Everyday AI Podcast – An AI and ChatGPT Podcast: EP 131: Build a Brand Live with AI (Part 2)

This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its

power to your fingertips.

Listen daily for practical advice to boost your career, business, and everyday life.

People often ask, what can artificial intelligence do for my business?

This whole AI thing doesn't really seem like for us or for me.

Well, I think we showed you that yesterday.

General AI can not only help your business, but you can grow your entire company or grow

a business with AI.

And we're following up with part two today.

Thank you for joining us.

My name is Jordan Wilson, and this is Everyday AI.

We are your daily livestream, podcast, and free daily newsletter, helping everyday people,

you and me, just learn AI, but how we can actually leverage it.

So that's what building a brand with AI, parts one and parts two, is all about.

So I'm excited for the follow-up of part one.

So before we get into that, let's first go over the AI news.

And hey, if you're joining us live, let me know.

I always like to know where you're joining us from.

And if you're joining us on the podcast, don't worry, check the show notes.

We're going to have yesterday's great show in there as well as other important topics

for you to cover.

So let's talk what's going on in the world of AI news first.

So first, will AI finally be regulated by the US government?

So the Biden administration is setting up to unveil a long-awaited executive order

on artificial intelligence.

So this new executive order from the Biden White House aims to regulate AI technology

and address potential risks it poses to society.

So the order will require government agencies to assess and incorporate AI into their work,

ease barriers for skilled workers, and increase a focus on cybersecurity.

Some government officials and lawmakers are concerned about the potential role of AI

in job displacement and the impacts on privacy and democracy.

My take, I don't think this is really going to be anything.

We've talked about this on the show for long enough.

Right now, generative AI and artificial intelligence in general means way too much

for the US economy for any sort of actual regulation or legislation.

All right, next in AI news.

So a chipmaker startup has grabbed $27 million to help the generative AI chip shortage.

So the new startup named CENTML has secured $27 million in US in seed funding from investors

such as Google's, Google's Gradient Ventures, NVIDIA, Deloitte Ventures, and Thompson-Rudder's Ventures.

So the startup CENTML is developing a solution to help companies facing AI hardware shortages

by optimizing their ML models.

So they're taking a little bit different approach, kind of using AI in the hardware

to help make more of these chips that are powering the generative AI movement.

All right, and then last, but not least, in AI news, Facebook is reportedly bringing AI writers

to your posts.

Yay. All right, so multiple users on social media have reported a new right with AI feature on Facebook.

This is similar to a feature that LinkedIn rolled out a couple of months ago,

where essentially you can enter, you know, when you go to write a new post on Facebook,

you'll be able to enter bullet points or a natural language prompt, something like,

you know, write a post about why AI posts will create a lot of mediocre content on social media,

and you hit enter and it writes it for you.

So no word yet on official release.

All right, are y'all ready to build something live with AI?

I hope you are.

As a reminder, if you missed yesterday's show, make sure to go to youreverydayai.com.

Sign up for the newsletter.

We've got it all there.

You can always go back and read all of our old newsletters.

But let's, let's real quick, before we do a quick recap of what we covered yesterday,

I want to shout out everyone joining us and thank you.

Thank you for joining us.

And let me know, I always like to know, especially when there's people that,

you know, have been saying, hey, been watching the live stream for months.

It's like, all right, shout it out.

Shout it out.

Let me know where you're joining us from.

So Maybrit said, early check-in yesterday's episode was jaw-dropping.

So excited about this one.

You're giving super insightful tips on how to use chatGBT for better marketing win-win.

Yeah, I'm going to, I'm going to recap real quick.

If you didn't catch yesterday's episode, don't worry.

I'm going to recap very quickly everything that we did and everything that we created.

Brian just said, Brian, thanks for joining us.

He said, just revisited part one this morning.

This everyday AI is next level.

Can't wait for part two.

All right.

So thank you all for joining us.

Let's get to it.

Actually, someone is asking, I have a question about ClickUp AI.

What is your hot take on that?

We use, we use ClickUp actually, haven't used the AI much.

We have, I think we pay like, you know, $2 a month or something to use it.

Haven't used it much, but let's talk about what we did.

Yesterday on the show, I, I had a great time yesterday building a brand from scratch,

but let's, let's quickly go over and recap really what we did.

All right.

So here's what we did.

We created a brand literally from scratch.

It was great.

So we started the show by asking our live audience again,

if you're joining us on the podcast, this is going to be one of the ones

click on the, click on the notes in the show description, watch this video.

I'm going to try to do my best to describe, you know, everything we do live,

but it's going to be best to watch this.

But, you know, yesterday we created a brand live from scratch.

Thanks to your suggestions.

So we had different suggestions for different products that we could create.

And we ended with a stable sip.

So a kind of more portable, you know, travel mug, you know,

I could, I could be using it right now.

Here I got my, my clear mug and it doesn't, I mean,

it's actually kind of stable.

It's got a wide bottom, right?

So we created a brand called stable sip

and we used a chat GPT pretty much exclusively to do this whole first half.

We're going to use a little bit of chat GPT today as well,

but we're going to be using some other tools.

So stick around for that.

So we also identified competitors for our brand

and we helped create stable sips unique selling proposition.

So this is great.

And we use chat GPT for a lot of use cases yesterday

that I don't think people understand when talking about building a brand

or even how you can grow your company.

People think chat GPT is just something that you go into to write content.

Like, I need to, you know, bang out this, you know, presentation quick.

So let me feed it some old content and have it updated.

All right.

I want a blog post about, you know, traveling to Tahiti, whatever.

You know, people think that's what chat GPT is for.

And I'll tell you guys, I don't really use chat GPT for that.

Hardly at all.

That's probably less than 10% of what, you know, myself and even our team

uses chat GPT for is content creation.

Yes, it can do that.

Yes, that is the low hanging fruit or the obvious use case

for a large language model like chat GPT.

But that's, I mean, it's, I always make the comparison of a cell phone, right?

That's like if you're using a brand new iPhone and you buy the newest one

and literally all you do is make a phone call, right?

You don't do anything else.

No text messaging, not taking photos, not using apps.

That's, that's the equivalent of using chat GPT just for content creation.

So one of the best things, and I think what we showed yesterday

through this was specifically using plugin packs, right?

