Inside Goodby Silverstein’s AI R&D Lab

Technically Creative by KoobrikLabs Podcast Transcript

Title: Inside Goodby Silverstein’s AI R&D Lab

Orlando Wood (00:00)
Every few years, advertising re-invents itself. And right now, that
re-invention is being led by the technologist sitting inside the
creative week on Technically Creative, I’m joined by Martin Ludvigsen,
Director of AI and Creative Technology at Goodby Silverstein and
Partners, one of the best advertising agencies in the world. At
Goodby, Martin runs a department that lives inside the creative floor
plan, not off in IT.
It reports directly to the chief creative officer and its job is to
prototype the impossible. Think of it as a creative R and D program
that sits between idea and execution. Supporting creatives as they
develop concepts, but then also liaising with outside vendors to bring
those ideas to life. It’s the connective tissue between imagination
and production. And it kind of, it’s one part creative development,
one part pitch material creation and ideation in that respect, and
then one part production department.
We talk about how it operates within the traditional agency structure,
how GSP’s Ask Dali project let museum visitors literally chat with
Salvador how the agency’s recent BMW Real for Real campaign turned the
idea of AI Slop into a sharp cultural statement about trust. Martin
explains why the most successful uses of AI in advertising put AI at
the center of the idea rather than just another cost-saving tool.
Because when AI becomes the subject of the idea, not just the software
behind it, it can elevate creativity, not just kind of automate it.
This is a conversation about where imagination meets our engineering
and what happens when a 40 year old agency decides to act like a
startup again. I think you’ll really like it. It’s very informative.
Here’s my chat with Martin Ludvigsen.
Orlando Wood (01:44)
Well, Martin Ludvigsen, know I’ve butchered that and not sent it in
the proper Danish way, but thank you for for being here. I’m super
happy to have you on the podcast.
Martin Pagh Ludvigsen (01:49)
Ha ha ha ha.
I’m so excited to be here and I’ve lived in the United States for
about 15 years so I have 15 years of experience with people butchering
my last name so I’m not at all offended by it. ⁓
Orlando Wood (02:07)
Well, I appreciate
your grace on that. But also, we have, I know I’m working towards a
deadline here because, Denmark is playing exactly at the end of this
hour. You guys are playing your first World Cup match against
Scotland. So I’ll be very respectful of your time as well.

your grace on that. But also, we have, I know I’m working towards a
deadline here because, Denmark is playing exactly at the end of this
hour. You guys are playing your first World Cup match against
Scotland. So I’ll be very respectful of your time as well.
Martin Pagh Ludvigsen (02:22)
Thank you so much. am wearing my team jersey and I’m ready to shift to
that. as I told you in the green room, I’m nervous, not for this
interview, but I’m nervous because this is the start of our World Cup
qualification. So big day to be Danish.
Orlando Wood (02:30)
Yeah.
Yeah, absolutely. Well, I promise we’ll get you out on time. You are
the director of AI and creative technology at Goodby Silverstein and
Partners, which is a ⁓ huge role, you know, especially kind of now.
And what’s so interesting to me is that Goodby Silverstein and
Partners is, certainly one of the agencies that has their face on the
Mount Rushmore of advertising, but particularly as well is based in
San Francisco. So.
it seems like it’s the perfect place to be on the front lines of AI
and technology. you have a background, were BBH Mullen Lowe, now
you’re at Goodby. mean, what, do you think Goodby and being in San
Francisco is particularly special in terms of your role?
Martin Pagh Ludvigsen (03:18)
Absolutely, absolutely. ⁓ There’s no doubt about that. It goes up and
down. I will say the agency has been around for a little over 40
years. So obviously, the tech boom of San Francisco has probably been
around for about that same time. And over the years, big tech, which
is probably a new term, didn’t exist back then. But technology
companies have played a big role in the work that we have done both
our clients along the way, there’s been many of those, but also as a
contributor to the tools that we can use to display our creativity, to
create some of the amazing ads that we’ve made over four decades of
flagship creative advertising.
Orlando Wood (04:00)
What does your team look like and how does it integrate with the
agency?
Martin Pagh Ludvigsen (04:03)
So the most important point to make is that my team sits under the
creative department at our agency. are creatives. I answer to our CCO,
Margaret Johnson. So we’re very much part of the creative department.
And that might be a little different from where…
AI sits at a lot of other agencies where it sits elsewhere. And you
know, mentioned my title as Director of Creative Technology and AI and
that really is…

first and foremost, how we think about AI, about generative AI. How
can it be used to build the creative that we make for our clients? How
can it be used to tell stories in new and unexpected way to work for
the brands that we work for? So I have a team of creative
technologists ⁓ and a producer who’s part of our team. And we really
work across almost the entire portfolio of clients at our agency.
Orlando Wood (04:37)
Mm-hmm.
Mm-hmm.
Martin Pagh Ludvigsen (05:00)
agency on figuring out new and unexpected ways to bring ideas to life.
And that goes both for the existing client portfolio that we have, but
also very much for the new business that we participate in. There’s a
lot of new business going on all the time and a lot of ways where we
as an agency demonstrate how, should you pick Goodby Silverstein and
Partners as your new agency has to do with our capabilities to bring
their brand story to life in new and unexpected ways.
Orlando Wood (05:06)
Mm-hmm.
Martin Pagh Ludvigsen (05:27)
And very often creative technology and AI is part of that.
Orlando Wood (05:32)
Yeah, yeah, I mean, we’ve been kind of saying to people that your IT
department now is a creative resource in a way it wasn’t before these
models came along. But I think that that also misses the point. I
think that, as you say, sitting in the creative department and being
trusted for your creative input on how to use these technologies is
incredibly important and making sure that
The why to use the technologies is respected rather than just the how.
Martin Pagh Ludvigsen (06:02)
absolutely. I think one thing that’s always been the case at Goodby’s
Silvestian and parties is that we have the best creatives in the
world. And they come up with the most amazing and crazy and wonderful
ideas, unlike anything you could imagine.
And those ideas across all sorts of media. Obviously, we are, as an
agency, historically, we’ve done a lot of great TV work. We always
have a lot of work in the Superbowl. So that is very much a part of
what we do. But really, the big ideas can come to life in a lot of
other channels. And that’s where creative technology comes in. So a
lot of conversations really start with an individual creative, a
creative director, whatever, coming to my department.

