July 12, 2023 • 24 mins
This week we return for the second part of our conversation centered around Artificial Intelligence– how it's transforming the marketing landscape, and what agencies and brands need to know in order to keep up with these rapidly evolving technologies.
Welcome to Part 2 of exploration of the role of AI in marketing on Partners in Possibility featuring special guests Jeff McDonald and Jason Harris.
Jason is the Co-Founder and CEO of the award-winning creative agency Mekanism, and host of the Soul and Science podcast.
Jeff is an AI thought leader, and the Director of Social Strategy at Mekanism. He's been named one to watch in this space by Ad-Age.com.
We’ll discuss how Chat GPT has democratized AI and become an accessible AI tool that anyone can use. The ease of use also lends itself to being able to work more efficiently individually and across teams. Speaking of teams-we also explore which department AI tools are destined to be a part of.
Learn more about Mekanism: https://mekanism.com/
Read Sprout Social’s interview with Jeff MacDonald: https://sproutsocial.com/insights/ai-social-media/
from plus company
Welcome to Part 2 of exploration of the role of AI in marketing on Partners in Possibility featuring special guests Jeff McDonald and Jason Harris.
Jason is the Co-Founder and CEO of the award-winning creative agency Mekanism, and host of the Soul and Science podcast.
Jeff is an AI thought leader, and the Director of Social Strategy at Mekanism. He's been named one to watch in this space by Ad-Age.com.
We’ll discuss how Chat GPT has democratized AI and become an accessible AI tool that anyone can use. The ease of use also lends itself to being able to work more efficiently individually and across teams. Speaking of teams-we also explore which department AI tools are destined to be a part of.
Learn more about Mekanism: https://mekanism.com/
Read Sprout Social’s interview with Jeff MacDonald: https://sproutsocial.com/insights/ai-social-media/
from plus company
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:02):
What is the level of the amountof humanity that we want to see in
the creative work that we look at, even if it's just an ad not
to say adds that, but Ithink that's important In an ever changing world,
campaigns are all about collaboration and connection, from an initial brief to the
(00:23):
follow through. We explore how innovativethought leadership makes a campaign success more than
a possibility. Welcome to Partners andPossibility. I'm your host, Brett Marshawn,
CEO of Plus Company, Except onceagain, I'm not actually Brett.
I'm an AI generated version of hisvoice. Today we're diving into part two
of our exploration of AI and it'squickly evolving role in marketing. In Part
(00:45):
one, Real Life, Brett Marshawnsat down with Jeff McDonald and Jason Harris
to discuss how chat GPTA made generativeAI something anyone could use and how marketers
should be using the II to worksmarter and faster. Jason is the co
founder and CEO of the award winningcreator of Agency Mechanism and host of the
Soul and Science podcast. Jeff isan AI thought leader and the director of
(01:06):
Social Strategy at Mechanism. He's beennamed one to Watch in this space by
that Age. Today we continue thatdiscussion and talk about some of the ways
Mechanism has used AI to encourage creativitywithin their agency, how this recent AI
boom is different than Web three,and what the future might look like.
Just this year, A plus Company, we brought a group together with representatives
from our various agencies to share AIbest practices, governance recommendations, and look
(01:30):
at partnerships. We decided to callit the Community of Practice. Here's Jeff
McDonald. I think the word practiceis so important in there. I think
getting in there and trying it outin different ways, finding it shortcomings and
finding where the bias lies, orwhere you're going to get a response that
(01:51):
isn't useful to you at all,or where you think it's not going to
save you time you could have cameup with that assumption faster is important,
so that way we can start findingout where those shortcomings lie in the tools
and find where we can make optimizations. I think one of the things we
did really early on at Mechanism iswe tried to get as many people onboarded
to AI as possible by just teachingthem a little bit about the tools that
(02:14):
are often in a fun way,we did this movie poster contest where people
generated movie posters, they generated thelog lines and the titles in the even
a fake movie review, and itquickly showed people what the opportunities were in
a fun, unique way. Butsecretly we also onboarded an entire agency on
mid Journey and chat chept so thatway they could start practicing and really starting
(02:36):
to see where it could implement,where it could be implemented in their work.