So using plugins mode and giving chat GPT access to the internet

and using proper prompting to help it research our competitors,

to help develop USPs, all of these things that would normally take

a lot of people or maybe a whole separate company.

Obviously, we didn't go into great depth because we were doing it live.

We were working from your feedback, going back and forth,

you know, kind of throwing out polls.

Hey, which one do you like more?

But I think it was a great use case of what chat GPT should actually be used for

and not just for creating mediocre content.

So we also use chat, chat GPT to identify potential customers.

We did craft some marketing copy and we're going to wrap that up today,

actually.

And we also created a logo inside chat GPT with Dolly free.

So huge, huge momentum yesterday by creating a literal brand from

stretch going from ideation strategy, you know, working on market position,

everything, you know, things that you maybe might not be using

generative AI for right now, but you probably should.

All right.

So that is a quick recap of yesterday.

All right.

I'm going to, I'm going to do something here.

I'm going to take a sip because yesterday I was, I was tired after the show.

I was tired.

It was a long one.

So I got my, my extra caffeine and I'm ready to get going on part two.

What about you all?

Like what are you, you know, for our live audience,

what are you looking forward to learning today in this process of building

a brand live?

I'm excited.

Let's get going, y'all.

Let's get going.

So here's what we're going to do.

And again, today it's, it's going to be a lot of jumping around from program to

program FYI.

So please, please excuse any delay that we might have as I'm sharing

different screens.

So I'll try to, I'll, I'll try to also do my best for our podcast

audience to describe what's going on.

Okay.

So I am jumping in chat GPT.

I have a prompt up.

I just spit it out.

I'll quickly explain what we're doing.

We did kind of end the show with saying like, Hey, write some marketing copy.

Right.

But that was just to explain something, but now we're actually going to use this,

this marketing copy.

So what I'm doing is I'm telling chat GPT.

So I'm saying use the, using the above or using the above, please perform a

competitive analysis on our competitors websites.

Huge, huge use case for chat GPT.

This is something I use all the time for competitive analysis.

If you want to run a competitive SWAT, right?

Strengths, weaknesses, opportunity threats, like you can hire a management

consult, consultancy company to do that, or you can do a very basic,

but very powerful one in chat GPT just by running a prompt like this.

And then I'm saying after a full analysis, right.

So synced marketing copy for landing page for steady sip.

Again, that is our brand that we created from scratch.

Steady sip.

I think I have our logo pulled up here.

Here we go.

Here's our logo.

We'll get more into that here in a minute.

So steady sip.

Essentially, if you are joining us new, joining us for the first time and you

didn't jump in yesterday, it is a different kind of travel mug.

It is geared more toward being very on the go, not, not spilling.

So a sturdy base, super secure vacuum sealed top, something that's on the go and

really gearing it toward kind of people that live kind of almost like a remote

lifestyle, right?

So those people that work hard at home and then they go in the office and maybe

go to the gym or they're traveling and working.

So that's kind of who we're, you know, one of our target demographics.

So essentially I'm asking though, so you don't have to read this,

this prompt kind of line by line.

I'm asking for five different options for each section of marketing copy on a website.

That's a great thing.

It's something a lot of people don't do.

They just ask for one.

You know, hey, write me an about page from my website.

All right.

Well, let's give some more information to chat.

Let's give it access to the internet so it can read our website or read our

competitors website so we can find what is unique about our product that we're

building.

I'm also telling chat to use some copywriting techniques such as problem

match station solution features and benefits and data.

All right.

So you'll see right here.

Chat, you can always check.

I told it to use browser up.

Someone asked this question yesterday because I do have at all times I always use

three plugins.

Usually I have two separate internet connected plugins browser up and box

script.

And then I'll have a different plugin enabled as well, just depending on the

use case.

So right here it's doing a competitive analysis.

It did it just right here.

So different kind of travel mugs or different popular mugs.

So mirror, I think it is here's yeti.

Most of us know yeti.

Here's hydro flask.

I think that's what this one is.

This large, my comically large water mug.

And then clean canteen and ember.

So here we go.

Here is our copy, right?

So I'm sharing it on the screen.

And if y'all can read this, let me know what you think.

So I'm asking for a CTA subhead features, who it's for, and about steady sip.

So there's an extra bold approach.

There is a minimalist approach.

And this is for the copy, right?

So as an example, I'm going to read just the CTA, right?

Or maybe the CTA and the subhead.

So extra bold and let me know which one.

If you're joining us live for our live audience, let me know which one is best

and we'll use this one.

That's the thing.

We're building this live.

I'm not choosing.

Y'all are choosing.

So let me know which one you like best.

So extra bold approach.

The CTA is grab your steady sip now.

The last travel mug you'll ever need.

Minimalist approach.

Discover steady sip.

Simplicity meets functionality.

Adventurous approach.

Embark with steady sip.

Your adventure companion.

Tech savvy approach.

Get smart with steady sip.

The future of beverage containers.

Eco-friendly approach.

Choose sustainable.

Choose steady sip.

Good for you.

Good for the planet.

All right, which one, which one do y'all like?

I, I haven't been paying attention to the comments.

Let me jump in.

Which one of these should, should we use?

What do you guys,

what are you guys thinking?

And also good morning, everyone joining us live.

Gabe, thanks for joining us.

LinkedIn user.

Sorry, whoever's joining us from Allentown PA.

Thanks, Val.

What's going on?

Woozy.

Thank you for joining us.

What do you guys think?

Which, which one of those,

which one of those should we be using?

Let me know.

Personally, personally, I'm leaning toward the,

the minimalist approach.

I don't know if anyone else has any thoughts.

Jay says he likes tech savvy, Gabe says, the one that says,

grab your steady sip now.

That's the extra bold approach, eco-friendly.

Okay. Well, we, we have a couple of votes.

We have a couple of votes.

So I think we had a kind of a vote for four different ones.

So I'm going to say we're going to run with the extra bold approach.

So I'm going to say, okay, for the extra bold approach.

Let me say, please use this.

And write content for the entire website and all relevant sections.

All right.

That's just doing,

we're just doing a very quick one right there.

Okay.

So we will check back on our marketing copy.

That was kind of first on our list.

So as a reminder, here's what we did yesterday.

We created a brand.

We identified competitors and stable Sips USPs.

We identified potential customers.

We crafted marketing copy a little,

and that's what we're wrapping up now.