And those ideas across all sorts of media. Obviously, we are, as an
agency, historically, we’ve done a lot of great TV work. We always
have a lot of work in the Superbowl. So that is very much a part of
what we do. But really, the big ideas can come to life in a lot of
other channels. And that’s where creative technology comes in. So a
lot of conversations really start with an individual creative, a
creative director, whatever, coming to my department.
we’re called the labs at at good be sylvester and partners and saying
i have
an idea or I have the beginning of an idea or I have a provocation, I
have a thought. Do you think this is something that we can do, we can
bring to life with the use of technology? And the answer is almost
always yes. At least we will try our best to make it happen. That
doesn’t necessarily mean that the idea will make it all the way
through and be something that the public can see one day. But at the
very least, we want to give them as much support as possible to
realize
the vision that they have for their idea. That also means that we are,
as a department, very technology agnostic because for us it’s not a
matter of, okay, let’s figure out how we can bring this idea to life
in this platform or this technology that we know really, really well.
It’s about, okay, what is the best way to execute this idea and let’s
figure out what technology is needed to support that. So over the
years, I really have become a jack of all trades.
Orlando Wood (07:23)
Mm-hmm.
Martin Pagh Ludvigsen (07:42)
when it comes to creative technology and sort of know enough about any
technology that can be used creatively to at the very least know how
to ask the right questions, talk to the right people, find the right
resources, build a prototype or whatever it takes to demonstrate how
the given idea will come to life through a creative use of technology.
Orlando Wood (07:42)
Mm-mm. Mm.
one of the last times I was in an agency, that was in the time where
it was really fashionable to hire a head of integrated production, or
there was this idea that there would be an integrated producer who
would be just as good at,
flash banner ads and things like that as they would be at television
production. And then of course, over time, everybody realized they
were two separate specialisms and it was more likely that you were
gonna have some people who were gonna be naturally better at both. I
think that that seems to be the same when it comes to a department
that would understand creative technology, that it’s a niche and a
special skill that needs to be respected.
Martin Pagh Ludvigsen (08:41)
I think so, and I’m very happy to work at an agency that shares that
vision that creates technology is such an important part of what we
do. Even though the majority of our work is in channels that we
understand really well, and we have the producers who are specialized
in that, and they don’t need someone like me to help them out because
this is a technology and a platform that we know well. So that’s not
what we are a part of. But as soon as we start moving in,

I think so, and I’m very happy to work at an agency that shares that
vision that creates technology is such an important part of what we
do. Even though the majority of our work is in channels that we
understand really well, and we have the producers who are specialized
in that, and they don’t need someone like me to help them out because
this is a technology and a platform that we know well. So that’s not
what we are a part of. But as soon as we start moving in,
into
technologies that our individual producers are not experts on. Or when
we cross over between different platforms, that’s when we need to
figure out how to be the liaison between the creatives, the producers
in figuring out how to actually bring this to life. Even though I will
say in full transparency, most of the time our lab department is not
the department that actually builds and produces the final idea. That
is most of the time something that we hire third party studios,
production companies to help with. But we very much need to be there
to first and foremost say, yes, this can be done. So you’ve got to
pass that synergy check. This is actually an idea that is possible to
make with existing technology. And then hopefully we also make an
experiment, a prototype that demonstrates this is this it could look
something like this as a finished product. And then after, we were
very important. Part of what we do is also selling that vision, that
idea through to the client, including
Orlando Wood (10:00)
Yeah.
Martin Pagh Ludvigsen (10:08)
giving them the confidence that this is something that can be built
and can be scaled to whatever the ambitions for the scale are, and
then finally identifying the right part to actually go out and build
this for us in a secure, scalable manner that lives up to all the
expectations that we have for the craft of any idea that comes out of
our agency.
Orlando Wood (10:29)
So in practice, you’re quite similar to the graphic design studio
mixed with the production department. You’re one part creative
development internally, one part liaising with outside vendors and
figuring out how something’s actually going to be executed. That’s a
really interesting place to be able to sit.
Martin Pagh Ludvigsen (10:48)
I’m very happy with where I sit. I’m very happy to be part of so many
conversations. ⁓ I actually…
In the typical separation of roles at the agency, the people who are
there from the very beginning to the end are the account people.
They’re the ones who have the initial contact with the client and all
the way through development and delivery and production evaluation of
the idea. Obviously, they are the ones who are there from start to
finish. But we’re like right after that. We’re probably involved in
the projects more than our creatives are because the creatives will
typically come in

In the typical separation of roles at the agency, the people who are
there from the very beginning to the end are the account people.
They’re the ones who have the initial contact with the client and all
the way through development and delivery and production evaluation of
the idea. Obviously, they are the ones who are there from start to
finish. But we’re like right after that. We’re probably involved in
the projects more than our creatives are because the creatives will
typically come in
after a brief stage which our strategists take care of. But we will
very often be involved before that because what we contribute, what we
know will also contribute to that brief stage. It’s already there
we’re exploring what are the technologies that could be interesting
for a brief like this for solving the business problem that we have
for our clients.
Orlando Wood (11:41)
Mm-hmm.
how much does the brief come directly to you and how much does the
brief go to a set of creatives who think they have an idea that might
require some tech and then they sort of sidle up to you and go, oh,
they look like we think we kind of want to present this idea, but
we’re not quite sure it’s even possible. How does that creative
workflow work?
Martin Pagh Ludvigsen (12:06)
⁓
As much as I would like to be the one to come up with the brilliant
ideas to begin with, most of the time that is not how it works out.
And actually over the years, that has probably shifted a little bit
where earlier in my career, I’ve had other director of creative
technology positions. I was at Mullen Lowe before this and I would
more frequently contribute in the idea generation phase at that point.
But I think I learned that this was not necessary.
Orlando Wood (12:15)
Mm-hmm.
Mm-hmm.
Mm-hmm. Mm-hmm.
Right.
Martin Pagh Ludvigsen (12:33)
the best use of my time of of my team’s time because
You know, I’m not saying I’m bad at coming up with ideas, but I’m also
saying that the creatives that we actually have at our agency are
better than I am at coming up with it. That is why they are the
creatives. They are the creative geniuses who come up with the really
cool stuff that I did not think of. So I’m there to support them right
after the light sparks in their brain about, wouldn’t it be kind of
cool if you… And some of the best ideas I’ve ever worked on have
been

You know, I’m not saying I’m bad at coming up with ideas, but I’m also
saying that the creatives that we actually have at our agency are
better than I am at coming up with it. That is why they are the
creatives. They are the creative geniuses who come up with the really
cool stuff that I did not think of. So I’m there to support them right
after the light sparks in their brain about, wouldn’t it be kind of
cool if you… And some of the best ideas I’ve ever worked on have
been
Orlando Wood (12:41)
Right.
Yeah.
Yeah.
Martin Pagh Ludvigsen (13:05)
those where someone has
literally found me in the hallway or walked over to my desk or you
know recently hit me up on Teams and said hey we we have this crazy
thought. Do you have five minutes? And I always those are the best
conversations. I love those conversations. Yes, I absolutely have five
minutes to listen to your crazy idea. I think one of the ideas in
recent years that I’ve been the proudest of started that way. It’s a
project that we’ve done for the Dali Museum in St. Petersburg, Florida
called
Orlando Wood (13:16)
Yeah. Yeah. Yeah.
Yeah, I was gonna
ask about that.
Martin Pagh Ludvigsen (13:34)
Dali
and that is a project where can visit the museum, walk up to a
recreation of Salvador Dali’s famous lobster telephone and then you
can pick up that lobster telephone and have a conversation with
Salvador Dali.
Orlando Wood (13:42)
Yeah. Yeah.
Martin Pagh Ludvigsen (13:47)
That project started exactly like I just described, that we have these
two amazing creatives, Pedro and Fabio. They hit up my team in a group
chat and said, hey, do you think it would be possible to have a…
conversation with Salvador Ali using AI. And I think now in September
2025, everyone would know that the answer to that is yes. But in
September 2023, which is when I, if I remember correctly, so two years
ago, this idea was first asked. That was the answer. Not everyone knew
that the answer was yes. Let’s just say it that way. But what we knew
because of our