And I think that's with any agency. I think anyone or anyone in
the world should start at least seeingwhat it can do, and that way
they can really start understanding what theimpact is going to be. Yeah,
pret a good point. Kay,So where do you see this? Like,
when are we going to be ata point where we're going to see
campaigns that are AI generated the video, etc. Or will that happen?
(02:58):
Let's look two, three, fouryears down, which is a long time
in the world right now, Butwhere do you think this will go?
I mean, I think there's goingto be campaigns completely created by AI and
I think the sticky point right nowis it's similar to when songs were you
(03:21):
know, created samples or mashups,right, and and those songs were out
there and no one really said saidanything about it, and then all of
a sudden, those artists wanted paymentand they wanted to be compensated because part
of that song was a piece oftheir original work. And so I think,
you know, we're a couple ofyears away from that happening because and
(03:46):
I think there's some organizations that aretrying to create that that system of if
you use a portion of this imagewas created from this, you get a
fractional payment. And until there's thatsort of centralized resource that protects the brands
and the agencies from any liability,that is when you'll start to see work
(04:12):
created through those tools and put outthe world. And I think, you
know, I don't know why thisis not related to advertising, but I
always go back to um. Youknow that New York lawyer that used chat
ept for legal research. I thinkhe used it for some filings to liken
it to these case examples, butthose cases were made up, and so
(04:39):
that I think anyone in any serviceindustry had scared the hell out of them,
because all of a sudden, thisguy, I think his name was
Schwartz was like infamous and the judgeyou know, kind of threw the book
at him from having you know,bogus citations. So I think you're like,
okay, well these tools, we'removing too fast for this stuff to
(05:01):
be put out into public consumption.So someone so there's going to have to
be some organization that creates that thatyou know, legal entity to allow you
to create work that is actually publicfacing. And I think we're a couple
of years. Yeah, well thisand this is where the intersection with web
three and and the blockchain can bereally interesting, right because I think protecting
(05:27):
IP you know, whether that bevisual video, music ideas, etc.
I think, you know, itwill be really important. But also I
agree with you m thinking about guidelinespolicy process within organizations, which is something
we are doing, by the way, is really important because you don't want
to be in a situation where abrand has you know, is now on
(05:53):
the firing line because they have usedwhatever from somebody else that they don't have
the rights to use, right andit could be really dangerous. I think
there's two sides to that as well. There's the can we do it?
And right now inserting the can wedo it is Adobe. They're coming in
and saying, if you use AdobeFirefly their image generation algorithm, they're going
(06:15):
to indemnify you as a business fromany further lawsuits that happen from copyright claim
holders because they're so sure in thealgorithm that they've trained on photographs that have
no copyright their open public domain,or that they own the copyright, and
that they're compensating on their end thatthey're willing to protect you as a business.
They've already so they're going and takingthe step that says, we understand
(06:38):
this is new territory, and youwant to use this for commercial purposes,
and we're betting everything that we havethat you'll be able to do that without
legal liability. So that's amazing thatit Adobe is doing that. They're already
taking a jump forward more than anyonehas the capability to, which is great.
But that's the can we? Sothat's answering that the should we?
Or I think is the next stageof that conversation. What is going to
(07:01):
be public perception when you create anAD that has generated exclusively through AI tools.
When you hear a voiceover and youknow that that voiceover wasn't recorded by
an actual artist, or you watcha commercial that's got video and you know
that they didn't actually seld in thefilm, creative film, any of that.
What is the acceptance level that wehave as a culture to say that
(07:23):
that is what we want to beseeing that we only want to see image
generated by other computers. I thinkthat's the next step that we have to
come to an agreeance upon as acommunity. As a society, is what
is the level of the amount ofhumanity that we want to see in the
creative work that we look at,even if it's just an ad not to
say ads that, but I thinkthat's important. No, it's a good
(07:45):
point. Right, Barack Obama endorsessomething and you know it's Barack Obama,
you have a very different feeling aboutthat product, end or service that was
endorsed. Then if an AI generatedversion of Barack Obama did the same thing,
which lacks credit, you know,at least if you knew you would,
you would not feel the same wayfrom a credibility standpoint about that product
(08:09):
or service. So it is somethingthat I don't have to really think about.
