone, welcome back.
So you're already using chatGPT in your business.
That's fantastic, but now getthis.
Imagine if you could have likeyour very own version of chat
GPT one that's laser focused onyour business.
Speaker 2 (00:14):
Yeah.
Speaker 1 (00:14):
Like think of it as I don't know, like giving a rock
star new employee an instantcrash course in everything about
your company.
Speaker 2 (00:22):
Right.
Speaker 1 (00:22):
That's the power of custom GPTs, right?
You're essentially getting allthe brainpower of AI, but it's,
all you know, hyper-focused onexactly what you need.
Speaker 2 (00:32):
Yeah.
Speaker 1 (00:32):
Ready to like?
Jump in and make a differenceright away.
Speaker 2 (00:34):
Yeah.
Speaker 1 (00:35):
And here's the kicker it's not some crazy expensive,
time-consuming thing.
Speaker 2 (00:38):
No.
Speaker 1 (00:39):
It's actually shockingly easy and affordable.
Speaker 2 (00:41):
It really is.
I mean, we're talking about atotal game changer here, moving
from those you know, generic AItools to something that feels
like it was you know, custombuilt just for you, totally Okay
.
Speaker 1 (00:53):
So let's break this down a bit.
Yeah, how do we actually createthese custom GPTs?
Speaker 2 (00:55):
What are the like the nuts and bolts.
Sure, so there are two mainways to do it.
The first is using somethingcalled custom instructions and
the second is called fine tuning.
Speaker 1 (01:02):
Okay, think of custom instructions, like giving your
GPT a bit of a personality tweak, you know.
Speaker 2 (01:09):
Oh, interesting.
Speaker 1 (01:10):
You're basically saying hey, this is how I want
you to talk, this is the tone touse.
And here are some reallyimportant things to remember,
Like, for example, if your GPTneeds to be talking to your
engineers, you might tell it goahead and use all the technical
terms you want.
Speaker 2 (01:24):
Right.
Speaker 1 (01:24):
But if it's talking to customers, you probably want
to say keep it simple.
You know plain language please.
Speaker 2 (01:29):
Yeah, absolutely so.
It's like setting the groundrules for how your GPT interacts
with different folks.
Speaker 1 (01:36):
Exactly.
Speaker 2 (01:36):
Okay, that makes sense.
But what about this fine tuningthing?
That sounds a bit more intensemaybe.
Speaker 1 (01:43):
It is, and that's where it gets really interesting
.
Fine-tuning is like sendingyour GPT to a super specialized
training camp to become a totalexpert in your business, and you
do this by feeding it all yoursecret sauce, your proprietary
data.
Speaker 2 (01:59):
Stuff like your internal manuals, product specs,
customer service logs.
Even those brilliant strategiesyou jotted down on a napkin,
like years ago.
Speaker 1 (02:08):
So it's like uploading your company's DNA
into the AI.
Speaker 2 (02:11):
Exactly, and with fine tuning, the possibilities
are like almost endless.
You can have a GPT that knowsall your industry jargon,
understands your products insideand out and can even predict
what your customers need beforethey even ask.
Speaker 1 (02:26):
Now that sounds like straight out of a sci-fi movie,
but I gotta ask how complex isthis process really?
Like, do I need a whole team oflike data scientists to make
this happen?
Speaker 2 (02:37):
Not at all, and that's what's so great about it.
I mean, openai has made thisunbelievably user-friendly.
Speaker 1 (02:42):
Oh good.
Speaker 2 (02:43):
Really, you just gather your data, you upload it
to their platform and theyhandle all that crazy technical
stuff.
It's actually surprisinglystraightforward, even if you're
not a tech expert.
Speaker 1 (02:51):
That's a huge relief.
Speaker 2 (02:53):
Yeah.
Speaker 1 (02:53):
I mean, I'm already seeing how this could be like a
total game changer forbusinesses like ours.
But let's dig a little deeper.
Let's really get into thenitty-gr of the benefits.
You know the real meat andpotatoes.
What can we actually expect tosee?
Speaker 2 (03:08):
OK, let's say, for example, you're running a
manufacturing company.
You've got all this.
You know this complex machinery, all this specialized
terminology, mountains oftechnical documents Right yeah,
I get.
Get, well, a fine-tuned GPT.
It could be a total lifesaverfor your employees.
They could ask it anything fromhow do I fix this thing To
(03:29):
where is that part number inthis 500-page manual?
Speaker 1 (03:32):
So it's like having an on-call expert who knows
literally every single detail ofhow things work Right, and that
could save tons of time andfrustration.
I mean, we've all been there,right?
Oh?
Speaker 2 (03:43):
absolutely.
And it's not just about savingtime either.