conversation with Salvador Ali using AI. And I think now in September
2025, everyone would know that the answer to that is yes. But in
September 2023, which is when I, if I remember correctly, so two years
ago, this idea was first asked. That was the answer. Not everyone knew
that the answer was yes. Let’s just say it that way. But what we knew
because of our
Orlando Wood (14:11)
Mm-hmm.
Yeah.
Martin Pagh Ludvigsen (14:27)
deep interest in cutting edge technology and everything that comes out
was that at that time, 11 Labs had just released a beta version of a
foundational model called multilingual V2, where they claimed that you
can, with a small sample of a voice, you can clone it and have a very
convincing conversation with it. So, two years ago, which AI years are
longer than dog years, so that is a very, very long time in AI time,
Orlando Wood (14:38)
Mm. Mm-hmm.
Martin Pagh Ludvigsen (14:56)
We got access to that beta and we found some samples of Salvador Alis
speaking English and then we worked with GPT, probably four, like an
earlier version of GPT at the time, to like come up with what are the
words he’s going to say based on his writing and things like that. And
then we built a prototype where we put those together. And I still
remember like we did a demo to these creatives and they were like
Orlando Wood (15:05)
Yeah.
Hmm.
Martin Pagh Ludvigsen (15:19)
wow. And very, very quickly, like brought this to Jeff Goodby, who
also said, we have to show this to the museum. And then we went to the
museum and we showed it to them and they were, yep, this,
Orlando Wood (15:20)
Yeah.
Yeah.
Martin Pagh Ludvigsen (15:30)
This is something we need to do with you guys. So this little ping on
Teams from someone, hey, we have a crazy idea. And now two years
later, it’s been an incredible success. We’ve won great awards for it.
It is one of our most talked about projects. It is still live at the
museum. We have lots of exciting ideas for how we wanna… That’s
true.

This is something we need to do with you guys. So this little ping on
Teams from someone, hey, we have a crazy idea. And now two years
later, it’s been an incredible success. We’ve won great awards for it.
It is one of our most talked about projects. It is still live at the
museum. We have lots of exciting ideas for how we wanna… That’s
true.
Orlando Wood (15:35)
Yeah. Yeah.
Yeah. There was a phone number you could call. I’m not sure it’s still
active,
but if you were there in person, you could pick up the lobster phone,
which is very cool. But you could also just from your mobile, like
call up a chat with Dolly, which is just a very cool thing to do. And
like you said, at the time, incredibly transformative. And now
everybody kind of takes that sort of stuff for granted, but
unbelievable that, know, it’s still very cool. Sorry, I wasn’t meaning
to imply that it was it. Yeah.
Martin Pagh Ludvigsen (15:56)
Yeah. Yeah.
Well, I mean, it’s still conversational AI. I think it’s cool, but
it’s also
like finding out.
Not every brand has a… Let’s for a moment call this the Salvador
Dalí Museum in St. Petersburg, Florida, a brand. They have an
incredible brand spokesperson, Salvador Dalí. The real Salvador Dalí
passed away in 1989, although some will say he’s still alive, and he’s
obviously still alive through his art. But the idea that you can now
have a conversation with him, really is transformative. The phone
number you mentioned, 772-ASK-DALÍ.
Orlando Wood (16:23)
Yeah.
Yeah.
Mm-hmm. Yeah. Yeah. Yeah.
Martin Pagh Ludvigsen (16:45)
You can actually call the phone number, but it goes to voicemail now,
which by the way is a fun experience. So I encourage
your listeners to call it. For a while around his birthday, this year
is 121st birthday, we wanted to bring the experience outside of the
museum. So the idea that you can now pick up your own phone, type 772-
ASDALI and call him was something that we put together fairly quickly.
But again, it was a huge success. And one of the things that I love to
see about this, I’ll tell you, we did

Orlando Wood (17:00)
Mm-hmm.
Martin Pagh Ludvigsen (17:12)
not record the conversations because we respect the privacy. But what
we could see was that we had a five minute limit on these phone calls
⁓ for like cost and scaling reasons so that people wouldn’t. But we
could see that in the middle of the night, the same people would call
back time and again and again until they had spoken to Salvador Dalí
for an hour in the middle of the night. So that tells me that, wow, we
made something here.
Orlando Wood (17:30)
Hmm.
yeah. Yeah.
Yeah. Yeah.
Martin Pagh Ludvigsen (17:37)
that really connected with people. imagine
creating a brand experience. If you could go out and say, hey, we
created an experience where people in the middle of the night by their
own choice type in your phone number and talks to your brand for an
hour. Just imagine being able to sell something like that to another
brand in the Salvador de Lima Museum. That’s a pretty big deal.
Orlando Wood (17:48)
Yeah. Yeah.
Well,
it’s the very definition of branded entertainment. Those people are
being fundamentally entertained by what’s happening. They are
participating in it, which gives them a sense of ownership. It gives
them a sense of interaction. It gives them a sense of being respected.
Listen, I think it’s a phenomenal project. It was very cool when it
came it’s funny because it brings up a couple of things. think that
another project that you guys have just recently done is the…
Martin Pagh Ludvigsen (18:02)
It is.
Yes. Yes.
Orlando Wood (18:21)
the BMW real be real, I think it’s called. ⁓ Real for real, sorry. ⁓
But you know, it’s interesting to me because obviously AI is such a
potentially incendiary topic, especially in advertising, especially in
advertising production. But what’s interesting is that there are
certain times where AI is used and everybody is totally fine with it.
And other times when AI is used and everybody is not.