So let's turn a little bit tothe future, because I know that
we're doing a bunch on the AIfront as our you know, other agencies
and marketers. You know. Oneof the things we have created as this
community of practice around AIM because there'syou know, there's it's happening in all
(08:30):
kinds of places and often behind closeddoors. Maybe talk a little bit about
the community of practice and where yousee AI helping us, helping our clients,
but also helping clients for themselves.Yeah, absolutely, I think Internally,
well, you know, we've talkeda lot about how AI tools can
help us just be more efficient atour jobs and do better at getting us
(08:52):
from a blank page to the workfaster. I think also some of the
things that it can do is makeus stronger at our work, not just
making us more efficient, but literallytrain us to be better. One of
the things that I love is anexample of how AI is being used in
education is that we've got two schoolsof thoughts happening. One where AI is
being blocked at schools, in thesecond where schools are actually implementing it into
(09:16):
their curriculum. And in one ofthe schools that's implemented it into their curriculum,
when students are writing persuasive essays,they have the AI write the argumentative
the opposite side, so that waythey can be better at writing an argument.
We're doing the same thing here instrategy briefs. We're putting in briefs
inside of AI to make it stronger, so we can question our assumptions,
(09:37):
so we can make it so thatwe're looking deeper at details, or maybe
we need to bring in some dataand facts to make our argument stronger.
So I think that's one of thethings that we're looking to do within training
other people inside the agency is howis it not just doing work for you
or generating ideas for you? Howis it making the ideas that you already
have better? That's super interesting.I mean, you know, we do
(10:00):
a campaign with a client and youcould ask AI to figure out what the
likely responses will be from the competitors, which is very difficult to do,
but they're probably going to be moreaccurate on something like that, which is
which is kind of interesting. Wehave used it in certain aspects to help
us prepare for pitches, where wewould query what questions clients might ask about
(10:24):
something that we're we're presenting, whichhelps you get the team. You know,
I'm big on rehearsal for pitches,but it helps me create questions to
ask the team to get them preppedfor pitches. So we've used it in
that way as well, Yeah,which is fascinating. I think the other
interesting thing is that it's gonna it'shaving impact in so many areas, right,
(10:45):
I mean, we are using itto be better and more accurate and
timely on expense reports, on timesheets, you know, which generate how we
charge clients for production and pre production, as we talked about earlier, for
writing social posts, as I talkedabout with the citizen example. For that
(11:07):
mean, we're as as you mayknow, we're actually using it to to
do optimization of actual client campaigns tofigure out, you know, how much
to spend where again against what targetsand messages, et cetera, which I
think could really make marketing much moreefficient and successful. But it's happening in
so many different places. My pointof view, and I'd be interested in
(11:31):
your point of view, is Ithink this is one of those things it's
tough to centralize, right because it'sjust happening everywhere. What do you think
is the right approach? Yeah,I think it has to be. That's
what's hard right, Like, rightnow, we think we're creating sort of
that center of excellence for AI,but ultimately it's going to touch every aspect
in every department. It's almost whenyou used to have when you used to
(11:54):
think of an agency is as asyou know, just digital or just social,
and all of a sudden you're like, well, everything's everything lives on
social platforms or everything's digital. Nowthat seems what's happening with AI. Where
it's it is it going to bea department Now it's going to be woven
(12:15):
into every single department in how thatdepartment is going to use AI to you
know, save time, deliver youknow, better quality, take on more
clients or more projects. So it'sgoing to be in kind of across every
service line, but it's going tobe used in different ways. We'll be
(12:39):
back after the break to talk aboutdata protection when it comes to AI and
what we can expect from this quicklyevolving technology in the coming years. Welcome
back to Partners and Possibility. I'mBrett Marshawn. One concern that seems to
(13:03):
come up again and again in discussionsabout AI is the issue of privacy.
But Jeff McDonald thinks that won't bean issue much longer. Everyone is so
concerned about this AI and the datathat they put in it being not protected
or being used in a nefarious way. And I think what's really interesting is
(13:26):
the next stage is people are goingto be rolling their own AIS as agencies
and as companies, and I thinkthat's really going to help loosen up what
people use the AI for. Likejust yesterday, I started rolling my own
stable diffusion of my own computer,and so everything I generate stays on this
computer. It's going to be thesame thing with companies. Everything they generate
stays within their companies, and thatknowledge set's going to stay within the company.