And then we created a list of the brands that we created.

And then we created a logo with Dolly.

All right.

So as a reminder, here's our logo.

So now here's, here's what we're going to do today.

Right.

So we already just started.

We're showing up our marketing copy with chat GPT.

Right.

Then we're going to ask next.

We're going to ask chat GPT for some image prompts in mid journey.

And this is where I need y'all.

This is where I need to hear from you all to say, all right,

what should we be, what should we be doing here?

So I'm going to get this ready, this prompt here.

So we have to wait for chat GPT to finish our,

to finish our text here.

It's not, again, sometimes you'll see here.

I just did a quick prompt and it's not writing.

It's just doing it section by section.

So we'll have to go and clean it up when it's done,

but I want to move on.

All right.

So now what I'm doing is I'm asking essentially for an image prompt.

All right.

That's another fun thing that you can do inside chat GPT is actually

generate text based prompts that you can then use inside,

you know, something like mid journey or something like dolly.

So right now I just put in a big prompt.

It said, based on all this, please describe the actual stable sit mug

as specifically as possible.

This is a longer prompt, but I'm essentially steering chat GPT

to create a simple text based image prompt

that I can use in multiple programs.

And we're going to throw this in.

And this is kind of the backbone, right?

This is like, we got to get some strong product images.

So we're going to get some AI generated images here.

So I'm going to go ahead and we're going to start throwing these in

kind of one by one.

So sorry, here we go.

So we were in this one, I'm grabbing this first prompt

and I'll just, I'm not going to read these all,

but this one says this text prompt for an AI image says

that I generated with chat GPT based on all of the content

that was already in our chat, right?

So chat GPT can go back about 8,000 tokens,

which is about 6,400 words.

So this text prompt says straightforward studio lighting,

product photo shoot for stable sip, a matte black travel mug

with spill proof technology, matte black finish,

wide base spill proof lid, ergonomic handle,

stainless steel interior, no logo, stable sip only.

So I kind of said in the prompt,

I didn't want like 50 other things in the image.

So that's what I was trying to get towards.

So now I'm going to jump in.

I'm going to go into Dolly first.

Okay.

So now we're in Dolly and I just put that,

just copy and pasted that prompt straight in there.

All right.

And we're going to, we're going to jump around here today, y'all.

We're going to do a lot of jumping.

I'm going to try to do my best to explain what's going on.

All right.

So I'm taking that same prompt now,

that same prompt that I just read.

I just opened mid journey.

Mid journey is also fantastic tool.

I think by far it is probably the leader in AI image generators.

Dolly actually is not too far behind,

but I did the exact same prompts.

And let me see if I can resize my window here so I can see.

There we go.

And the difference is all I have to do in mid journey,

if you're new, you just type dash backslash.

Imagine that it says enter your prompt.

I entered it and bam.

And mid journey, and I'll kind of go over the pros and cons,

but let me know which one you guys are liking so far.

So mid journey is kind of a still finished rendering.

We have four different options.

So mid journey gives us four different options.

Let's jump back into, let's jump back into Dolly.

Dolly's taking it sweet time today.

That's okay.

We'll give it some time.

So here's what we have in mid journey.

I'll try to do my best to describe these images.

It's, they're pretty good, right?

Some of them threw a logo with an S on it,

even though we didn't want a logo,

but we have four different choices to choose from.

Let me know which one you guys like.

I'm going to run another prompt here,

but there's four different, very minimalistic.

It kind of put a, like a wood, you know,

like the steady sip mug on a wood table.

It threw in some other things,

even though I didn't necessarily want it,

we could refine this with a little bit more prompting,

but we're doing this live,

but some very sleek looking mugs, right?

So I'm going to go ahead,

I'm going to grab our next prompt and let's throw this in.

And let me know, this is our first option.

Do you like these black ones?

What do you think?

We're, I'm not going to go through all of these prompts.

I just want to show how we can work

with some of these images a little bit.

So let's, let's check on Dolly, Dolly finish.

Good job, Dolly.

All right.

So same thing in Dolly.

So we have four different images.

Dolly, I think, and let me know what you think.

I think Dolly did a little bit better of a job

in terms of following the prompts specifically,

because I asked for nothing else in, in this photo,

except for the mug.

And here we go, Dolly, four, four different,

four different images.

Same thing with mid-journey,

but I think these are pretty clean.

So different physical character,

characteristics of the mug as well.

So, you know, we did ask for a handle

and it looks like in Dolly,

only two of them got a handle on the actual mug.

It looks like one has a handle on the actual lid,

which is interesting.

And our original ones, our original ones in Dolly,

or sorry, in mid-journey, it looks like only,

oh, okay, three, three had a lid, I think.

It actually gave us some, some different sizes.

So I'm going to jump into the comments here.

Let me know which one y'all are liking so far.

And I'm going to also go ahead and put the next,

the next prompt.

I think we're just going to maybe do two.

We're not going to do all four.

So I'm going to get the next one going in Dolly.

But which one do you guys like?

I think these images are pretty good, pretty good.

We have a lot of wows so far in the comments.

A lot of wows.

Brian saying simplicity meets functionality.

Which, which one of these?

Tanya really likes the logo.

Thanks, Tanya.

Jay, Jay says love mid-journey,

even though it created an image of the box.

Yeah, mid-journey images, I think,

look a little more picturesque in this, in this example.

But I think Dolly actually did a little bit better of a job

at following the prompt.

We really wanted a product photo shoot with nothing else in it,

with studio lighting, simplicity.

And I actually think Dolly followed the prompt a little better,

even though I think if I were to just use one image,

I think maybe the mid-journey was a little bit better.

All right, yep.

A lot of people are saying Dolly's better.

Michael says Dolly looks better for e-commerce product.

Gabe says Dolly.

Yeah, a lot of people are saying Dolly.

Lower left.

So Gabriel's saying lower left, this one right here.

I like this one too.

This one, lower left hand corner, says stable sip.

It has the mug.

It looks very sturdy on the bottom.

I like this lower left one too.

Let's, let's look at one more.

All right, here we go.

So this is an interesting one.

So let's try this one more time.

I'll tell you what's happening.

Hey, this is Jordan, the host of Everyday AI.

I've spent more than a thousand hours inside ChatGPT,

and I'm sharing all of my secrets

in our free Prime Prompt Polish ChatGPT course

that's only available to loyal listeners like you.