the BMW real be real, I think it’s called. ⁓ Real for real, sorry. ⁓
But you know, it’s interesting to me because obviously AI is such a
potentially incendiary topic, especially in advertising, especially in
advertising production. But what’s interesting is that there are
certain times where AI is used and everybody is totally fine with it.
And other times when AI is used and everybody is not.
Martin Pagh Ludvigsen (18:25)
Real for real. Yes.
yes.
Orlando Wood (18:45)
I obviously mentioned, for example, the Coca-Cola ad that was
reproduced entirely in AI. You know, I think the fear around that is
that, it, you know, potentially proves that, productions can be done
with less, less money, less human intervention, things like that. but
then you have projects like the real for real.
and Ask Dolly. And there’s another one, there was a one out of the UK
that was an old lady who speaks on the phone and tricks ⁓ scam callers
into staying on the line and incredibly frustrates them. I mean,
that’s just fun. And I would love it if one day they released the
audio files. Yeah.
Martin Pagh Ludvigsen (19:12)
Yes.
That campaign, the way, I…
for sure, that is an incredible campaign. I’m an executive judge for
the Webbies, and I was very proud to award that campaign. That is one
of the best uses I’ve seen of AI. I think… you know, we have our two
examples. I can briefly talk about what BMW Real4Real is. It came out
this week, Tuesday this week, so the day after Labor Day.
Orlando Wood (19:38)
Yeah,
Martin Pagh Ludvigsen (19:42)
And the interesting thing is this is an ad for BMW certified pre-owned
program. Obviously, buying used cars in general is a wild west and I’m
sure we all imagine horror stories of like driving off the lot and the
wheel falls off immediately and all that.
Orlando Wood (19:47)
Mm.
Yeah.
Martin Pagh Ludvigsen (19:57)
a certification program, in particular one coming from BMW, you can
trust that that is not going to happen to you. So that is sort of the
product that we’re selling here. Now the inside is one, well, let’s
look at what is the Wild West in culture right now. And I think the
Wild West in culture is AI Slop, my number one contender for term of
the year in 2025.

a certification program, in particular one coming from BMW, you can
trust that that is not going to happen to you. So that is sort of the
product that we’re selling here. Now the inside is one, well, let’s
look at what is the Wild West in culture right now. And I think the
Wild West in culture is AI Slop, my number one contender for term of
the year in 2025.
Orlando Wood (20:02)
Mm-hmm, mm-hmm, yeah.
Mm-hmm.
Martin Pagh Ludvigsen (20:22)
It is content that is filling our feeds and it’s not great. And I
think…
Orlando Wood (20:28)
Yeah.
Martin Pagh Ludvigsen (20:29)
As an industry, we are probably contributing to that. But as you
mentioned earlier, a lot of these tools are now at a level where
literally anyone can use them, which means that is anyone can create
AI, and that means that anyone can also create AI. Slop and that,
unfortunately, is most of the time what it is. And that is what fills
our feeds. And that is boring. So the premise of the campaign is that
we have these two women who one of them is looking at her phone
and she sees something incredible. The scenes are, we have a pigeon
riding a skateboard and doing an ollie. We have a rock formation that
looks like a thumbs up. We have some butterflies picking up a wedding
cake. And she’s like, oh my God, this is amazing. And then very
quickly in the spot, it’s revealed that there, that’s AI. It’s AI. Oh,
really? So, so.
Orlando Wood (21:17)
Yeah.
Martin Pagh Ludvigsen (21:20)
So in a world where you can’t really trust anything anymore, you know,
the premise is then, but you can trust BMW’s certified program. And I
think it’s fair to say that we’ve really hit a nerve with this
campaign at a time where there is a lot of debate about AI slob, about
just content out there that you can’t really trust anymore. And
sometimes it looks bad. Sometimes it’s just uncanny and you don’t
really know what’s real. What is even real? Well, BMW is real for
real. And that’s the premise here.
Orlando Wood (21:22)
Hmm. Hmm.
Mm-hmm.

Hmm. Yeah.
Martin Pagh Ludvigsen (21:47)
I think we have worked with generative AI for a really long time. I
know it’s popular to say in advertising, well, we’ve worked with
artificial intelligence for a while. And as an industry, we have.
There’s a lot of artificial intelligence going on behind the scenes.
Orlando Wood (21:59)
Yeah, and there’s a big difference between, you know,
a lot of machine learning technology and, you know, and those sorts of
things that have been around for ages. And then suddenly everybody
comes around and generative AI is all they’re talking about. But
generative AI isn’t the whole story. Yeah.
Martin Pagh Ludvigsen (22:04)
Yeah.
It’s not, but generative
AI is, we’re one of the few agencies who can even say that we’ve been
working with generative AI for a long time because our first
generative AI project also for the Dali Museum came out as far back as
in 2019 using generative adversarial networks, still generative AI,
just an earlier generation of it. So we have on our team decades of
experience with this and not only do we have experience with the
technology, we also understand the culture really well. And when you
understand culture really well,
well and you start to see a reaction out there that okay there are
actually people who aren’t quite on board with this that’s an
opportunity to produce something that is counterculture to that so BMW
real for real is as far as I’m aware the first like AI counterculture
piece that has come out there and I think I think it’s fair to say
I’ve been working with automotive advertising for a very long time you
don’t
ever see certified program advertising getting picked up by all the
industry media, but this one did because we really hit a nerve with
what we’ve done here.
Orlando Wood (23:13)
Mm. Yeah.
Yeah, it seems to me that if I were to create a rule, and obviously,
the purpose of creativity is to undermine rules, but so I’m sure it’ll
be true. It’ll be false the moment it comes out of my mouth. But it
seems to me that AI injected at the strategy level and the creativity
level seems to do very well. Whereas AI injected into the production
and execution level.

Yeah, it seems to me that if I were to create a rule, and obviously,
the purpose of creativity is to undermine rules, but so I’m sure it’ll
be true. It’ll be false the moment it comes out of my mouth. But it
seems to me that AI injected at the strategy level and the creativity
level seems to do very well. Whereas AI injected into the production
and execution level.
seems to really turn people off. think part of that is about the
humanity. Part of it is about the general laziness or the perceived
laziness of the, the use of it. But the other one I’m reminded of is
the, the Heinz campaign where they, they asked for, they asked AI to
produce ketchup and every single thing that it produced was basically
a Heinz bottle. And they were like, well, this is, you know, what this
proves to us and to everybody is that
Martin Pagh Ludvigsen (23:53)
Mm-hmm.
Orlando Wood (24:05)
ketchup and Heinz are basically synonymous, using one of the major
bugs of video models to their advantage. Yeah.
Martin Pagh Ludvigsen (24:08)
Amigo
Kraft Heinz
is, as an organization, incredibly smart about generative AI. And this
is an example of output that goes out into the world where AI is very
much at the core of the idea. It’s also an example of an idea that
couldn’t exist without ⁓ AI. think a lot of the time what you see a
negative reaction to is when you see work that has been created with
AI that…
Orlando Wood (24:16)
Mm-hmm.
Mm. Mm-hmm.
Yeah. Yeah.
Martin Pagh Ludvigsen (24:36)
could have and maybe should have been created with more traditional
production means because you get a sense that this is now they’re just
taking shortcuts to get to the final result and the result the final
result
Orlando Wood (24:41)
Mm-hmm.
Martin Pagh Ludvigsen (24:48)
is not as good as it would have been if you had produced this the
right way. But when we have ideas like the Heinz idea that you
mentioned, our own Astalli, even BMW real for real, this idea would
not have been possible without the use of generative AI. That’s when I
think you have a much stronger case because now you are actually
creating original work using a new instrument in the orchestra, in
this case, generative AI.