It's about accuracy too, right?
Let's say, you're a financialadvisor, you know and you need
to give really precise advice.
It needs to be a hundredpercent compliant and, you know,
really tailored to each client.
Speaker 1 (03:58):
I see where you're going with this A custom GPT
trained on like all thefinancial regulations and all
your firm's specific investmentstrategies.
That could be incrediblyvaluable.
Speaker 2 (04:08):
Absolutely Like.
It's almost like having a superpowered compliance officer and
an investment analyst all rolledinto one.
Speaker 1 (04:14):
Exactly, and when you're dealing with, you know,
sensitive financial information,big decisions, that level of
accuracy, that consistency, it'spriceless.
And we can't forget about thecustomer experience, can we?
I mean, these days people wantyou know, instant answers, they
want that personalized touchyeah, absolutely, and a custom
(04:35):
GPT can actually make thathappen right it can like.
Speaker 2 (04:38):
Imagine our retail company right and they train a
custom GPT on their wholeproduct catalog, all the
customer feedback, even theirbrand guidelines exactly now
when a customer like has aquestion, they don't have to,
you know, wade through somegeneric FAQ page or wait on it
forever right.
Speaker 1 (04:57):
They just get like an instant answer, totally
accurate and, you know, tailoredto what they need and all in a
tone that fits the brandperfectly.
Speaker 2 (05:06):
It's almost like having a personal customer
service rep for every singlecustomer, but it's all powered
by AI, which is kind of crazywhen you think about it.
It is.
Speaker 1 (05:13):
Okay, so I think it's time to get down to brass tacks
here.
I know our listeners are allabout taking action.
Speaker 2 (05:18):
Sure.
Speaker 1 (05:19):
So let's give them a roadmap.
How do we get started withthese custom GPTs?
What's step one?
Speaker 2 (05:25):
Well, the first thing is you've got to figure out
what are your goals Like.
What do you actually want thiscustom GPT to do and who is it
going to be talking to?
Is it for your customers?
Is it for your team internally?
Speaker 1 (05:37):
So it starts with having a clear vision, right,
yeah, you know like, what are wetrying to achieve here?
Speaker 2 (05:41):
Exactly.
Speaker 1 (05:41):
Okay, so then we move on to the data.
Right yeah, curating the data.
What's the secret sauce there?
What do we need to know?
Speaker 2 (05:47):
The big thing is quality over quantity.
You don't have to, you know.
Just dump every single file youhave into the GPT and hope for
the best.
Think of it like you'retraining a new employee, right.
Speaker 1 (06:00):
Right.
Speaker 2 (06:01):
You wouldn't just hand them a stack of random
papers and say, ok, good luck,become an expert.
You'd give them the like, theessential stuff, the things they
really need to know to do theirjob.
Speaker 1 (06:10):
Okay, that makes sense.
So if I'm like a softwarecompany and I'm building a GPT
for, let's say, employeeonboarding, I'd want to feed it
things like our company handbook, training materials, all that
internal stuff about our cultureand how we do things.
Speaker 2 (06:27):
Exactly, and once you've got all the right data,
then you just upload it toOpenAI's platform and kick off
that fine-tuning process okay,and that's where it gets a
little little techie, right.
Speaker 1 (06:42):
Do I need to be a coding whiz to make this happen?
Speaker 2 (06:44):
not at all.
That's honestly like one of themost amazing things about this
whole thing okay, good.
I mean open AI has made it souser friendly.
It's like you know.
Using any other online tool,you upload your stuff, you tweak
a few settings and OpenAI justtakes care of the rest.
All their fancy algorithms doall the hard work behind the
scenes.
Speaker 1 (07:02):
So we don't need to go out and hire a bunch of data
scientists to get this up andrunning.
That's a relief.
Speaker 2 (07:08):
Nope, not at all.
And the speed is another hugeplus.
We're not talking about likemonths and months of development
here.
You could have a custom GPT,you know, ready to go in a few
hours, maybe even minutes.
Speaker 1 (07:19):
Wow, that's incredible.
Okay, so we've got our cleargoal, we've got our data all set
.
We've started the fine-tuningprocess.
What happens next?
Speaker 2 (07:27):
All right.
So now comes the really crucialpart Testing and iterating.
You know, this is where you getto see your custom GPT in
action.
Start by asking it somequestions, giving it some
prompts, seeing what it comes upwith.
Think of it like a trainingexercise.
You know You're helping your AIteammate reach its full
potential.
Speaker 1 (07:47):
I like that, like you're having conversation with
your custom GPT, right Seeinghow it thinks yeah, making some
adjustments as you go.
Speaker 2 (07:53):
Exactly and that back and forth.
It's really key to getting thebest results.
