00:03
Hi everyone, welcome to The Creative Mindset, a podcast about what the future holds at the
intersection of creativity and technology. I am Reina Moto, the founding partner of I&CO,
a global innovation firm based in New York and Tokyo.
AIを活用したクリエイティブな活動
Generative AI has been the topic of the year, not just in the creative industry, but various
industries and how it might impact the way we work and us as human beings. Since the
explosions of tools like ChatGPT, Meet Journey, and other generative AI tools, there have
been many attempts at leveraging AI in the creation of content. However, at the enterprise
or brand level, there haven't been that many viable or successful case studies of
AI for and by creatives. Today's guest is Pum Lefebvre, the co-founder and the Chief
Creative Officer of DesignARMY in Washington, D.C., where she oversees all creative, creative
coming through the agency's doors. In part one, we talked about how she used AI in actual client
work, and she shared with us specifics of how she built her campaign using Meet Journey,
how long it took, and the advantage and disadvantage of using AI for a creative process.
If you haven't listened, please have a listen. In part two, we go into how we, as creatives and
designers, should maximize our creativity in the age of AI. So, let's get started.
If you were to produce this in a traditional way, the way you had done it, and you had done
many, many photos, I mean, you've done beautiful work for the same client in the past, and all of
them, it's design armies work, beautiful photo shoots, and beautiful video shoots, great-looking
models, and great-looking fashion, all those kind of things, right? How long would that have taken?
Like a usual...
Like a campaign, like a year ago, like how much time did you have to spend?
At least this one, four months.
Four months. And on average, a client, and you don't have to be specific, but a client like this,
you know, this size, a regional client, how much would they have paid for something,
including production?
No, we can't talk about that.
I knew you were going to say that. I was trying, I was pushing my luck. Yeah. But, you know,
so let me put it this way. Having done this sort of thing myself, right? Like,
I would say, you know, on the lowest end, I would say 150, you know, 100 to 200k. Like,
for, you know, to me, 400k would be a better budget to work with. Like, in the past couple
years, that's the kind of shoots that I've been either directly or indirectly involved.
But even with, I say, it was video involved, you know, it becomes very expensive.
I mean, imagine this campaign.
Yeah.
Shoot for real, that's half a million for at very bottom, for the people, for the class, for,
right? But this is all digital. It's all not exist. The model doesn't exist. You know,
the land doesn't exist. So I, it's, you know, it's good and bad. Good because it's fast,
but then, you know, we have to protect our creativity and we need to put the value and
dollar sign into it.
I think everyone tried to figure out how to charge the client using AI. I mean, at the end of the day,
it's still the idea, right? Adventures in AI is an idea. It is a campaign. It is to help someone
selling the product, help consumers visualize the glasses that is probably not that inspiring
on the white background, but now you put on someone's face. Like, oh, you know, with that
skin tone, with red colors, bring out your blue eyes or whatever. You know, it's very, very, very
easy to do. I think, you know, it's a very, very good idea. You know, it's a very, very good idea.
I think, you know, it's a great idea. I think, you know, it's a very, very good idea. I think, you know, it's a very, very good idea.
whatever um you know that's still creativity yeah so i mean we just need to be really good
at like you know we need to charge the client and it's still time it doesn't generate itself
so by the time you presented you know within the first week and a half or two weeks or so
was the first presentation relatively close to the final output that you had or you still had
to sort of iterate off like what was the first presentation like it was done that's the different
you know like in the for us as a creative agency you usually skip concept pull mood board and then
you have to go to hire director photographer scout wardrobe makeup cast all that right
it skipped that process completely because as you concept
it's already creating the final product as you go so you you skip the photo shoot aspect of it
really so you concept it out this is what it looks like we might adjust typography but it's pretty
much done ready to use it's going to be a hard time for us to win with speed you know like the
speed and the speed and time right i mean the quality has to be there now we are going to
compete with each other
AIによるデザインの民主化
on the speed and time how fast can you get your this project done beautifully right and then you
know how long and cost i mean imagine i mean we do photo shoot a lot this is expensive campaign
it's you're gonna have to fly into arizona or desert you have to get 50 different models
for the right we have to
see you know the fashion part and there's a limitation like how can i i'm not gonna have
funding to create you know a calm the gazon ish with jellyfish dress you know in real life so
what it does to me uh this process is make me realize that ai can democratize design so if you
don't have money or time you know or money or resource i'm not going to be able to do that