is not as good as it would have been if you had produced this the
right way. But when we have ideas like the Heinz idea that you
mentioned, our own Astalli, even BMW real for real, this idea would
not have been possible without the use of generative AI. That’s when I
think you have a much stronger case because now you are actually
creating original work using a new instrument in the orchestra, in
this case, generative AI.
Orlando Wood (25:05)
Yeah. Yeah.
Martin Pagh Ludvigsen (25:16)
that actually enables you to create new types of ideas that wouldn’t
have been possible before. Whereas, as you said, the Coca-Cola example
from before was a recreation shot by shot of something that had been
done before. it’s like, well, is that OK? That’s a technology demo.
That’s all it is, right? It’s not something. It’s not an interesting
idea anymore. It’s just a demo.
Orlando Wood (25:28)
Yeah, yeah.
Yeah.
Yeah, yeah, it’s interesting. We had somebody on the podcast named
Doug Shapiro, who’s an analyst and a sort of thought leader in
entertainment. And one of the things that he said was actually,
whereas a lot of people are concerned about legacy brands, legacy
agencies, legacy creative organizations, he said, well, actually,
think that in a world where you can’t trust a lot of what’s coming
across your eyeballs,
trust is the new oil. If the social media era was defined by data and
algorithm being the oil that powered social media, well then trust is
going to be the new thing that empowers what he called, you know,
we’re moving from an attention economy to an intention economy because
the only way you’ll be able to know whether your campaign was
successful is not whether I watched it.
Martin Pagh Ludvigsen (26:07)
Yeah.
Mm-hmm.
Orlando Wood (26:29)
but actually whether I made an intention, I moved closer to buying and
that I trusted you enough to get closer to the point where I might
give you my credit card, where I might interact with you in a real
way, where I have to do something and spend either part of my time or
part of my hard earned money to interact with you. And I think that
that’s what, know, Real4Real certainly, illustrates very well is we’re
moving into a place where trust and authenticity is

actually gonna be the thing that drives a lot of this because we’ll
have so much empty calories thrown at us that we’ll need to know what
is actually nourishing.
Martin Pagh Ludvigsen (27:08)
absolutely. I mean, that’s I haven’t listened to that particular
episode, but it sounds like I should go back and listen to it because
what you just said is something that that that I agree so strongly
with. Trust is obviously incredibly important. And the challenge with
trust is that it takes a really long time to build it up, but it’s
very easy to tear it down and create ⁓ great mistrust instead. And if
a brand does that, if suddenly you can’t trust a brand anymore, then,
you you’re you have a problem.
Orlando Wood (27:27)
Mm.
Martin Pagh Ludvigsen (27:36)
because now you’ve gone from someone who could potentially be your
customer to someone, it will take a very, very long time for you to to
come back to you. And people…
Orlando Wood (27:37)
Mm-hmm.
Martin Pagh Ludvigsen (27:45)
seeing AI slob as are lot of people who see that and think I can’t
believe that brand XYZ would would use AI to do this. They have
violated my my trust in their authenticity. So so I’m now going to be
turned away by it. That that that is an incredibly important factor to
to take into consideration here and one that recently we’ve realized
how important that is. So so I fully agree with
And I think the reason a lot of people in business are so excited
about generative AI, well, it’s kind of obvious, It’s that they see a
way to increase the efficiency of getting things done. But that…
Orlando Wood (28:25)
⁓ Yeah.
Martin Pagh Ludvigsen (28:28)
In this case, comes at the cost of effectiveness because you might
have produced the work and gotten your idea out in the world in the
cheapest and fastest possible way. But when it reaches your target
audience, they may see it, but then they’re like, wait a minute, what
just happened here? Why did my brand do this to me? So you have failed
at the goal of your campaign, which is to build brand trust with them
and potentially getting them to take out that credit card or whatever
it is you want them to do.
Orlando Wood (28:55)

Orlando Wood (28:55)
Yeah, it’s interesting because we spoke to Sergio Lopez as well, know,
who’s CEO of OmniCom production. Obviously you guys are within the
OmniCom network. One of the things that I did think that he said that
I thought was really interesting was how time now is actually a
creative resource in terms of, you know, your ability to output
something that is culturally relevant in a world that’s moving so
fast. it’s so easy
Martin Pagh Ludvigsen (29:00)
Yeah.
Indeed.
Orlando Wood (29:21)
call them financial efficiencies or workflow efficiencies that are all
about a margin and a bottom line. But actually there’s a creative
impact there as well. And I think that that’s really interesting.
mean, he, one of things that he said is if a standard campaign takes
into 18 months, think about where we were 18 months ago. We still 18
months ago had Joe Biden running for president we hadn’t even gotten
into the Kamala Harris era.
if you are wanting to be as creative as possible, shortening that
timeline is actually a very smart strategy.
Martin Pagh Ludvigsen (29:52)
indeed. Yeah. And that’s obviously, I mean, you said this before that
generative AI and the strategy and creative phases of the work that we
do has just been incredibly powerful for us. There’s no turning that
back. Being able to very, very quickly iterate through briefs and
ideas to visualize things, to show things and be able to much more
quickly have these reviews that lead to the final work, even if the
final work is not produced with generative AI at all, but the creative
development, the strategy development phases,
Orlando Wood (29:56)
Mm. Mm.
Mm. Mm.
Martin Pagh Ludvigsen (30:20)
We’ve really, with Generative AI, reached an incredibly potential for
what we can do in the same time as before, just bigger output in that
time, even if the final, we’re creating one film, right? But we can
experiment a lot more within that idea than we’ve been able to before,
thanks to Generative AI.
Orlando Wood (30:35)
Mmm.