You might need to tweak thedata a bit, refine your prompts,
maybe even retrain the modelbased on you know what you're
learning.
Speaker 1 (08:06):
So it's not like a set it and forget it kind of
thing.
You got to stay involved.
You got to keep refining it.
You know, make sure it's reallyworking for you.
Speaker 2 (08:13):
Exactly, and once you're feeling good about how
your GPT is performing, then youcan start thinking about you
know how to actually integrateit into your business.
Speaker 1 (08:21):
Okay, so this is where things get really exciting
.
I mean, I can already see thepossibilities A custom GPT for
marketing, one for sales,another for customer support,
each one like a specialistworking 24 7 to make our
business like more efficient,more personalized and ultimately
, you know, more successfulthat's the whole point.
Speaker 2 (08:41):
It's not a one-size-fits-all thing.
It's about creating likebespoke AI solutions just for
you.
Imagine having a whole team ofvirtual assistants, each one a
total expert in their area,working around the clock to help
you reach your goals.
Speaker 1 (08:56):
And all of that without having to like, hire and
manage a whole bunch of peopledeal with payroll, all that
stuff.
Speaker 2 (09:02):
Exactly.
It's like having an army ofskilled employees, but they're
all AI powered.
Now I know some people mighthear all this and think well,
this is great for those bigcompanies with tons of money,
but what about us smaller guys?
And that's where I think one ofthe most exciting things about
custom GPTs comes in.
Speaker 1 (09:21):
You're talking about like accessibility?
Speaker 2 (09:23):
Yes, open AI has made this surprisingly affordable,
even for medium sized businesseslike yours.
I mean, the cost of fine tuningis really quite low and the
potential ROI it's huge.
Speaker 1 (09:35):
So it's not just about, you know, being on the
bleeding edge of tech.
It's about getting a realcompetitive advantage, no matter
what size your company is.
Speaker 2 (09:43):
Exactly, and the best part is you don't have to like
jump in all at once, right?
You can start small, maybe withjust one GPT for a specific
area of your business, and asyou start seeing results, then
you can gradually expand, buildout your whole AI ecosystem.
Speaker 1 (09:56):
Like that approach Start small experiment, learn
and then scale up when you'reready.
It's a really low risk, highreward kind of thing.
Speaker 2 (10:03):
It is, and let's be clear here this isn't just like
a fad or something.
This is a total game changer.
It's a fundamental shift in howbusinesses operate.
The companies that get on boardnow, the ones that invest in
custom GPTs they're going to bethe ones who define the future.
Speaker 1 (10:20):
So the message is loud and clear Don't wait, don't
hesitate, jump in and startexploring this world of custom
GPTs.
Speaker 2 (10:28):
Yeah.
Speaker 1 (10:28):
It's time to unlock the full potential of AI, you
know, really transform yourbusiness.
Speaker 2 (10:33):
All right.
So we've talked about whycustom GPTs are such a big deal
and we've walked through how toactually make them.
But for our listeners out therewho are thinking, ok, this
sounds amazing, but like wheredo I even start?
What's the first, you knowpractical step they can take?
Speaker 1 (10:47):
Well, I think it all begins with picking a problem.
You know, a specific problemthat you really want to solve.
Look at your business and thinkwhere are we spending way too
much time, or where are wemaking the most mistakes, or
even where are we hitting a walltrying to grow?
Speaker 2 (11:02):
Yeah, like zeroing in on those bottlenecks, those
pain points, the stuff.
That's like driving you crazy,exactly.
And once you've got like thatproblem in your sites, then you
can start thinking about OK, howcould a custom GPT actually fix
this?
Let's say, for example, you'rerunning an e-commerce business
right, medium size.
Speaker 1 (11:21):
Yeah.
Speaker 2 (11:21):
And you're getting like buried in customer service
emails.
Speaker 1 (11:25):
Oh, I feel that I bet a lot of our listeners can
relate to that one.
Speaker 2 (11:28):
Yeah, I mean, picture this right You've got a custom
GPT that's been trained on yourwhole product catalog, your
return policy, all those likeFAQs, right, and it can just
instantly answer questions,handle returns, even make like
personalized recommendations.
Speaker 1 (11:42):
It's like having like a whole team of customer
service superstars, all poweredby AI.
Speaker 2 (11:48):
Right, that's the idea, and that's just one
example.
The possibilities are reallyhuge, yeah, okay, so you've got
a clear picture of what you wantto do, then what You've got to
get that data ready.
Speaker 1 (11:58):
Yeah, we talked about it earlier, but any like pro
tips on actually picking theright data.
What do people need to keep inmind?
Speaker 2 (12:05):
Think of it as building a brain for your GPT.