(13:48):
I think that's going to be superimportant. So you're saying that information
is not global, it's local.It stays on a localized machine. So
imagine if you were logged into theVPN for Plus company. We're all using
Plus COGPD, and I have itwrite a creative brief using a bunch of
secretive information from our clients that stayswithin our server in our secure enclave.
(14:11):
No one can see that data generation, but is what is helpful is that
everyone within the network gets to benefitfrom that by the AI getting smarter by
us seeing the responses and the promptsthat they wrote. So that's the next
stage of this is closed AI.Closed AI, which I know is separate
(14:31):
from where I normally preach from,which is this open, democratized network.
But I think so many people haveconcerns around privacy and especially around what the
EU put out with the AI Actthis week proposal. I think that's going
to be the next step. Ithink that's fascinating though, because we're going
to get asked by clients that question, right, are you using the tools?
(14:52):
And if so, what security measuresdo you have on those tools?
And we are thinking about this,by the way, like with Project Catalyst.
One of the reasons that we havedone things like the fifty five Rush
acquisition is because we are going tohave first party data and client data which
is going to be closed, andwe're going to combine that with you know,
(15:15):
the trillions of pieces of data thatare you can access in the you
know, on the Internet, inthe open market, in in things like
chatchyput and mid Journey. But whenyou combine the two of them, you
can actually create something that's you know, not just unique, but actually uniquely
powerful insightful and proprietary. Yeah,so I completely agree with you, Jeff.
(15:39):
I think that really is the future. I'd love to hear your point
of view and discuss a little bitthe difference between AI and Web three,
because you know, Web three inthe metaverse was everywhere. I mean,
companies like Facebook change their name,so and it feels like it's sort of
not gone because I don't get hasgone, but it's definitely waned and AI
(16:03):
has become the topic. Is thisjust a fad? Is how do you
see AI versus Web three? Job? Well, I think one of the
key differences here is that if youread through any of the headlines that were
written about with Web three, allof the applications that we people were talking
about in the news or around Idon't get and I don't want to mix
(16:26):
words because there's a lot of legislationgoing on in the US about this right
now. But speculative assets, let'sjust call it that around the NFTs in
cryptocurrencies, and I think one ofthe issues that we lost sight of in
Web three is we've talked about thoseobjects and we didn't talk about the underlying
technologies that could truly revolutionize the waythat consumers own their data. By having
(16:52):
you have the ability to own yourpersona and your data on the internet,
the ability to have frictionalist transactions wherethere's no intermediary. We lost the ability
to talk smartly about the blockchain andthe ways that the blockchain can impact every
single aspect of way we do business, even what we're doing with AI,
(17:14):
and it became just about that topic. And I think people very much are
willing to push all of Web threeout the door and say, oh,
well, because cryptocurrencies are down andbecause NFTs are down, web three is
just a fat It's not going tocatch on. It is in the sense
that maybe we you know, thenext time you work with a brand or
you are a brand, you maynot be recommending an NFT launch right now.
(17:40):
But I don't think AI is thesame way, in the sense that
blockchain is not also going away.AI is here. It's impacting the way
that we do our work. It'simpacting the way that we're going to be
looking at data, and obviously theway we're going to view images on the
Internet. We're going to have tobe a little bit more critical of whether
or not those images are real orif they've been created by an AI or
(18:02):
at least modified. So my hopeis is that we've learned what we need
to learn from the rise of Webthree and use it as a learning template
for how AI can be more impactfulin the background and be used at an
infrastructure level. And we don't seeit as this, Oh, let's just
(18:22):
generate something with AI and then nevertalk about it ever again. We need
to look at how AI can beimplemented in every part of our roles and
not just as this flashy new thing, because if we do that, it
will be a fad. It won'tbe something that we jump onto and use
as a way to revolutionize the waythat we do work. Yeah, someone
tells me i's not a fad.I think I read some stat that three
(18:45):
percent of the population own an NFTand almost thirty percent have tried AI.
So if you if you look atthat kind of shows that you know,
the adoption has been much much faster, and that feels like it's going to
(19:07):
be here to stay. Yeah.I mean, we could have a whole
other episode about the intersection of Webthree, especially both the blockchain and DOWS
combined with AI, and then addquantum computing when that comes along, and
you know, that could be awhole new world, But we'll say that
maybe for another episode. How's that. AI is definitely having a moment and
(19:32):
the most cutting edge tools are availablefor basically anyone to use me. AI
Brett is a perfect example. Wejust interview audio and recorded consent. We
were able to create a relatively convincingAI version of my or Brett Marshawn's voice.