this is can be a great use for someone who has idea imagination but not necessarily have
you know resources or budget to fly to arizona with 50 people so it sounds like just to break
down what you just talked about there's the idea that you you a human being prune you came up with
but based on what you talked about because it was you i think this happened like if it was
a non-creative person who wasn't able to do it it was a non-creative person who wasn't able to do it
to think through this in a very quick way yeah you know like you've been doing this 25 years or
whatever but it's the experience that you have as a human being as a creative talented smart
expansion division like if it was a 25 year old creative person i don't know if it would have
happened this quickly you know so that was that and and also you being able to sort of connect
okay what is the client trying to do and what do they need to say what's one of the product
uh benefits and being able to connect to that so there's the idea right
the second
part you also talked about was the quality right and that's also based on your experience your eye
your talent and you know what's good versus what's bad like you're able to judge because
yeah you need a little certain taste to be able to judge right taste yeah yeah uh but the time and
the cost part i mean time is money as well but i guess i assume that you know you get paid by
this client for your time or for your time and you know you're able to judge because you're able to
educate yourself for your team's time right but when it comes to production did they pay you for
production or were they just paying your basic fee well there's an animation aspect to it
you know so so there is okay yeah yeah production is a part of it too so it was just like a you
know a experiment project you know i had to convince my client to let us do this so but it is
新たな視覚言語の可能性
a you know paid project obviously and um production time is just like a kind of typical
Yes, because generate it alone, photo alone, that's not enough.
We still have to put glasses in the photo.
We still have to animate it.
We have to write caption.
We have to make sure that the episode and the story is right.
I mean, we add music.
We had a music designer creating a score,
original score for the sounds of Eileen, you know?
Do you think that the same client or another client would come to you and say,
hey, we only have four weeks and we don't have a lot of money,
but we know you can do great work in AI.
Do it.
Like, what would you say to them?
I can do four weeks, but then you will have to have money.
Yeah, yeah.
I can do it quite well.
Yeah.
You know, everyone have access to Midjourney,
but then...
And we have done it so beautifully.
Yeah.
It's funny because, you know,
we got a lot of press for this from Ad Age or Ad Week.
Literally one week after we launched the campaign and it was on press,
I got three calls from new clients.
New clients.
Yeah.
I think clients, they're curious and they're very excited about the process.
And it's just, you know, fascinating.
Not so much of the speed.
I think the result was...
More of the, you know, the result of the campaign.
Yeah.
It's almost like a new visual language in a way.
It's like the AI look, right?
Yeah.
It's just a new look.
It's a new style when you think about it.
So it's not so much about like how fast or how inexpensive.
I mean, that's probably important, but just the way it looks is interesting.
It's new.
クライアントの反応と今後の展望
Yeah.
You know, we are creating brand new feelings.
But were they interested in...
That kind of look or were they interested in, oh, wow, I can get this kind of look this quickly?
No, they're just intrigued by, you know, I have never seen like AI campaign done this way.
Right, right.
And then it was kind of interesting to, you know, it's like a new thing to play with.
Yeah.
Everyone just pretty much curious.
So what would you say if the same client came back to you, you know, three months from now saying,
hey, you know, I need something, what would you say?
I mean, we'll, you know, continue because like fashion, you have to do spring season and fall season.
Yeah, we are continuing.
I mean, this is actually part of the pitch is like we're going to do the next adventure in AI.
It's not just going to stop because, you know, it's have to keep going in the fall.
In the summer, the concept right now, it's going to go to different land.
Yeah, yeah.
You know, it's more like beautiful with flowers.
So we are going to keep going to the fall.
As experimentation and see, you know, what kind of result we get.
And it's going to change, you know, it's going to change.
And that's why it's already changed.
And I already not like Mid-Journey 5.
So like for me as Mid-Journey, I mean, when I first started working on this is version 4.
In March, version 5 came out and I don't like it.
Why is that?
Because it was too real.
Yeah.
It was so real.
Real, realistic, right?
So you see the picture of the Pope wearing Montclair jacket went viral because it's so real.
And then people can't, some people didn't think it was for real.
It's like a real photo.
Yeah, yeah.
So I, given the same prompt from version 4, version 5, and the result is different.
No.
And I'm not liking now as much because there is imperfection of,
for version 4, whether people have six fingers or, you know, this, the other thing they can't render very well, it's finger, you always see six or seven fingers always with, with a human rendering, but now it's getting too good in a way that it's kind of losing it charms a little bit.