What does it need to know to bereally good at its job?
If you're training a GPT tohelp with sales, you'd probably
want to give it you know allthose product descriptions,
sales scripts, testimonials fromhappy customers and you know
data on your biggest deals.
Speaker 1 (12:25):
So you're giving it the information that makes it an
expert right.
The better the data, the betterthe GPT.
Speaker 2 (12:30):
Yeah, exactly, yeah.
Oh.
And here's a little extra tiporganize that data, you know,
make it easy for the gpt todigest.
Clear headings, bullet points,all that good stuff.
It's like um, it's like givingyour gpt a textbook instead of
just a random pile of notes,right?
Speaker 1 (12:43):
right, that makes a lot of sense.
Okay, so we've got our problem.
Our data is all prepped.
What's next?
Speaker 2 (12:49):
time to upload all that data, to open a eyes
platform and kick off thefine-tuning process.
And this is where things get.
This is where the magic happensright.
Speaker 1 (12:57):
The magic of opening eyes, incredible tech that makes
all this surprisingly simple.
Speaker 2 (13:02):
Exactly.
You really don't have to belike a coding genius or anything
to do this.
Their platform is likeseriously intuitive, even if
you're not a tech expert.
You just upload your data, youchoose a few settings and let
OpenAI handle the rest.
Speaker 1 (13:16):
It's like having a whole team of brilliant AI
engineers working behind thescenes just for you.
Speaker 2 (13:23):
Right, and speaking of speed, you know how long does
this actually take?
Yeah, that's a good question.
It really depends, you know, onhow complex the task is and how
much data you have.
But a lot of times you can havea custom GPT ready to go in
just a few hours, sometimes evenless.
Speaker 1 (13:42):
Wow, that's incredible.
We're not talking about, like,months and months of waiting.
You know this is something anybusiness can do.
Speaker 2 (13:48):
It really is OK.
So now you've got your customGPT all trained up, right Now
it's time to put it through itspaces.
Speaker 1 (13:55):
Right, testing, refining, all that good stuff.
Make sure it's reallyperforming before we, you know,
set it loose.
Speaker 2 (14:01):
Exactly.
Think of it as like a dressrehearsal.
You want to make sureeverything's perfect before the
big show.
Start by, you know, asking itsome questions, give it some
prompts, see what it does.
Speaker 1 (14:12):
Yeah, like have a conversation with it Right, get
a feel for how it thinks.
Speaker 2 (14:16):
Exactly and don't be afraid to you know, try
different things, differentprompts, different data,
different settings.
See what works best.
Open AI gives you, like, a lotof flexibility to really fine
tune it until you're happy.
Speaker 1 (14:31):
So it's a, it's a process, right.
Speaker 2 (14:32):
Yeah.
Speaker 1 (14:32):
Constant learning and improvement.
Speaker 2 (14:34):
Absolutely, and that's what's so powerful about
it.
You're not just building a tool, you're building, like a
learning system, something thatcan adapt and evolve with your
business.
Speaker 1 (14:45):
So let's talk big picture for a second.
We've covered a lot of groundhere the steps, the benefits,
how easy it is to do, but whatdoes this all mean for the
future of business?
What's the takeaway here?
Speaker 2 (14:54):
I think this is like a fundamental shift.
Honestly, Custom GPTs, they'renot just a shiny new toy.
They're changing how we work,how we compete, how we create
value.
Speaker 1 (15:03):
Yeah, it's like the early days of the internet or
mobile phones.
You know it's that big.
Speaker 2 (15:09):
It is, and the companies that get this, the
ones that you know invest inthis and really embrace it.
They're going to be the oneswho define the future.
Speaker 1 (15:16):
Right.
So the message is loud andclear Don't wait, dive in,
explore this whole world ofcustom GPTs.
Speaker 2 (15:23):
Yeah, it's time to unlock the full potential of AI
and really transform yourbusiness.
The future's here and it'scustom made.
Speaker 1 (15:32):
And that wraps up our deep dive into the world of
custom GPT's.
We hope you found thisinsightful and you know, most
importantly, that you canactually use this information.
Speaker 2 (15:42):
Absolutely, and we encourage you to, you know,
check out the resources onOpenAI's website and start
playing around with it.
Yeah, give it a try, and weencourage you to check out the
resources on OpenAI's websiteand start playing around with it
.
Speaker 1 (15:50):
Yeah, give it a try.
And, as always, thanks forjoining us on the Deep Dive.
We'll catch you next time foranother exciting exploration.
Hey everyone, welcome back.
So you're already using chatGPT in your business.
That's fantastic, but now getthis.
Imagine if you could have likeyour very own version of chat
GPT one that's laser focused onyour business.