So many of the tools and ofcreation and the tools of image generation,
(19:53):
things like that. For years ofbeing gatecap around it being too technical
and it's too hard for you tolearn. Oh you want to become a
video editor because you're really interested inbeing a film director, Well, the
tools are pretty technical. And allthat we've been doing over the last several
decades is making it so we canclose that gap and make it so that
every tool is democratized an easier touse, and AI just does that at
(20:15):
every level. It makes it sothat you can go into Photoshop and type
exactly what you want, or movinginto video editing, you're going to be
able to go into a video editingtool and type exactly what you want.
I just saw a really amazing exampleof a website generation tool where you can
type in exactly what a website isthat you want. This next generation of
(20:36):
creators, they're going to have thisopportunity that no one has had in the
past where if they can think it, they can make it. Yeah,
they don't have to no code,right exactly. I have a friend and
who lives in Silicon Valley who's gottwo younger kids who once chet YPT came
out, he took his kids outof coding classes because he's like, this
is not going to be necessary.Oh yeah, that's amazing. I think.
(20:56):
I think one scary thing about umAI and and technology is and that
we're facing in the the agency levelis you know, there's a there's a
declining sense of belonging in the USUM with close friendships on massive decline intimate
(21:18):
relationships. I think with the proliferationof technology and coupled with the difficulty of
of O RTO and getting people tointeract and socialize. One flag for me
as an agency you know leader isit it's a it's a tool that is
another tool of um you know ina sense isolation where you can kind of
(21:42):
have that do the work for youand it's less of a I'm I'm stuck.
We need we need to crack thistogether. Um, So that's that's
just something in the back of mymind. Yeah, especially when you add
bias in the bias of AI tothat. Right. So I think it's
a very good point and well,hopefully Jeff's right, and what AI is
(22:03):
going to be about is getting peopleoff their computers. So I think it's
spend more time connecting socially, whichis super important for happiness as we all
know, but it's also super importantfor creativity, you know, which is
to talk to other people with differentpoints of views and find intersections, which
I'm not sure AI can do.Frankly, that's the next stage of AI
tools in my opinion, Like ifyou wanted to take a bet and you
(22:26):
were to generate an AI tool today, I would make a social AI platform
where people are using generative AI togetherto make a better idea because you're bringing
in multiple taste the same way thatAI works by bringing in thousands of data
points all at the same time tomake a decision. If you add that
layer and you have twenty team membersat your agency helping drive where that AI
(22:48):
is headed with let's say the nextpitch deck is created using AI. That
makes it that much more powerful becausehe's you have all of those extra inputs,
those better inputs coming from all thosepoints of views. So my hope
business that instead of AI tools beingthis social isolation tool where you're constantly just
chatting with this computer and not areal person to help make your ideas better,
(23:10):
that will actually become an opportunity foryou to be able to have a
second person over your shoulder when you'reusing chat TBT and you can be writing
in your inputs or responses at thesame time. That's where I hope that
it heads next, and as westart to see how things only get better
as we bring in more human elementsinto this artificial intelligence world. Right listen,
this has been really interesting. Iwas already super excited about AI and
(23:36):
what's going on and it's future andas you guys have said, what it
can do for us and for ourclients. But I'm even more excited now.
And I'm glad we have you guyson top of this stuff, by
the way, So congratulations for whatyou've done up until now, and good
luck with all the work that's comingand all the opportunities that are in front
of us. So thanks for joiningme, Thank you for listening to Partners
(24:02):
and Possibility. That was part twoof my discussion with Jason Harris, co
founder and CEO of Mechanism and directorof Social Strategy Jeff McDonald. I hope
this has been useful in helping youcreate and develop your own relationship with AI
and keep an eye out for thatcollaborative AI that Jeff predicted. I think
you might be onto something
What is the level of the amountof humanity that we want to see in
the creative work that we look at, even if it's just an ad not
to say adds that, but Ithink that's important In an ever changing world,
campaigns are all about collaboration and connection, from an initial brief to the
(00:23):
follow through. We explore how innovativethought leadership makes a campaign success more than
a possibility. Welcome to Partners andPossibility. I'm your host, Brett Marshawn,
CEO of Plus Company, Except onceagain, I'm not actually Brett.