Yeah, that’s the other part I’d like to get into, because obviously
we’ve kind of talked about where AI is used in the strategy of the
campaign or the actual idea and the creative. it sounds like you and
your team are sort of such a support to the entire creative
organization in that way. But then, you know, there’s also the sort of
nuts and bolts of creative ideation. know, creative development is, is
basically R &D. if a
a car manufacturer would have an R &D department. It would be burning
through money, kind of coming up with new ways to light the interior
or break, using curves or whatever it is. but in a creative
organization, creative development is you spend a lot of time, a lot
of effort, a lot of money just iterating coming up with it. You know,
on a brief, you might send three or four different creative teams out
to to come up with with ideas that fit the brief.
you’re taking on board that time most of the time. Most brands don’t
pay for creative development. So you as an agency are, carrying that
time. What how do you work with the how does your team work with the
creative teams on that ideation, pitch deck creation, all of that kind
of stuff?
Martin Pagh Ludvigsen (31:47)
Well, don’t know exactly how the chemistry in the brain works, But the
idea, at some point, it’s a spark inside ⁓ the head of a creative
person. And that idea now needs to be articulated, put into a format
where other people can also get that spark lit in their head. this
really is, in my opinion,
Orlando Wood (31:51)
Right.
Mm-hmm.
Martin Pagh Ludvigsen (32:07)
has always been the biggest challenge and the most time consuming part
of our work. Actually, we are, in our case, paid for creative
development of ideas, which I think is important. I know that that’s
not always the case and it’s different from contract to contract, of
course. But the development of ideas, the actual creative magic, that
is our output. It is a very, very important ⁓ thing.
Orlando Wood (32:21)
Mm. Mm-hmm.
Mm-mm.
Martin Pagh Ludvigsen (32:32)
in what we do. And there are no generative, no large language models
that can come up with good ideas yet. They will regurgitate something
that has been done before, think of it in a new way. But this idea
that cuts, that based on a razor sharp insight, that cuts against
culture and really gets noticed, that still comes out of the brains of
creatives. In fact,

in what we do. And there are no generative, no large language models
that can come up with good ideas yet. They will regurgitate something
that has been done before, think of it in a new way. But this idea
that cuts, that based on a razor sharp insight, that cuts against
culture and really gets noticed, that still comes out of the brains of
creatives. In fact,
It’s funny, so many times when I’ve attended conferences and heard
presentations from some of the leaders of the great AI labs, that is
what they say. is what creativity is the thing that will survive,
depending on how dystopian you want to look at it, the transformation
into the AI age, because we still need our way of thinking, the way
our brain works, the mysterious way the chemistry in our brains works
to actually come up with the idea.
Orlando Wood (33:05)
Mm.
Martin Pagh Ludvigsen (33:20)
So that is still a very important thing. But then selling the idea,
that is where generative AI has just become so important for us.
Orlando Wood (33:30)
Because surely mood videos
are dead. mean, if just to explain mood videos used to be these like
cut ups, everybody would do that would maybe have the song that you
were going to use in the ad and it would have scenes from a million
different films that kind of generally got the vibe that the ad was
going to be. And that was a large presentation tool that people would
spend time. I would imagine that’s, you know, kind of an outdated
thing to do now. Like you should be almost. Yeah.
Martin Pagh Ludvigsen (33:34)
Well, yeah.
Yeah.
It is, it is, it
is. Because now, to GenerAV AI, we have the capabilities to instead
make something that is, in fact, much closer to the final film that
we’re going to make. And that is absolutely what we do, I think, very
early on before we really put this into system.
Orlando Wood (34:05)
Mm-hmm.
Martin Pagh Ludvigsen (34:13)
we would, our lab would be very involved in like working with various
open source models to help our creatives generate visuals, stills in
that case, like working through the various versions of flux and
stable diffusion and all that, because that’s where really the cutting
edge magic was happening. And then lately, obviously now that
generative video has gotten to a level where it’s actually worth your

we would, our lab would be very involved in like working with various
open source models to help our creatives generate visuals, stills in
that case, like working through the various versions of flux and
stable diffusion and all that, because that’s where really the cutting
edge magic was happening. And then lately, obviously now that
generative video has gotten to a level where it’s actually worth your
time. Our creatives can also do that. I think we have a, because we’re
part of Omnicom, we have access to a platform where everyone working
at our agency can now generate images and videos and, you know, a safe
environment where we do have access to the latest cutting edge models
from, know, OpenAI and Google and Runway and things like that. So
suddenly, instead of creating rep mood videos that we’ve done for, you
know, a long time, we can now make something that is closer to the
finished
Orlando Wood (35:08)
Mm-hmm.
Martin Pagh Ludvigsen (35:11)
thing so that when we present the idea to our client, they will see
not a mood, but something that is like this is 50, 60 % of what we
imagined the final idea will be. And by showing them that they have
the confidence to provide feedback that is relevant because you’re now
giving feedback on the idea and give the thumbs up or maybe sometimes
the thumbs down because they see it. Okay, I see your idea. I see what
you mean, but this is not what we’re looking for. And it’s not based
on like a mood.
Orlando Wood (35:25)
Mm.
Martin Pagh Ludvigsen (35:40)
or
vibe even you could say but instead it’s based on a much more precise
articulation of the idea and that really has unlocked something for us
in the creative industry.
Orlando Wood (35:42)
Yeah.
Mm-hmm.
Yeah, and I think the creatives love it because they, they get that
first kind of, know, they get to see it come alive, which, which to be
honest, most of the time up until generative AI came around, they
would pitch and pitch and pitch and pitch a pitch. sometimes a
creative team might not, if they were having a bad run, might not
produce something in a year. And so they don’t ever get to see an
output. And very often it can, the blame can kind of fall for them.

Martin Pagh Ludvigsen (36:09)
Yeah, true.
Orlando Wood (36:16)
on the process or the client or the this or the that. And now at least
you get to see your idea come to life in some way, which allows you to
fall more in love with it, become more passionate about the selling of
it and all that kind of stuff. And that’s all of a benefit, I think.
Martin Pagh Ludvigsen (36:24)
for sure.
But it also…
I think one important thing is that sometimes the opposite also
happens, that you see it come to life and then you as a creative,
that’s when you realize, hey, wait a minute, this is actually not
working as well as I thought it would when this was the spark inside
my head. So that also allows you to like kill an idea and move on to
the next one. And I think that is just as powerful as being able to
see a good idea come to life, being able to kill bad ideas because you
see them visualized and realize, hey, wait a minute, this isn’t quite
working.
Orlando Wood (36:40)
yeah.
Yeah. Yeah.
Mm.
Mm.
Martin Pagh Ludvigsen (37:01)
Maybe it’s not even your creative director who kills at that point.
You can do that yourself because you see for yourself that this isn’t
working. Let’s move on to the next one.
Orlando Wood (37:05)
Yeah.
Yeah, you mentioned in there working with like OpenAI it’s funny to me
thinking of OpenAI as a startup now with its, you know, nearly $500
billion valuation, but you know, yes, they are a startup, but you
know, in terms of, you know, your position in the Bay Area and
obviously you being in technology such a long time,
Martin Pagh Ludvigsen (37:17)
Well, technically they are, but yeah.