Check out what Mike, a freelance marketer,

said about the PPP course.

I just got out of Jordan's webinar.

It was incredible, huge value.

It's live.

So you get your questions answered.

Pretty stoked on it.

It's an incredible resource.

Pretty much everything's free.

I would gladly pay for a lot of the stuff

that Jordan's putting out.

So if you're wondering whether you should

join the webinar, just make the time to do it.

It's totally worth it.

Everyone's prompting wrong.

And the PPP course fixes that.

If you want access, go to podppp.com.

Again, that's podppp.com.

Sign up for the free course

and start putting ChatGPT to work for you.

Let's see.

All right.

Oh, I entered the second prompt inside Dolly.

Okay.

And this prompt returned back an error.

I said straightforward studio lighting,

product photo shoot for StableSip,

a sleek silver mug designed for portability,

sleek silver exterior taper design,

slide lock lid, rubber grip, double walled insulation,

no logo, StableSip only.

And that return with an error from Dolly.

It says, I apologize for the inconvenience,

but I was unable to generate images

for the provided description due to our content policy.

That's one thing I've noticed with Dolly,

their content policy.

First of all, you can't read a lot about it,

but I think the word sleek for whatever reason

is triggering kind of this moderated message.

So I took the word sleek out

and we're going to see if we can go ahead

and generate a second one.

All right.

So let's go ahead while we wait for Dolly

to see if it's a little better.

Here is the second prompt inside mid journey.

So this one brought up images.

It's again, completely different prompts.

So we're getting completely different images.

These four are a little more silver.

Again, this is the prompt that I just read,

kind of the sleek design with kind of the shiny silver.

Again, these images, I think look great inside mid journey,

but they didn't really give us what we wanted, right?

But we have kind of this,

this to me looks more like a home bug.

So let me know, let me know.

Do we like the second output from mid journey?

This looks like to me,

this doesn't look like something that's on the go.

It does look kind of cool and simple and modern.

However, this doesn't look to me exactly like a mug

that you would want to take on the go, but that's just me.

All right.

So again, we got an error.

We got an error bummer from Dolly.

So let's try one more.

Let's go ahead and try one more for Dolly,

just to see if we can see multiple generations from Dolly.

All right.

And then after this, let me preview kind of while we wait.

Next, we are going to be using these product images.

So we're going to save the ones that we like.

And then we're going to do a little cleaning up in Canva.

We're going to go over Canva's AI magic suite,

just for a couple of things.

And then after that,

we're going to create hopefully some actual,

more like lifestyle images,

both in Dolly and also in mid-journey.

All right.

So what are we liking so far?

What are we liking?

Yeah, thumbs down.

I think Gabe is saying thumbs down from mid-journey.

So far, I would agree.

We're not getting great results so far.

Yes, Jay.

We did say no label, no logo label on the mug.

Sometimes when you have a lot of different instructions

or directions like I did,

early on, I think Dolly gets a lot of things right,

but then it'll get some things wrong.

I did have that on the back end of the prompt.

Normally, when you're quote-unquote prompt engineering,

what you would do is you would do some testing.

And you would move that like no logo up closer in the prompt

to see if it would remove it,

which a lot of times that works,

but it might forget something on the back end.

All right.

So let's go ahead.

So here is the second image generation.

This one actually worked inside Dolly.

So this prompt says,

product photo shoot for Stable Sip,

a textured gray mug with thermal insulation,

textured gray exterior, angular-based twist on lid,

finger grooves,

ooh, thermal insulated walls,

minimalistic studio lighting, Stable Sip only, no logo.

So you'll see, yes, it did include, unfortunately,

logos on all four, even though we asked for no logos.

But again, I'm not surprised

because we did have a lot of directions in there.

And we kind of quote-unquote buried

that no logo kind of at the end.

All right.

So let's talk about it.

Do we like these better?

Are these better?

Do we like those better?

Or do we like these black ones?

Personally, I think someone said

like lower left on the black ones.

I'm kind of feeling that one.

Let me know which one y'all are liking.

And then let's look at this one too.

And I'm going to go ahead.

I'm going to do one thing as we wait here as well.

So give me one second because shout out,

shout out, Douglas.

Douglas shot me over a tip.

He said, hey, I saw this, saw something you can do

with custom instructions.

And I was like, all right, let me put it to the test.

So I'm going to go ahead.

I'm going to rerun our first prompt.

And I'm going to explain to you what's going on

and why we're doing this.

All right.

So I'm going to go ahead and grab this first prompt

that we used that gave us these nice four black sleek images

that I think turned out really well in Dolly.

All right.

So I am opening another Dolly tab.

And here's why.

I have some custom instructions and I'll explain.

So these custom instructions, I wrote these

that essentially says label these images.

So again, shout out, Doug, and someone else on Twitter.

Figure this out a couple of weeks ago.

So let's run this and I'm going to show you why

we're going to be doing it this way.

And also let's get back to who loves what.

So Monica says loves the second iterations.

All right.

Tanya says the finger groves like that one.

Black.

Gabe says black.

Yeah.

Which one?

Which one were you guys liking more?

We'll give this Dolly a second.

Let's look at these again.

Are you all feeling the black version more?

Or are y'all feeling this gray version?

I do like the groves.

I think that looks nice.

However, at least for me personally, I'm feeling the black ones in Dolly.

What are you guys feeling?

What are you guys feeling?

Let me know.

Again, if you're joining us in the middle,

thank you for joining us.

We're doing this live.

We're building a brand.

That's why we have some 10 to 20 second pauses here

while we wait for generative AI to do its thing.

So yesterday, we literally created an idea live from the audience.

We built a brand called StableSip, went through the USP,

created marketing copy, did competitive research,

ran some competitive analysis, things like that.

So black.

Okay.

Most people are saying black.

Brian is saying the black version is better.

Two Bryons are saying the black version is better.

Monica says, is the goal what we like better?

Or which prompt was more accurate?

The goal, okay, two separate things.

The goal right now is which one do we like better?

And we can talk before and after about which one actually followed the prompts better.

But for now, let's decide which one we actually want to use.

Michael says black and Dolly, I agree.

Yeah, we can launch the StableSip 2.0 on Kickstarter.

Yeah, who wants to make this?

Who wants to make this?