AIと人間の役割
So like, I'm like not even upgrading my, to version 5 right now.
I just have to like add that I want to render in version 5.
I'm keeping version 4 because I don't want to lose that 1950 or like, you know, much more like artistic.
Oh, I see, I see.
Also, because now it's just become real.
It's like a little bit like when you watch Cinderella, Disney animation, you know, back then versus three-dimensional Cinderella now, they look different.
Right.
And I think there's something about like the drawing and imperfection that make the piece of art.
Art look like art.
Now it's become like so perfect, so perfect, so real that you can't tell a difference between machine and human anymore.
Yeah.
Yeah.
I mean, good or dangerous is depend on how, how you want mid journey to, to generate it.
Yeah.
But maybe it's good to like, you know, looking at the Instagram and see how the technology evolve visually.
Yeah, yeah, yeah, yeah.
But you might not be doing as many actual video.
I'm photoshoots.
As you've done in the past moving forward.
No, not true.
I think it's dependent.
Yeah.
It depends.
I mean, I just, you know, came back from a week long photoshoot.
If you want something real, raw, authentic, human, you have to work with human.
Okay.
It's going to be tough creating something that look real, but it's fake because there's something about the AI in it.
You can tell that it's.
you know an ai iterated image there's something that is still quite photoshoppy and photo
manipulation and you don't get the sense of like the wind the sunlight that hit the skin tone it's
just a feeling right and i think we all human can tell the difference yeah so so on that point
this is a little bit more philosophical than the conversation that we've been having but
what part of creative process or what part of creativity do you think ai will can replace
and what part you know should humans do human should you know we we we are the author
of a story so human will continue to dream and ideate you know idea and um
ai for me is a little bit like machine washer and dryer right it's like you know when you do
laundry in the in the past you would have to do it by hand you know it's just like throw the laundry
in yeah and it just automate and and and and get it done for you we human needs to
know how to make a call of like which kind of you know let's say mid-journey
you got four different a b c d choice which one is the one that feel right and it's to me it's
come back to the feeling does it feeling right you know does it look fake does it you know you
have to make a decision you know on on like what image you want to put forward um to to your
audience and what kind of image that's gonna work with the audience and the brands um so
i mean you know it's funny because i was really excited when i create this campaign you know with
mid-journey and i show to my daughter who is 17 years old gamer right it's 18 she's 18 she's a
gamer she's always you know like this kind of look and feel and i show it to her i was like
look sophie isn't it amazing i'm so excited about this campaign and of course she say
okay i think it's cool but it's not as cool as the campaign that you have done previously
for georgetown optician or hong kong ballet why why was that why was that she say it's lack
クリエイティブの価値
emotion and it's done by machine and i she's like you know i love the campaign that you
use real actor and you know go and shooting with the team um so much more you know
something good or not do you do put the value into human hand you know the time that it's
created the people that the community of human that coming together to create a campaign
you know versus like this campaign that done by pum and then you know six different designer
in the team and we'll be still working as a team but it's just a different role so like we as a
human has to make a decision what is good same thing like let's say louis vuitton bag right
louis vuitton bag costs certain price nine thousand dollar because everyone every single
piece of the bag is cutting by hand and there's a human behind that that is hand stitching
each one hand painted the side of the bag and each one is collectible
and it has a value when you buy and sometimes chanel bag increase in value you know as it go
same thing as watch right it's done by human hand so there's a value to that the bag can be done by
machine make in china and produce really fast and you know look like kind of like real louis vuitton
but it's not quite so there's a value into the human time and creativity
art of the craft um that should be considered of what food is in the future i mean how how do you
クリエイティブの将来
see i mean i think you know what keep me up at night is relevancy when you when you are having
you you when you have a your own agency for 20 years you always want to make sure that you stay
relevant and you not always just celebrate your past success right and i think that's
what i was doing now with this ai thing because i didn't even want to do it at first but then i don't
any choice but pushing you know that's keep me up at night and you know i do worry about the client
is not willing to pay what we worth that's
should that we should all be worried about because now you know the creativity ability
to tell the story ability to creating beautiful design if at the end of the day is about how many
viewer we can get on tick tock we are not judging on the craft anymore we are judging on the
popularity so you know so that's something that i'm fighting for our industry to you know we need
to create value for ourselves
and we need to not give away because of its past it doesn't mean it's cheap because our idea is not
cheap right that was part two of my conversation with pum lefebvre the co-founder and the chief
creative officer of design army in washington dc my three key takeaways from my conversation
with pum war number one it's the point of view that counts number two the human touch is what
makes the difference and number three