I'm an AI generated version of hisvoice. Today we're diving into part two
of our exploration of AI and it'squickly evolving role in marketing. In Part
(00:45):
one, Real Life, Brett Marshawnsat down with Jeff McDonald and Jason Harris
to discuss how chat GPTA made generativeAI something anyone could use and how marketers
should be using the II to worksmarter and faster. Jason is the co
founder and CEO of the award winningcreator of Agency Mechanism and host of the
Soul and Science podcast. Jeff isan AI thought leader and the director of
(01:06):
Social Strategy at Mechanism. He's beennamed one to Watch in this space by
that Age. Today we continue thatdiscussion and talk about some of the ways
Mechanism has used AI to encourage creativitywithin their agency, how this recent AI
boom is different than Web three,and what the future might look like.
Just this year, A plus Company, we brought a group together with representatives
from our various agencies to share AIbest practices, governance recommendations, and look
(01:30):
at partnerships. We decided to callit the Community of Practice. Here's Jeff
McDonald. I think the word practiceis so important in there. I think
getting in there and trying it outin different ways, finding it shortcomings and
finding where the bias lies, orwhere you're going to get a response that
(01:51):
isn't useful to you at all,or where you think it's not going to
save you time you could have cameup with that assumption faster is important,
so that way we can start findingout where those shortcomings lie in the tools
and find where we can make optimizations. I think one of the things we
did really early on at Mechanism iswe tried to get as many people onboarded
to AI as possible by just teachingthem a little bit about the tools that
(02:14):
are often in a fun way,we did this movie poster contest where people
generated movie posters, they generated thelog lines and the titles in the even
a fake movie review, and itquickly showed people what the opportunities were in
a fun, unique way. Butsecretly we also onboarded an entire agency on
mid Journey and chat chept so thatway they could start practicing and really starting
(02:36):
to see where it could implement,where it could be implemented in their work.
And I think that's with any agency. I think anyone or anyone in
the world should start at least seeingwhat it can do, and that way
they can really start understanding what theimpact is going to be. Yeah,
pret a good point. Kay,So where do you see this? Like,
when are we going to be ata point where we're going to see
campaigns that are AI generated the video, etc. Or will that happen?
(02:58):
Let's look two, three, fouryears down, which is a long time
in the world right now, Butwhere do you think this will go?
I mean, I think there's goingto be campaigns completely created by AI and
I think the sticky point right nowis it's similar to when songs were you
(03:21):
know, created samples or mashups,right, and and those songs were out
there and no one really said saidanything about it, and then all of
a sudden, those artists wanted paymentand they wanted to be compensated because part
of that song was a piece oftheir original work. And so I think,
you know, we're a couple ofyears away from that happening because and
(03:46):
I think there's some organizations that aretrying to create that that system of if
you use a portion of this imagewas created from this, you get a
fractional payment. And until there's thatsort of centralized resource that protects the brands
and the agencies from any liability,that is when you'll start to see work
(04:12):
created through those tools and put outthe world. And I think, you
know, I don't know why thisis not related to advertising, but I
always go back to um. Youknow that New York lawyer that used chat
ept for legal research. I thinkhe used it for some filings to liken
it to these case examples, butthose cases were made up, and so
(04:39):
that I think anyone in any serviceindustry had scared the hell out of them,
because all of a sudden, thisguy, I think his name was
Schwartz was like infamous and the judgeyou know, kind of threw the book
at him from having you know,bogus citations. So I think you're like,
okay, well these tools, we'removing too fast for this stuff to
(05:01):
be put out into public consumption.So someone so there's going to have to
be some organization that creates that thatyou know, legal entity to allow you
to create work that is actually publicfacing. And I think we're a couple
of years. Yeah, well thisand this is where the intersection with web
three and and the blockchain can bereally interesting, right because I think protecting
(05:27):
IP you know, whether that bevisual video, music ideas, etc.