Maybrit says Dolly's black output.

Yeah, I think I agree.

Brian says we need to have it in AI blues.

Yeah, I think mostly everyone is feeling the black version a little bit better.

So I'm trying to bring this up here, y'all, where we can do this with our custom instructions.

So let me explain why and hopefully this works.

So I'm zooming in.

I'm taking the exact same prompt that gave us these nice images,

the nice black ones right here.

I'm just rerunning this same prompt that gave us the four black images that we seem to all like.

And I'm running it inside a new chat in Dolly 3.0 with custom instructions enabled.

So if you are joining us live, custom instructions, I wouldn't even say love hate relationship.

Really, just like hate and don't need it.

But this is one interesting use case.

So custom instructions are essentially, if you have it enabled before any prompt is run inside chat GBT,

it actually looks at your custom instructions or refers to your custom instructions first.

So in these custom instructions, I essentially told chat GBT to assign a unique identifier to each image

so I can better identify it.

Because otherwise, you can't really keep working.

And again, someone on Twitter kind of figured this out.

In mid-journey, this is called seeds.

So essentially, each image in mid-journey, you can grab a seed.

So you can continue working with it.

And so you can essentially tell whether it's mid-journey or Dolly, hey, this is the image,

but let's make some modifications.

In my limited testing, seeding or working with a seed image works much better in mid-journey.

Dolly has some kind of mixed results.

But I just wanted to show you this.

So now we have four separate images.

Let me know.

Do any of you like this one better?

We have four different images.

One of them is more of an over-the-top view, which I don't know.

If you can like, I'm clicking on one, it put the logo of a horse's head on the flip top,

which is interesting.

I kind of like the lower right-hand corner of this one, y'all.

But actually, which one of these?

We're actually going to use one of these, and I'll tell you why.

We're going to use one of these so we can try to work with it

and try to continue to build off of it.

But just real quick, because we've got to keep going.

I don't want to accidentally turn this into a three-hour live.

That would be terrible.

So do we like this one better?

Do we like this second version better here in Dolly?

So pretty similar images to our first iteration,

but now we have the custom instructions.

I personally, I'm seeing if anyone's shouting out which one people like.

All right, Gabe says right lower.

That's the one I was feeling too, Gabe.

So let's go ahead.

So now I'm looking, and it says, here are the product photos,

and it's giving me reference numbers.

So I can now say, and again, sometimes this works, sometimes it doesn't.

I'm assuming since I'm doing this live, it's not going to work.

So I'm going to say, let's take image A4, right?

So that's our bottom right-hand corner.

So it's assigning a unique identifying code

that I can hopefully continue to build off of.

Otherwise, you just have to kind of keep going back and forth

and you can't really build off of an image.

So I'm going to say, let's take image A4,

which is the bottom right-hand corner, our new stable SIP,

and put it in a product shoot with two people working outdoors.

I'm going to say, it doesn't need to focus on the people,

but just show the people working and make the focus the image A4 bug.

All right.

So we'll see if this works.

I had varying results, and don't worry.

We're still going to be able to hopefully get some great product images.

We're going to take it into one or two other programs here.

All right.

So let's see if this works.

And let me know.

No, I don't think anyone wants it three hours.

Tanya says three hours.

Yay. I don't want to take up too much of your day.

We do this live if you are new to the show.

Normally, we have about 25, 30-ish minute episodes.

Obviously, these ones are going to be a little longer

since we're trying to build a brand live.

All right.

So we're going to try to think if we should wait here for a second.

Sometimes Dolly is super quick.

So again, we're in Dolly.

We're trying to use kind of this custom instructions benefit that it gave us,

these image identifiers, to see if we can continue building off of them.

But what I'm going to go ahead and do for now,

I'm going to download this one in the lower right-hand corner.

And I'm going to jump over in mid-journey.

And we're going to try to do something similar.

Ready?

So I'm going to do this.

I'm typing backslash describe in mid-journey.

And I'm uploading this image that we just did in Dolly.

I'll tell you why here in a second.

So what I'm trying to get.

So you can run this describe command inside of mid-journey.

For a couple of reasons.

Okay.

It's failed the process.

I'm going to try it one more time.

Mid-journey was being a little weird today.

So let me try it one more time.

All right.

So one thing that you can do,

and we've showed this in shows before,

and I'm going to go ahead and do it.

So you can upload an image into mid-journeys.

So people are always like,

okay, what if I already have an image and I want to keep working with it?

Well, you can upload it in mid-journey with the slash describe.

And then what happens is it gives you a prompt.

So then you can then use that prompt in mid-journey

to try to recreate anything.

So, you know, if you have a real life photo shoot

and it's like, ah, you didn't get things right

and you want to bring it into mid-journey,

you can try your hand at then using text prompts

to try to make it better.

But actually what I'm doing is I'm just clicking on this now.

And because I uploaded it,

now I'm right clicking and I'm saying copy image address.

Okay.

I'm going to explain what I'm doing here.

So now I'm typing in backslash imagine.

And now I copied that image address.

So I pasted it.

So essentially all it is is it's a long link.

Okay.

And then I'm going to go ahead here.

I tried this before, hopefully this works.

So here's what I did.

So now it just says I'm running a command in mid-journey.

So backslash imagine.

I pasted that image link.

And then I put a description.

So now what I'm doing is I'm trying to stack.

I'm trying to stack something here.

So I'm both uploading an image into mid-journey

and I'm using the source image as inspiration.

And then I'm saying a man in his mid-30s

working on a computer with a mug on his desk,

picture-esque view, inspiring.

And then I put Canon 5D Mark 4.

So hopefully it can be more of like an image.

So now what you'll see here as these render, right?

Are we getting it here, y'all?

So it didn't get it 100% right.

But now we have that base image.

And now are you seeing how we're pulling this together, right?

Hey, if you're a creative out there,

if you're an advertising, marketing, whatever,

this is like such a great hack.

We just used a really good image from Dolly.

We saved it.

We uploaded it to mid-journey using the slash describe command.

So we could grab the URL,

then we use that as like a base or a seed

to then use that mug that just, it was a mug only.

And now we are building scenes around it, right?

Are we getting?

Are we getting how cool this is?

All right.

So which one, which one do you guys like here?

So I'm sharing my screen.

We did get, it did work.

And I think it looks pretty good.