the most important thing is that the client is not obsessed about staying relevant the part of the conversation that caught my attention was when we were talking about how we should charge the client as you heard especially from the first episode the time it took for her to create this campaign drastically reduced from the normal process that she would go through with her team versus in this case she was the one who was doing the work the majority of the work the heavy lifting was done by herself and the
Hollywood
of all the
people
for
more
people
and
more
in
in
things that may have taken years for them to gain the skill to do but the difference here she said
that the work that she created using AI was better than the work that the younger generation did
because she knew how to give the prompts she knew how to revise the prompts to get what she was
looking for much more effectively than say younger designers who may not have the vocabulary who may
not have the experience to direct and give feedback in another conversation that I had a while ago
with PJ Pereira who had written a novel about AI and martial arts recently he talked about AI today
is like a horse instead of a bicycle and what he meant by that was that controlling AI could be as
tricky as working with AI
as riding a horse you need that experience you need the skill to know how to ride it so that
difference between knowing how to provide directions and feedback coming from the experience
and on top of that I would say in her case and this is what you pay her for is the taste that
she has she has a unique taste coming from Asia studying in the US and building a career
and she has a unique taste coming from Asia studying in the US and building a career and building a career
and working around the world with clients around the world that has helped her cultivate that kind
of taste that she has and that's what you pay for so even though the production aspect of it
the making aspect of it in a normal circumstance a campaign like that could have easily cost
five hundred thousand dollars just to produce and could have taken several months to concept
prepare and produce she did that almost by herself you know with some help from her own experience and
from a few designers that she has so in this case she was able to condense the process from say four
month to four weeks to go from start to finish however it is what you pay for is her for her
AIの限界と人間のタッチ
expense and for her taste and for her eye so not everybody would be able to produce the level of
work that she was able to produce so that's number one it's the point of view that counts
key takeaway number two the human touch is what makes a difference what was really insightful in
my conversation with Pom was that she realized relatively quickly the strength as well as the
weakness of AI and the fact that AI wasn't able to produce or generate the very product that it
was trying to visualize which is these pairs of glasses that this brand Georgetown Optician
sells I found that insightful and ironic at the same time because if we are trying to use AI to
speed up the process and the very thing that you want to depict is say in this case the product the
glasses but the fact that I couldn't is very telling but again this comes from her experience
that she realized that you know what this needs to be done by humans and she quickly adjusted the
approach to produce the glasses by photographing and then photoshopping that in to the images
that she generated so the models the fashion that these models were the setting in this case because
it was set on an image on an imaginary planet called AI that setting the scene those things
were producing AI but the most important aspect of the campaign the glasses
were made by or photographed by humans and then composited in so that kind of human touch is what
makes a difference I would add on another note really vital human touch is the story so the
関連性を気にしないこと
strength of this work that she did wasn't just the fact that it was the speed at which she was
able to produce this with a fairly limited resource but also creating the narrative crafting the
narrative so that
it wasn't hey you know what look at my AI campaign but it was hey look at this story about humans
traveling uh to outer space and on this planet you need these glasses so there was that kind of
background story fiction but story that was behind it so that it made sense and then it gave a reason
for this AI work to exist and that's the human touch that made the difference in this case so
the human touch is what makes a difference and finally number three and this is perhaps the
most important point obsessed about staying relevant as I said earlier AI tool is a very
Democratic tool in that anybody can use it you know many tools are free and it removes the barrier
of execution so in the past you may have years of expense to be able to produce the kind of thing
and this one that with the use of mid Journey that removed that barrier of experience and the skill
so think of these technology tools AI and beyond AI these are the tools that can help us stay
仕事と関連性
relevant regardless of age experience industry profession what have you staying relevant as a
professional and a human being is one of the most
things that we need to obsess about. Oftentimes, technology does make workers irrelevant because
it takes away from jobs, but AI can give us an opportunity to transform what we do and how we
work, and it can help us stay relevant. So to summarize my three key takeaways from my conversation
with Pumar, key takeaway number one, it's the point of view that counts. Number two, the human
touch is what makes the difference. And number three, obsess about staying relevant. If you are
listening to this on Spotify, there's a Q&A field, so please do send us your questions and comments.
And if you like our podcast, please leave us a five-star rating. We'd be so grateful.
I'm Reina Moto, and this is The Cray Mindset. See you next time.