Orlando Wood (37:32)
Do you think it’s easier to collaborate with these startups and are
they as excited to work with you as you guys would be working with
them? does the Bay Area just as an environment make that an easier
conversation?
Martin Pagh Ludvigsen (37:47)
I think so. You know, as I mentioned, I’ve seen spoken on stages at
South by Southwest and other places that they creatives to bring their
ideas to life. think it’s actually, there’s a…
a researcher at Google DeepMind, Doug Eck, he introduced an idea
earlier today, I saw that in a video of AI as an instrument, as a
musical instrument. And the comparison he made was, if you think of
Roland, the Japanese music company, when they made the 808 drum
machine a long time ago, they…
Orlando Wood (38:10)
Right.
Martin Pagh Ludvigsen (38:22)
could never have imagined what creative people would turn that into.
And I think that’s such a wonderful way to think about it. mean, if I
had a quarter of every time I heard someone say AI is a tool.
Orlando Wood (38:25)
Yeah. Yeah.
Martin Pagh Ludvigsen (38:34)
I actually don’t agree with that. I don’t agree that AI is a tool.
It’s more than that. It’s a paradigm shift. Generative AI is a
paradigm shift. not just another Photoshop. But what I do see it as is
an instrument. So if you compare it to a piano or a drum machine…
Orlando Wood (38:36)
Mm.
Mmm.
Martin Pagh Ludvigsen (38:50)
You know, if my four year old sits down at the piano I have next to
me, what comes out of that? She’s having a good time, but I don’t
think a lot of people would enjoy listening to what comes out of that.
But if someone else, through Virgil Rosso, would sit down at the exact
same piano and play it, everyone would love to listen to what comes
out of that. And obviously that wouldn’t be possible without the piano
being there right next to me, none of those scenarios, but it also
wouldn’t be possible without the talent of the person playing.
Orlando Wood (39:01)

Mm.
Yeah.
Mm-hmm.
Martin Pagh Ludvigsen (39:20)
that instrument and there’s just so much wonderful stuff that comes
out of that. I think that the most interesting question isn’t what can
AI do, it is what new genres of creativity can we invent by playing
this new instrument? That’s where I think the synergy between us as a
flagship creative agency and the AI labs that come to us because they
do. OpenAI came to us to launch Dali 2,
Orlando Wood (39:21)
Mm-mm.
Yeah.
Martin Pagh Ludvigsen (39:48)
launched with the project called Dream Tapestry at the Dalí Museum.
Earlier this year, we released Giraffes on Horseback Salad, a film of
a of a lost script that Salvador Dalí wrote in the 1930s. The
Hollywood studio said it was impossible to make. But thanks to Vio
from Google Deep Minds, again Google came to us, asked what can you do
with this platform that we’ve created, we were able to realize this
masterpiece that had been, you know, it took almost a hundred
years from Dali imagining it to us bringing it to life. I think the
technology does not replace artists. It elevates.
Orlando Wood (40:25)
Mmm.
Martin Pagh Ludvigsen (40:26)
us, I’m calling myself an artist now, that might be a bit, so it
elevates them. But it automates the
Orlando Wood (40:26)
Yeah. Yeah.
Martin Pagh Ludvigsen (40:31)
brute force of execution and puts an incredible premium, in my
opinion, on the things that only humans can provide, the vision, the
taste, and the soul. And that is what we bring and what we will
continue bringing to creativity through AI.
Orlando Wood (40:45)
Yeah, I mean, in terms of your team as well, mean, how do you balance,
you know, experimentation, especially with technology that’s moving so
fast is super fun and, you know, and isn’t that great? But obviously
in creative organization like Goodby, you have the need for stable,
repeatable, and also highly, you know, specific workflows that allow
incredible amounts of controls.

Yeah, I mean, in terms of your team as well, mean, how do you balance,
you know, experimentation, especially with technology that’s moving so
fast is super fun and, you know, and isn’t that great? But obviously
in creative organization like Goodby, you have the need for stable,
repeatable, and also highly, you know, specific workflows that allow
incredible amounts of controls.
Martin Pagh Ludvigsen (40:56)
Mm-hmm.
Orlando Wood (41:09)
How do you balance that? have you struggled with that at all?
Martin Pagh Ludvigsen (41:13)
It’s always evolving is what I would say. also, one thing that I like
to say, as I said, my department is called labs within GSP. We are the
only department that is allowed to fail within our organization
because you can’t have a lab that does experiments if your expectation
is that everything succeeds, then it’s not a lab anymore. You’re not
working in unknown spaces. So we do experiments and sometimes those
experiments fail.
Orlando Wood (41:25)
Right. Yeah.
Martin Pagh Ludvigsen (41:41)
nothing comes out of them. But sometimes they succeed spectacularly.
And that is when we move on and create some of the great work that the
Labs team has created, including some of the people who’ve had the
director of Labs role before me. They’ve just done some incredible
work. When it becomes repeatable…
patterns as you described, something that other people can now do,
then that’s when it moves out of labs. So as I said, when two years
ago, we worked on the Xfinity Wicked spot with our creative team,
which was a really big idea, mean, it actually took a really long time
before it came out because it aligned with the release of the film.
But one of the challenges that we had there was that this was a really
big and
Orlando Wood (42:04)
Mmm.
Martin Pagh Ludvigsen (42:23)
expensive
production. involved Cynthia Rivo and Ariana Grande. They were in the
film. So a lot of stakeholders, a lot of people had to make some
decisions about making this commitment. And it was really important to
give them as much as possible to confidently say, yes, let’s go ahead
and do this. So a lot of experimentation was done on our end to
produce the assets that could help sell through this film. And that
was a success for us. We sold through the film. It’s a great film.
It’s real

production. involved Cynthia Rivo and Ariana Grande. They were in the
film. So a lot of stakeholders, a lot of people had to make some
decisions about making this commitment. And it was really important to
give them as much as possible to confidently say, yes, let’s go ahead
and do this. So a lot of experimentation was done on our end to
produce the assets that could help sell through this film. And that
was a success for us. We sold through the film. It’s a great film.
It’s real
Orlando Wood (42:37)
Mm-hmm.
Martin Pagh Ludvigsen (42:51)
tearjerker if you want to go out there and find it. But nowadays, they
don’t need to involve us to do something similar because we now have
given all of our creatives access to tools so that they can create
assets, storyboards, concept art, even video to help sell through
their ideas. So Labs is not needed for that process anymore. We can
instead move on to areas of technology and experimentation that are
still beyond the capabilities of our
Orlando Wood (42:52)
Mm-hmm. Mm-hmm.
Mm-hmm.
Mm-hmm.
Martin Pagh Ludvigsen (43:18)
rest of our creative department.
Orlando Wood (43:19)
Right. And are you using, know, are you, how much are you interacting
with the Omnicom? I know they’re, they’re obviously developing a lot
of AI tools and you know, and multimodal tools and things like that.
How are you engaging with, with those tools?
Martin Pagh Ludvigsen (43:33)
So the platform, think it was Sergio mentioned the name of it, it’s
called Omni AI, is actually a remarkable platform because…
Orlando Wood (43:38)
Mm-hmm.
Martin Pagh Ludvigsen (43:44)
The platform itself was developed by Omnicom by some very talented
engineers, but the underlying models come from the best AI labs in the
world. So that means that we can produce text, images, video, audio,
not by models that Omnicom has created or trained, but by models that
OpenAI and Google and Runway and even Meta and Amazon have trained.
that is the kind of stuff that we used to have.
to go to external platforms to work with those or even run open source
models on our own hardware. But now all of that is accessible through
a platform that anyone can access. And that has definitely been a game
changer for us. For, let’s say, the table stakes, generative AI stuff.