Again, it didn't get the mug exactly right,

but it got it pretty close.

Okay, it changed some different things

from picture to picture,

but we have now a grid for different pictures

that has our mug now with different scenes of,

you know, a man in his mid-30s working

and this kind of embodies what we were kind of working on

yesterday on, on how the stable sip should appeal

to kind of like a digital nomad and entrepreneur,

you know, someone that's always on the go.

And that's what we see here.

So which, which image do we like here?

Brian says this is pretty amazing.

Which, which, which one do we like?

Do we like?

And I know it's a little small on the screen.

I'm going to zoom in a little bit.

Do we like upper left, upper right, lower left, lower right?

And then we're going to go ahead and re,

we're going to upscale these images.

And then we're going to, we're going to go with this.

We're going to try, we're going to try to make this show

about the same time as yesterday,

but you got to let me know which one do you like the best?

This is an interactive one.

And again, if you're listening on the podcast,

please, please click the show notes, come back and watch this

because I think this is something that you really have to see

to understand, but I'm doing my best to describe it.

Kapesh says Dolly.

The second option is left.

Tanya says upper right.

Brian says upper left.

Maybritt says top, top right, upper.

Okay. So we have, we have multiple votes

for top left, top right.

Okay. So here's what we're going to do.

We're going to do a couple of things here, y'all.

So what I'm going to do in mid-journey,

there's, there's some more options that you don't have in Dolly.

So I'm actually going to run a variation of version two.

So version two is our upper right.

So I'm clicking, I'm clicking V2.

So that's going to run now for different variations of this.

And the reason why I'm doing that is because

if we have one that we like, then we can save them all

and then use them on the website that AI is going to create for us.

But if we have one that really looks good,

we can run some variations.

Hopefully the mug will look kind of the same.

And then we could even use multiple of those photos

if they are different enough.

All right. So let's give that a second.

Let's jump back in to Dolly here.

Let's see how it did.

All right. Here we go.

So as a reminder, here's what we did.

We were using similarly to kind of that process

that we just did in mid-journey.

We said, let's take image A4 and put it in a product shoot

with two people working outdoors.

Okay. So look at this.

So it kind of worked, but kind of didn't.

So this is where I said, you know, people on Twitter

are talking like, oh, it works, but it really doesn't.

Because we wanted this one in the upper right-hand corner.

So Stable Sip, it's a very sturdy mug.

This is kind of like really what we wanted.

But if we go down here, it followed everything we wanted,

except it didn't really bring that mug over.

So we have four different images that Dolly created,

but it did not really give us what we needed.

So I'm going to say, I'm just going to try.

So that is, I'm typing that as the wrong mug.

Everything else looks good.

I'm going to say, please regenerate using the,

what did we have?

This was image A1 using the mug from image A1 in the scene.

So again, varied results from Dolly, but no worries.

I think we have some good stuff here in mid-journey.

This is why you got to work with them both, right?

All right, here we go.

So we got some variations.

Same thing here.

It changed the mug kind of comically, right?

So some of these, now the mugs are like,

looks like 80 ounce coffee mugs.

We're not going to roll with those.

So I'm going to go back to this original one

in the upper right hand corner.

I'm going to upscale that so I can click you too.

So you can always upscale or run variations in mid-journey

or just do the regenerate.

So I'm going to do a couple of things at the same time

because we got to get this one going.

So I'm upscaling one of these images here,

which actually worked pretty well.

And actually now that I upscale it,

that mug also looks kind of big, right?

But that's okay.

I guess it's from a closer vantage point.

So we'll still roll with it.

So I'm going to save these images.

Here's our image.

I'm going to save it.

All right, we got our image saved.

And then we'll see.

I'm running some variations.

Thank you for sticking with us.

I know we're jumping around.

Let's see.

Hey, speaking of jumping around,

let's see how Dolly did.

So it looks like it kind of,

we re-ran this and asked them to specifically get the correct mug

because when it ran kind of using this seed technique,

it did not get it right even the second time around.

It really didn't.

However, this one in the upper left hand corner,

I think is kind of close.

So I'm just going to grab it just so we can still use it.

All right, so that was a lot in image generation.

We spent a little longer than I wanted to, but that's okay.

So let's do something real quick.

All right, let's jump in to Canva.

All right, so we're going to use,

we have our product images,

and here is our logo from yesterday as well.

All right, so we're going to go over some basics

because we got to clean this up.

So, hey, there is some human work involved here, y'all.

So I have the logo from yesterday that we kind of voted on.

There's a couple of things that didn't turn out

how we wanted them to, and a couple of changes we have to make.

So I'm going to click on this photo in Canva.

I'm going to click Edit Photo,

and we have these Magic Studios.

So Canva just did a pretty big AI release about two weeks ago.

We covered it in separate episodes.

So I'm going to click Background Remover

because there's a background,

and I want to get rid of this background.

And let's see.

It looks like, which is funny, it may or may not be working.

Okay, I think it might be working.

Okay, so I'm pretty sure we removed the background.

I'll have to see here in a second.

So now I'm going to click on this again.

I'm going to go to Edit Photo,

and I'm going to click Magic Eraser

because, essentially, it added the word graphic at the bottom.

So I'm going to now have the Magic Eraser tool

highlighting the word graphic, and it is going to delete that.

Obviously, that doesn't seem very impressive right now,

but this Magic Eraser tool in Canva is amazing.

This was very basic.

This is essentially no magic involved in this.

We're just deleting a word from a transparent image.

However, what it can do, Magic Eraser,

is probably one of my favorite parts of Canva's new Magic Suite.

It's fantastic because if there's other things in the background,

if I delete something, it uses AI to replicate

what pixels would be behind there, and it does a fantastic job.

All right, so now we have our logo there,

and let's go ahead.

I'm going to upload.

I'm going to bring in some of these images that we saved.

So here are some of our images,

and we're going to do a little more work here.

All right, so bear with me.

So we have one of our original product photos

that we really liked.

But like we talked about, we have a problem here

because this stable SIP is not the logo that we want.

So now we're going to be able to see that Magic Erase.

Sorry, I'm going to undo that.

I accidentally hit Background Remover,

which I did not want to do.

So I'm going to click Edit Photo.

I'm going to go to Magic Eraser.

So now I'm going to zoom in here.

If you just erase this, it wouldn't look good

because there's texture on this mug.