Speaker 2 (00:14):
Yeah.
Speaker 1 (00:14):
Like think of it as I don't know, like giving a rock
star new employee an instantcrash course in everything about
your company.
Speaker 2 (00:22):
Right.
Speaker 1 (00:22):
That's the power of custom GPTs, right?
You're essentially getting allthe brainpower of AI, but it's,
all you know, hyper-focused onexactly what you need.
Speaker 2 (00:32):
Yeah.
Speaker 1 (00:32):
Ready to like?
Jump in and make a differenceright away.
Speaker 2 (00:34):
Yeah.
Speaker 1 (00:35):
And here's the kicker it's not some crazy expensive,
time-consuming thing.
Speaker 2 (00:38):
No.
Speaker 1 (00:39):
It's actually shockingly easy and affordable.
Speaker 2 (00:41):
It really is.
I mean, we're talking about atotal game changer here, moving
from those you know, generic AItools to something that feels
like it was you know, custombuilt just for you, totally Okay
.
Speaker 1 (00:53):
So let's break this down a bit.
Yeah, how do we actually createthese custom GPTs?
Speaker 2 (00:55):
What are the like the nuts and bolts.
Sure, so there are two mainways to do it.
The first is using somethingcalled custom instructions and
the second is called fine tuning.
Speaker 1 (01:02):
Okay, think of custom instructions, like giving your
GPT a bit of a personality tweak, you know.
Speaker 2 (01:09):
Oh, interesting.
Speaker 1 (01:10):
You're basically saying hey, this is how I want
you to talk, this is the tone touse.
And here are some reallyimportant things to remember,
Like, for example, if your GPTneeds to be talking to your
engineers, you might tell it goahead and use all the technical
terms you want.
Speaker 2 (01:24):
Right.
Speaker 1 (01:24):
But if it's talking to customers, you probably want
to say keep it simple.
You know plain language please.
Speaker 2 (01:29):
Yeah, absolutely so.
It's like setting the groundrules for how your GPT interacts
with different folks.
Speaker 1 (01:36):
Exactly.
Speaker 2 (01:36):
Okay, that makes sense.
But what about this fine tuningthing?
That sounds a bit more intensemaybe.
Speaker 1 (01:43):
It is, and that's where it gets really interesting
.
Fine-tuning is like sendingyour GPT to a super specialized
training camp to become a totalexpert in your business, and you
do this by feeding it all yoursecret sauce, your proprietary
data.
Speaker 2 (01:59):
Stuff like your internal manuals, product specs,
customer service logs.
Even those brilliant strategiesyou jotted down on a napkin,
like years ago.
Speaker 1 (02:08):
So it's like uploading your company's DNA
into the AI.
Speaker 2 (02:11):
Exactly, and with fine tuning, the possibilities
are like almost endless.
You can have a GPT that knowsall your industry jargon,
understands your products insideand out and can even predict
what your customers need beforethey even ask.
Speaker 1 (02:26):
Now that sounds like straight out of a sci-fi movie,
but I gotta ask how complex isthis process really?
Like, do I need a whole team oflike data scientists to make
this happen?
Speaker 2 (02:37):
Not at all, and that's what's so great about it.
I mean, openai has made thisunbelievably user-friendly.
Speaker 1 (02:42):
Oh good.
Speaker 2 (02:43):
Really, you just gather your data, you upload it
to their platform and theyhandle all that crazy technical
stuff.
It's actually surprisinglystraightforward, even if you're
not a tech expert.
Speaker 1 (02:51):
That's a huge relief.
Speaker 2 (02:53):
Yeah.
Speaker 1 (02:53):
I mean, I'm already seeing how this could be like a
total game changer forbusinesses like ours.
But let's dig a little deeper.
Let's really get into thenitty-gr of the benefits.
You know the real meat andpotatoes.
What can we actually expect tosee?
Speaker 2 (03:08):
OK, let's say, for example, you're running a
manufacturing company.
You've got all this.
You know this complex machinery, all this specialized
terminology, mountains oftechnical documents Right yeah,
I get.
Get, well, a fine-tuned GPT.
It could be a total lifesaverfor your employees.
They could ask it anything fromhow do I fix this thing To
(03:29):
where is that part number inthis 500-page manual?
Speaker 1 (03:32):
So it's like having an on-call expert who knows
literally every single detail ofhow things work Right, and that
could save tons of time andfrustration.
I mean, we've all been there,right?
Oh?
Speaker 2 (03:43):
absolutely.
And it's not just about savingtime either.
It's about accuracy too, right?
Let's say, you're a financialadvisor, you know and you need
to give really precise advice.
It needs to be a hundredpercent compliant and, you know,
really tailored to each client.