They don’t need to come to us in labs anymore. And the individual
curious creative can now sit down at their own laptop and do it on
their own. And I think that that’s incredible. And then sometimes they
find that, okay, they reach a point, is not, I can’t take this as far
as I had hoped for. And then they might still come to us. Is there
something else that you can provide that can help elevate this beyond
what Omni AI can do for me? And then we still get involved. That
happens almost
on a daily basis.
Orlando Wood (44:58)
And is
part of that also just the comfort of knowing that it’s clean from a
copyright and data provenance perspective? Yeah.
Martin Pagh Ludvigsen (45:05)
Absolutely. Absolutely.
mean, I think we all saw yesterday that Mid Journey was sued now by
Warner Brothers. yeah, so add that to the list of lawsuits. ⁓ Mid
Journey is an interesting case as a diffusion model, right? Because I
think it was the…
Orlando Wood (45:13)
Again, by one of his, yeah.
Martin Pagh Ludvigsen (45:24)
first one, I mean, there was a breakthrough moment in November 22 and
chat GPT came out where the world found out about generative AI. But
mid journey was the first diffusion model that most people could get
access to, even though, you know, I was surprised to see that, you
know, creatives figured out how to the syntax to prompt it through
discord. I was like, okay, wow, impressive. It’s not, it wasn’t
exactly a great user experience to begin with, but they still figured
it out. But
Orlando Wood (45:45)
Yeah, yeah, yeah, yeah.
Martin Pagh Ludvigsen (45:52)
I think it’s fair to say that Mid Journey has not exactly…
made the same deals with the content providers that some of the other
diffusion models out there have. And now we’re really starting to see
the consequences of that. I understand why a lot of people still use
Mid Journey because it does produce incredible quality output. they’ve
added video recently. So that is interesting. I think you have to ask
yourself.

made the same deals with the content providers that some of the other
diffusion models out there have. And now we’re really starting to see
the consequences of that. I understand why a lot of people still use
Mid Journey because it does produce incredible quality output. they’ve
added video recently. So that is interesting. I think you have to ask
yourself.
Orlando Wood (46:01)
Yeah.
Yeah.
Martin Pagh Ludvigsen (46:25)
as a creative when you use a tool like that. First of all, I mean, who
knows how these lawsuits will go? Who knows how any lawsuit will go in
this day and age, right? But that platform could come crashing down.
Orlando Wood (46:31)
Yeah. Yeah.
Martin Pagh Ludvigsen (46:37)
It feels like any moment. So I would not recommend that anyone becomes
too dependent on it. And we have, in fact, the policy at our place is
don’t use mid-journey because we know it produces great output, but
that output may be decided in a court of law that it comes from stolen
content. And we just cannot legally be in a position where we base our
generative AR on stolen content. So that’s an interesting case,
obviously.
Orlando Wood (46:39)
Hmm.
100%.
Yeah, absolutely. Yeah, we’re following it very closely. Obviously,
we’re a Warner Brothers accelerator company at Kubrick and Kubrick
Labs. for us, yes, yeah, yeah, exactly. So we’re happy to see it.
think that we interviewed Bryn Muzher from Moon Valley and Asteria.
They’ve got their Mary model, which is a validly.
Martin Pagh Ludvigsen (47:09)
Yeah. ⁓ I didn’t know that. But yeah.
Orlando Wood (47:26)
only licensed material in the underlying data set. It sounds like the
Omni AI is the same. I think that the models are getting slightly
easier to train. There’s obviously a capabilities gap when you use
underlying data with propriety, but I think that things are at least
moving in the right way. And I think that that makes me certainly a
lot more comfortable about the future and protecting artists’ rights.
Martin Pagh Ludvigsen (47:35)
Mm-hmm.
Yeah.

Orlando Wood (47:52)
than I was before. But we’ll see, as you say, it’ll all come down to
how it goes. I Anthropic obviously won their free use argument in the
first round and now it’s moving on to the second. So who knows, but
it’s all a developing scenario. But it’s nice to know that you and
your teams are keeping such a good eye on it.
Martin Pagh Ludvigsen (47:55)
for sure.
Mm-hmm.
It’s not just for legal reasons. also think that advertising has
existed in symbiosis with the artist community for decades. We depend
on their craft and skill to bring our ideas to life and in return, we
pay them. That’s actually really helpful. Then with that payment, they
can move on to refine their artistry, refine their craft, create work
that may not pay as well as advertising does.
Orlando Wood (48:23)
Mm. Mm.
Yeah.
Martin Pagh Ludvigsen (48:41)
the next time they need to pay rent, can come to us again and we can
work with them. I think, plus we find so much inspiration in the
artist community, we wouldn’t be here without them. mean, is not art.
Sometimes we create a piece of work that sort of like transcends
advertising and becomes art. I think that in two decades of working,
maybe I’ve made two or three such pieces that could
Orlando Wood (48:49)
Yeah. Yeah.
Martin Pagh Ludvigsen (49:07)
that can be elevated to that. But at the end of the day, know, it’s at
best art-serving capitalism. The true artists out there, they will
jump in and out of advertising as they need and we’re very happy to
work with them.
Orlando Wood (49:14)
Mm.
Yeah.
I certainly think that the Dali AI thing or the Ask Dali thing
elevates to the level of performance art. ⁓ And also, mean, you know,
it’s very, very cool. Well, anyway, Martin, we have got to get you to
kick off. It’s, you know, it’s Denmark is playing. We said we’d get
you out of here. We’re now three minutes over time. You must go and
get kickoff and good luck. Break a leg. All of the things.

I certainly think that the Dali AI thing or the Ask Dali thing
elevates to the level of performance art. ⁓ And also, mean, you know,
it’s very, very cool. Well, anyway, Martin, we have got to get you to
kick off. It’s, you know, it’s Denmark is playing. We said we’d get
you out of here. We’re now three minutes over time. You must go and
get kickoff and good luck. Break a leg. All of the things.
Martin Pagh Ludvigsen (49:28)
Yeah, I appreciate it.
You
Orlando Wood (49:46)
I hope the team wins and ⁓ good luck and enjoy the kickoff of the
World Cup.
Martin Pagh Ludvigsen (49:51)
I will, and thank you so much for having me on. It’s been an inspiring
conversation.
Orlando Wood (49:55)
It’s been a pleasure. could have, you if they weren’t kicking off, I
would have kept you around another hour and just kept picking your
brain. But I really appreciate you giving us the time. think what
you’re doing is awesome and, you know, and very impressive and, you
know, good luck with it. And thank you again.
Martin Pagh Ludvigsen (50:10)
Thank you.