So hopefully, Canva is going to do a good job here

of filling in this texture,

which again, if you've ever used Photoshop

or something like this old-school Photoshop

before they introduced their AI features,

this would take a long time, and it wouldn't look good.

So I highlighted this logo, and let's see how.

That's amazing.

It took it away.

It got the sheer and the shine just right.

You literally cannot tell a logo was there previously.

All right, so I think it looks great.

So I'm going to click back.

So now what we're going to do, kind of old-school,

not old-school here, but little kind of old-school.

So I'm going to slap this logo on.

So again, this isn't going to look perfect,

but I'm going to slap this logo on.

There is a little sheer and shine,

so it's not going to look the best.

And maybe we'll make it white.

So little trick, FYI, I work in Canva a lot.

So I'm going to actually just crank the brightness

on this, I believe.

Bring the contrast down, highlights up.

We'll see if this works.

Whites up and blacks down.

Okay, there we go.

It worked.

So quick way.

Again, this doesn't look fantastic.

We're just trying to show how we can quickly replace the logo

with some simple graphic design.

So I'm going to go ahead.

I'm going to save this image.

I'm going to save this image.

And we're also now going to go into another tool.

All right.

All right, here we go.

So I'm going to duplicate this here real quick.

Now we're going to delete the background.

And let's see how this works.

And I'll tell you why.

All right, fantastic.

So now let me make sure I got to jump into our live stream here.

Okay, I needed to make sure we still had the right image up.

Okay, so now I'm going to export this.

So what we did just to recap, and I know this is a tough one

if you're listening on the podcast.

So we took our AI-generated image from Dolly.

We brought it into Canva's Magica Racer

because it put in a logo we didn't want it.

We used the Magica Race tool to get rid of it,

which it did a fantastic job.

Then we essentially slapped the AI-generated logo on the cup.

It doesn't look perfect, but I think it actually

looks pretty good.

I think when we put it in some other product images,

no one is going to be able to know the difference.

So now we're going to export this as a PNG,

and we're going to do it with a transparent background.

And I'll tell you why.

All right, we're moving along fast, y'all.

We're doing a lot.

Now we're going into Flare AI.

Okay, this is a great tool that helps you use AI

for some product photo shoots.

All right, so I already have an account.

It's free.

So all I'm going to do now is create a new project.

And this is pretty simple.

Okay, I'm going to go to assets and upload my product photo.

So I just, we just made this together.

Here is our stable-sip mug AI-generated,

along with the logo that we just did.

So it says place your product here.

I'm going to make it slightly bigger.

So we'll probably run a couple here.

And there's a lot of different templates that you can use.

So I see these templates on the left-hand side,

and they are like very sometimes overly picture-esque,

if I'm being honest.

Different settings for product or e-commerce shops,

but it's great.

So think of like a product mock-up,

but it has kind of these templates that you can just apply.

So I'm kind of scrolling through these different templates,

and I'm wondering which one is really going to appeal to our audience.

So a lot of them, you know, it's really for,

I think a lot of these look good for beauty products,

for outdoor products, you know, it has mountains.

Here we are in a vineyard, you know,

there's a table with a candle and a tablecloth.

So a lot of these look a little,

maybe not what we're looking for necessarily,

but I do still think there's some good ones.

So let's go ahead and use this one.

Let's see, sorry, I'm scrolling through here.

I want to find a good one here,

so we can hopefully actually use this.

All right, this might not be the best,

but I chose one,

and when you click on one of these templates in Flare,

it essentially gives you this pre-built prompt more or less,

and it says a stainless steel travel mug on a rocky cliff

in front of the ocean and blue skies

with an island in the background.

All right, so we're really pushing the digital nomad lifestyle here,

right, and showing, you know,

showing this in a kind of a travel scenario, right?

So this, again, this is Flare AI.

It has actually a great free plan,

and I think it's great to generate some pretty simple Ecom product images,

and bam, we got it.

So there is our, on the right hand side, we have our finished product.

It puts in a shadow and everything,

so I think it did a pretty good job.

So I'm going to click download image.

All right, and then let's go ahead and choose one more here.

So you can also, which I like,

you can just go up here,

click any part of this prompt, and do it yourself.

So it says on, so I'm going to say on, a office desk.

Right, and then said in front of, and the default is an island.

And so the default here is the ocean and blue sky.

So I'm going to say in front of a picture-esque view of the Chicago skyline.

So let's see how this does.

So in general, you can even use this to create your own prompts from scratch.

So I am essentially updating this.

I don't know how it's going to work, but we'll see.

So now the kind of the prompt that we're using inside Flair

to generate some of these EECOM product images is a stainless steel travel mug,

which is our photo that we uploaded.

So it's using our photo as the base on an office desk

in front of a picture-esque view of the Chicago skyline.

Not sure how it's going to work,

but we're going to download it and use it anyways.

So let's see, hey, let me know.

How are we doing?

Val just put in a comment, just the whole mind blown saying, wow.

What do you guys think?

Is anyone going to buy this now?

All right, so not bad.

Not terrible.

All right, so we got our image generated.

It looks Chicago-esque there.

I think that's actually, it's good enough to use, right?

We're just using this for the basics here.

All right, so we're moving along.

We're moving along, y'all.

I said we are going to try to cap this.

I know y'all are busy people,

whether you're listening on the podcast or listening live.

All right, so here, let me recap where we're at, y'all.

We have all of our content from yesterday.

Okay, so not only do we have all of our content from yesterday,

but now we have an actual product image

with a logo that we created.

We cleaned it up a little bit in Photoshop.

We use mid-journey to seed images.

We use Dolly to seed images.

So now we have, even though they're not 100% accurate,

now we have some images that we can use

for multiple parts on a website.

So now we're going to combine everything using Durable.

So I've actually done a review of Durable before.

It is an AI-powered website builder

and kind of their call to action here

is build a website in 30 seconds with AI.

And if I'm being honest, you can actually do that.

So I already have an account.

You can do this for free.

There's certain features you have to pay for,

but essentially you just click Create a new business.

And I'm typing in business name, which is Study SIP Industry.

Unfortunately, you have to pick one and there's no good ones.

So I'm just going to say Writing.

And then give me a second here.

I need to grab.

All right.

I'm going to say, please update this.

Just give me, sorry.