Speaker 1 (03:58):
I see where you're going with this A custom GPT
trained on like all thefinancial regulations and all
your firm's specific investmentstrategies.
That could be incrediblyvaluable.
Speaker 2 (04:08):
Absolutely Like.
It's almost like having a superpowered compliance officer and
an investment analyst all rolledinto one.
Speaker 1 (04:14):
Exactly, and when you're dealing with, you know,
sensitive financial information,big decisions, that level of
accuracy, that consistency, it'spriceless.
And we can't forget about thecustomer experience, can we?
I mean, these days people wantyou know, instant answers, they
want that personalized touchyeah, absolutely, and a custom
(04:35):
GPT can actually make thathappen right it can like.
Speaker 2 (04:38):
Imagine our retail company right and they train a
custom GPT on their wholeproduct catalog, all the
customer feedback, even theirbrand guidelines exactly now
when a customer like has aquestion, they don't have to,
you know, wade through somegeneric FAQ page or wait on it
forever right.
Speaker 1 (04:57):
They just get like an instant answer, totally
accurate and, you know, tailoredto what they need and all in a
tone that fits the brandperfectly.
Speaker 2 (05:06):
It's almost like having a personal customer
service rep for every singlecustomer, but it's all powered
by AI, which is kind of crazywhen you think about it.
It is.
Speaker 1 (05:13):
Okay, so I think it's time to get down to brass tacks
here.
I know our listeners are allabout taking action.
Speaker 2 (05:18):
Sure.
Speaker 1 (05:19):
So let's give them a roadmap.
How do we get started withthese custom GPTs?
What's step one?
Speaker 2 (05:25):
Well, the first thing is you've got to figure out
what are your goals Like.
What do you actually want thiscustom GPT to do and who is it
going to be talking to?
Is it for your customers?
Is it for your team internally?
Speaker 1 (05:37):
So it starts with having a clear vision, right,
yeah, you know like, what are wetrying to achieve here?
Speaker 2 (05:41):
Exactly.
Speaker 1 (05:41):
Okay, so then we move on to the data.
Right yeah, curating the data.
What's the secret sauce there?
What do we need to know?
Speaker 2 (05:47):
The big thing is quality over quantity.
You don't have to, you know.
Just dump every single file youhave into the GPT and hope for
the best.
Think of it like you'retraining a new employee, right.
Speaker 1 (06:00):
Right.
Speaker 2 (06:01):
You wouldn't just hand them a stack of random
papers and say, ok, good luck,become an expert.
You'd give them the like, theessential stuff, the things they
really need to know to do theirjob.
Speaker 1 (06:10):
Okay, that makes sense.
So if I'm like a softwarecompany and I'm building a GPT
for, let's say, employeeonboarding, I'd want to feed it
things like our company handbook, training materials, all that
internal stuff about our cultureand how we do things.
Speaker 2 (06:27):
Exactly, and once you've got all the right data,
then you just upload it toOpenAI's platform and kick off
that fine-tuning process okay,and that's where it gets a
little little techie, right.
Speaker 1 (06:42):
Do I need to be a coding whiz to make this happen?
Speaker 2 (06:44):
not at all.
That's honestly like one of themost amazing things about this
whole thing okay, good.
I mean open AI has made it souser friendly.
It's like you know.
Using any other online tool,you upload your stuff, you tweak
a few settings and OpenAI justtakes care of the rest.
All their fancy algorithms doall the hard work behind the
scenes.
Speaker 1 (07:02):
So we don't need to go out and hire a bunch of data
scientists to get this up andrunning.
That's a relief.
Speaker 2 (07:08):
Nope, not at all.
And the speed is another hugeplus.
We're not talking about likemonths and months of development
here.
You could have a custom GPT,you know, ready to go in a few
hours, maybe even minutes.
Speaker 1 (07:19):
Wow, that's incredible.
Okay, so we've got our cleargoal, we've got our data all set
.
We've started the fine-tuningprocess.
What happens next?
Speaker 2 (07:27):
All right.
So now comes the really crucialpart Testing and iterating.
You know, this is where you getto see your custom GPT in
action.
Start by asking it somequestions, giving it some
prompts, seeing what it comes upwith.
Think of it like a trainingexercise.
You know You're helping your AIteammate reach its full
potential.
Speaker 1 (07:47):
I like that, like you're having conversation with
your custom GPT, right Seeinghow it thinks yeah, making some
adjustments as you go.
Speaker 2 (07:53):
Exactly and that back and forth.
It's really key to getting thebest results.
You might need to tweak thedata a bit, refine your prompts,
maybe even retrain the modelbased on you know what you're
learning.
Speaker 1 (08:06):
So it's not like a set it and forget it kind of
thing.