So I'm jumping back and chat, GPT,

because we need to grab some of this content

that we wrote previously.

So that's what I'm doing here now off-screen.

So at least let's see three.

All right, perfect.

So we're going to grab that.

So now we're back in Durable.

We do have to choose an industry, unfortunately,

and we click Create Business.

So again, is this going to be the best AI-generated website ever?

Absolutely not.

If you don't know web design, web development,

WordPress, Squarespace, whatever, I don't speak code.

It's actually pretty good.

It very quickly creates a website for us.

I don't know if we're going to have time

to do every single piece, but what we are going to do,

what we are going to do is we're going to do the basics here.

All right.

So now what we have is we have an entire website already built.

Unfortunately, so one thing I wish Durable did a little bit

differently is gave you actually more options

on the front end to enter information about your company.

So you didn't have to kind of do all of this.

You didn't have to edit and update.

So I wish you could put in a couple paragraphs,

put in a couple photos.

So you didn't have to change it all,

but we're going to do this quickly.

So we do have a working website now for StudySIP.

So we can go in and we can edit each section here.

So now I'm going to upload our logo.

So I'm going to upload our logo.

Let's get our logo.

Let's see.

I think I actually need to jump into Canva here and grab our logo.

All right.

So that's what I'm doing now, FYI.

So I am re-exporting our StableSIP logo.

Again, we generated this with AI.

I need to make this transparent.

Obviously, any time you do something live,

it always takes 10 times longer, right?

Normally doing these things in Canva takes like 10 seconds.

All right.

So now we're uploading our logo.

Bam, we got our logo.

All right.

So our logo is in place.

There is our StudySIP.

All right.

So now what we're going to do is we are jumping back

and we're going to grab our content.

So I think we said bold approach.

So we are saying grab our StudySIP now.

So don't worry.

I'm grabbing all of these things kind of separately.

So don't worry.

I'm going to be grabbing everything, our call to action,

our headline, all of these things.

So we're going to do get your StudySIP now.

Bam.

Our subhead that we had ChatGPT generate is the only travel mug

you will ever need.

Bam.

There we go.

We are going to then do our style.

And then we can upload ready.

So we are going to replace and upload our new image, right?

So let's go ahead and put our kind of hero image.

I think we'll use the office, the Chicago office one.

I think that was a good one.

So there we go.

All right.

So I'm not going to go through, right?

I'm not going to go through and update every single section.

I'm only going to do a couple here.

So we can just kind of see this above the fold, above the fold action.

So let's go ahead and grab the rest of our content.

And we'll kind of look at at least the top version of this.

So we don't need to do the entire website.

But I wanted y'all to see that, yes, we can build an entire website here.

With AI, yes, we do have to do a little bit of editing in these sections, right?

Which I wish there was one that we could do a little bit faster on the front end, right?

Because it's not super easy to go through.

Or it's a little time consuming to go through and put all these in.

But we're at least going to finish the top half of the website here, all right?

So we can at least look at our creation.

And we can feel proud of what we built here, right?

So all right, we got this one here.

I'm just uploading.

So I'm uploading all of these images that we had before.

I'm uploading them so at least we can get a high level view

of what we created with Study SIP.

All right, here we go.

We're going to do our last one.

We're going to do our last one.

I know this isn't the range of products that we maybe wanted, right?

But let's just go ahead now.

So again, I am in.

I am in Durable.

We have at least the top part of the website kind of done.

I don't want to waste everyone's time and go through and pick every single piece.

But that's it.

It is done.

So let's go.

I'm going to click Exit.

I'm going to go ahead and click Change This.

So we can publicize this, right?

So after the show, when I'm not wasting everyone's time going through piece by piece,

we will do this.

But there's our website, right?

So again, we didn't get the time to put in every single piece of copy,

but we did create that website in just a matter of seconds.

So give it a second.

It's loading up here.

So as let's quickly, quickly recap what we did here, y'all.

And thank you if you're still sticking around.

I know this was a longer episode.

I appreciate y'all sticking around.

Ahmed said, Jordan, you are the best.

Jay, Jay, thank you.

Jay said, just built a site on Durable this weekend.

It did a great job.

Yeah.

So Durable to answer the question.

Val, he said, what is the name of the website builder?

Durable.

There's a couple of these different kind of AI generated website builders.

And here is kind of our preview of the website here.

So Brian says awesome stuff.

Thank y'all.

So very quick, let's recap.

And let me answer the question here.

Can AI, can chat GBT?

There is our website.

Again, I'm not going to enter this website into any design awards,

but we got something pretty good, y'all, pretty good.

So let's talk about this.

What do we do?

What didn't we do?

And this is why I wanted to do an episode like this.

Today's was a little longer, but people always say AI is not for me.

AI is not for my company.

My background, I've been getting paid to write for people for 20 years.

I've been designing websites for probably longer.

I've been helping companies with brand strategy.

Business operation, business development, you know, in my career,

I've worked in so many aspects of marketing, advertising, communication,

business development.

I was the executive director of a company before.

If you say AI is not for you, if you actually watch the last two episodes

where we built a brand live using generative AI, if you are still sitting there

and saying, I don't know how AI can help my company, help my brand,

you just saw it.

You just saw it, right?

What we did, again, we would, if we wanted to actually launch this as a real company,

we would have spent a little more time.

We would have done some fine tuning, a little bit more on the prompting,

a little bit more on the copy, probably a little more on the strategy side.

However, we at least went through the entire process.

Ideation, finding, running a competitive SWOT analysis,

you know, doing industry comparisons, doing target demographic research, crafting copy

that fits all of those different things.

And then taking all of those insights and making creative as we talked about today.

So using mid-journey, using Dolly for logos, imagery, product shots, lifestyle images,

and then using even Flare to kind of do some AI-powered kind of Ecom style product images,

which I think turned out very well.

And then using Durable as an AI website builder, which gave us a website

that you see now here on the screen in a matter of a minute or two.

So if you don't know, now you know.

Yes, AI can build a brand from scratch.

It can do it live and it can build your company and your career as well.

All right.

I hope you all enjoyed this.

And as a reminder, please go to your everydayai.com.

I promise y'all we don't do these hour episodes very often.

I think this is probably only our second one.

So I hope you join us back tomorrow and every day with Everyday AI.

Thanks y'all.

Machine-generated transcript that may contain inaccuracies.