You got to stay involved.
You got to keep refining it.
You know, make sure it's reallyworking for you.
Speaker 2 (08:13):
Exactly, and once you're feeling good about how
your GPT is performing, then youcan start thinking about you
know how to actually integrateit into your business.
Speaker 1 (08:21):
Okay, so this is where things get really exciting
.
I mean, I can already see thepossibilities A custom GPT for
marketing, one for sales,another for customer support,
each one like a specialistworking 24 7 to make our
business like more efficient,more personalized and ultimately
, you know, more successfulthat's the whole point.
Speaker 2 (08:41):
It's not a one-size-fits-all thing.
It's about creating likebespoke AI solutions just for
you.
Imagine having a whole team ofvirtual assistants, each one a
total expert in their area,working around the clock to help
you reach your goals.
Speaker 1 (08:56):
And all of that without having to like, hire and
manage a whole bunch of peopledeal with payroll, all that
stuff.
Speaker 2 (09:02):
Exactly.
It's like having an army ofskilled employees, but they're
all AI powered.
Now I know some people mighthear all this and think well,
this is great for those bigcompanies with tons of money,
but what about us smaller guys?
And that's where I think one ofthe most exciting things about
custom GPTs comes in.
Speaker 1 (09:21):
You're talking about like accessibility?
Speaker 2 (09:23):
Yes, open AI has made this surprisingly affordable,
even for medium sized businesseslike yours.
I mean, the cost of fine tuningis really quite low and the
potential ROI it's huge.
Speaker 1 (09:35):
So it's not just about, you know, being on the
bleeding edge of tech.
It's about getting a realcompetitive advantage, no matter
what size your company is.
Speaker 2 (09:43):
Exactly, and the best part is you don't have to like
jump in all at once, right?
You can start small, maybe withjust one GPT for a specific
area of your business, and asyou start seeing results, then
you can gradually expand, buildout your whole AI ecosystem.
Speaker 1 (09:56):
Like that approach Start small experiment, learn
and then scale up when you'reready.
It's a really low risk, highreward kind of thing.
Speaker 2 (10:03):
It is, and let's be clear here this isn't just like
a fad or something.
This is a total game changer.
It's a fundamental shift in howbusinesses operate.
The companies that get on boardnow, the ones that invest in
custom GPTs they're going to bethe ones who define the future.
Speaker 1 (10:20):
So the message is loud and clear Don't wait, don't
hesitate, jump in and startexploring this world of custom
GPTs.
Speaker 2 (10:28):
Yeah.
Speaker 1 (10:28):
It's time to unlock the full potential of AI, you
know, really transform yourbusiness.
Speaker 2 (10:33):
All right.
So we've talked about whycustom GPTs are such a big deal
and we've walked through how toactually make them.
But for our listeners out therewho are thinking, ok, this
sounds amazing, but like wheredo I even start?
What's the first, you knowpractical step they can take?
Speaker 1 (10:47):
Well, I think it all begins with picking a problem.
You know, a specific problemthat you really want to solve.
Look at your business and thinkwhere are we spending way too
much time, or where are wemaking the most mistakes, or
even where are we hitting a walltrying to grow?
Speaker 2 (11:02):
Yeah, like zeroing in on those bottlenecks, those
pain points, the stuff.
That's like driving you crazy,exactly.
And once you've got like thatproblem in your sites, then you
can start thinking about OK, howcould a custom GPT actually fix
this?
Let's say, for example, you'rerunning an e-commerce business
right, medium size.
Speaker 1 (11:21):
Yeah.
Speaker 2 (11:21):
And you're getting like buried in customer service
emails.
Speaker 1 (11:25):
Oh, I feel that I bet a lot of our listeners can
relate to that one.
Speaker 2 (11:28):
Yeah, I mean, picture this right You've got a custom
GPT that's been trained on yourwhole product catalog, your
return policy, all those likeFAQs, right, and it can just
instantly answer questions,handle returns, even make like
personalized recommendations.
Speaker 1 (11:42):
It's like having like a whole team of customer
service superstars, all poweredby AI.
Speaker 2 (11:48):
Right, that's the idea, and that's just one
example.
The possibilities are reallyhuge, yeah, okay, so you've got
a clear picture of what you wantto do, then what You've got to
get that data ready.
Speaker 1 (11:58):
Yeah, we talked about it earlier, but any like pro
tips on actually picking theright data.
What do people need to keep inmind?
Speaker 2 (12:05):
Think of it as building a brain for your GPT.
What does it need to know to bereally good at its job?
If you're training a GPT tohelp with sales, you'd probably
want to give it you know allthose product descriptions,
sales scripts, testimonials fromhappy customers and you know
data on your biggest deals.
Speaker 1 (12:25):
So you're giving it the information that makes it an
expert right.
The better the data, the betterthe GPT.
Speaker 2 (12:30):
Yeah, exactly, yeah.
Oh.
And here's a little extra tiporganize that data, you know,
make it easy for the gpt todigest.
Clear headings, bullet points,all that good stuff.
It's like um, it's like givingyour gpt a textbook instead of
just a random pile of notes,right?
Speaker 1 (12:43):
right, that makes a lot of sense.
Okay, so we've got our problem.
Our data is all prepped.
What's next?
Speaker 2 (12:49):
time to upload all that data, to open a eyes
platform and kick off thefine-tuning process.
And this is where things get.
This is where the magic happensright.
Speaker 1 (12:57):
The magic of opening eyes, incredible tech that makes
all this surprisingly simple.
Speaker 2 (13:02):
Exactly.
You really don't have to belike a coding genius or anything
to do this.
Their platform is likeseriously intuitive, even if
you're not a tech expert.
You just upload your data, youchoose a few settings and let
OpenAI handle the rest.
Speaker 1 (13:16):
It's like having a whole team of brilliant AI
engineers working behind thescenes just for you.
Speaker 2 (13:23):
Right, and speaking of speed, you know how long does
this actually take?
Yeah, that's a good question.
It really depends, you know, onhow complex the task is and how
much data you have.
But a lot of times you can havea custom GPT ready to go in
just a few hours, sometimes evenless.
Speaker 1 (13:42):
Wow, that's incredible.
We're not talking about, like,months and months of waiting.
You know this is something anybusiness can do.
Speaker 2 (13:48):
It really is OK.
So now you've got your customGPT all trained up, right Now
it's time to put it through itspaces.
Speaker 1 (13:55):
Right, testing, refining, all that good stuff.
Make sure it's reallyperforming before we, you know,
set it loose.
Speaker 2 (14:01):
Exactly.
Think of it as like a dressrehearsal.
You want to make sureeverything's perfect before the
big show.
Start by, you know, asking itsome questions, give it some
prompts, see what it does.
Speaker 1 (14:12):
Yeah, like have a conversation with it Right, get
a feel for how it thinks.
Speaker 2 (14:16):
Exactly and don't be afraid to you know, try
different things, differentprompts, different data,
different settings.
See what works best.
Open AI gives you, like, a lotof flexibility to really fine
tune it until you're happy.
Speaker 1 (14:31):
So it's a, it's a process, right.
Speaker 2 (14:32):
Yeah.
Speaker 1 (14:32):
Constant learning and improvement.
Speaker 2 (14:34):
Absolutely, and that's what's so powerful about
it.
You're not just building a tool, you're building, like a
learning system, something thatcan adapt and evolve with your
business.
Speaker 1 (14:45):
So let's talk big picture for a second.
We've covered a lot of groundhere the steps, the benefits,
how easy it is to do, but whatdoes this all mean for the
future of business?
What's the takeaway here?
Speaker 2 (14:54):
I think this is like a fundamental shift.
Honestly, Custom GPTs, they'renot just a shiny new toy.
They're changing how we work,how we compete, how we create
value.
Speaker 1 (15:03):
Yeah, it's like the early days of the internet or
mobile phones.
You know it's that big.
Speaker 2 (15:09):
It is, and the companies that get this, the
ones that you know invest inthis and really embrace it.
They're going to be the oneswho define the future.
Speaker 1 (15:16):
Right.
So the message is loud andclear Don't wait, dive in,
explore this whole world ofcustom GPTs.
Speaker 2 (15:23):
Yeah, it's time to unlock the full potential of AI
and really transform yourbusiness.
The future's here and it'scustom made.
Speaker 1 (15:32):
And that wraps up our deep dive into the world of
custom GPT's.
We hope you found thisinsightful and you know, most
importantly, that you canactually use this information.
Speaker 2 (15:42):
Absolutely, and we encourage you to, you know,
check out the resources onOpenAI's website and start
playing around with it.
Yeah, give it a try, and weencourage you to check out the
resources on OpenAI's websiteand start playing around with it
.
Speaker 1 (15:50):
Yeah, give it a try.
And, as always, thanks forjoining us on the Deep Dive.
We'll catch you next time foranother exciting exploration.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Las Culturistas with Matt Rogers and Bowen Yang
Ding dong! Join your culture consultants, Matt Rogers and Bowen Yang, on an unforgettable journey into the beating heart of CULTURE. Alongside sizzling special guests, they GET INTO the hottest pop-culture moments of the day and the formative cultural experiences that turned them into Culturistas. Produced by the Big Money Players Network and iHeartRadio.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.