I think, you know, itwill be really important. But also I
agree with you m thinking about guidelinespolicy process within organizations, which is something
we are doing, by the way, is really important because you don't want
to be in a situation where abrand has you know, is now on
(05:53):
the firing line because they have usedwhatever from somebody else that they don't have
the rights to use, right andit could be really dangerous. I think
there's two sides to that as well. There's the can we do it?
And right now inserting the can wedo it is Adobe. They're coming in
and saying, if you use AdobeFirefly their image generation algorithm, they're going
(06:15):
to indemnify you as a business fromany further lawsuits that happen from copyright claim
holders because they're so sure in thealgorithm that they've trained on photographs that have
no copyright their open public domain,or that they own the copyright, and
that they're compensating on their end thatthey're willing to protect you as a business.
They've already so they're going and takingthe step that says, we understand
(06:38):
this is new territory, and youwant to use this for commercial purposes,
and we're betting everything that we havethat you'll be able to do that without
legal liability. So that's amazing thatit Adobe is doing that. They're already
taking a jump forward more than anyonehas the capability to, which is great.
But that's the can we? Sothat's answering that the should we?
Or I think is the next stageof that conversation. What is going to
(07:01):
be public perception when you create anAD that has generated exclusively through AI tools.
When you hear a voiceover and youknow that that voiceover wasn't recorded by
an actual artist, or you watcha commercial that's got video and you know
that they didn't actually seld in thefilm, creative film, any of that.
What is the acceptance level that wehave as a culture to say that
(07:23):
that is what we want to beseeing that we only want to see image
generated by other computers. I thinkthat's the next step that we have to
come to an agreeance upon as acommunity. As a society, is what
is the level of the amount ofhumanity that we want to see in the
creative work that we look at,even if it's just an ad not to
say ads that, but I thinkthat's important. No, it's a good
(07:45):
point. Right, Barack Obama endorsessomething and you know it's Barack Obama,
you have a very different feeling aboutthat product, end or service that was
endorsed. Then if an AI generatedversion of Barack Obama did the same thing,
which lacks credit, you know,at least if you knew you would,
you would not feel the same wayfrom a credibility standpoint about that product
(08:09):
or service. So it is somethingthat I don't have to really think about.
So let's turn a little bit tothe future, because I know that
we're doing a bunch on the AIfront as our you know, other agencies
and marketers. You know. Oneof the things we have created as this
community of practice around AIM because there'syou know, there's it's happening in all
(08:30):
kinds of places and often behind closeddoors. Maybe talk a little bit about
the community of practice and where yousee AI helping us, helping our clients,
but also helping clients for themselves.Yeah, absolutely, I think Internally,
well, you know, we've talkeda lot about how AI tools can
help us just be more efficient atour jobs and do better at getting us
(08:52):
from a blank page to the workfaster. I think also some of the
things that it can do is makeus stronger at our work, not just
making us more efficient, but literallytrain us to be better. One of
the things that I love is anexample of how AI is being used in
education is that we've got two schoolsof thoughts happening. One where AI is
being blocked at schools, in thesecond where schools are actually implementing it into
(09:16):
their curriculum. And in one ofthe schools that's implemented it into their curriculum,
when students are writing persuasive essays,they have the AI write the argumentative
the opposite side, so that waythey can be better at writing an argument.
We're doing the same thing here instrategy briefs. We're putting in briefs
inside of AI to make it stronger, so we can question our assumptions,
(09:37):
so we can make it so thatwe're looking deeper at details, or maybe
we need to bring in some dataand facts to make our argument stronger.
So I think that's one of thethings that we're looking to do within training
other people inside the agency is howis it not just doing work for you
or generating ideas for you? Howis it making the ideas that you already
have better? That's super interesting.I mean, you know, we do
(10:00):
a campaign with a client and youcould ask AI to figure out what the
likely responses will be from the competitors, which is very difficult to do,
but they're probably going to be moreaccurate on something like that, which is
which is kind of interesting. Wehave used it in certain aspects to help
us prepare for pitches, where wewould query what questions clients might ask about
(10:24):
something that we're we're presenting, whichhelps you get the team. You know,
I'm big on rehearsal for pitches,but it helps me create questions to
ask the team to get them preppedfor pitches. So we've used it in
that way as well, Yeah,which is fascinating. I think the other
interesting thing is that it's gonna it'shaving impact in so many areas, right,
(10:45):
I mean, we are using itto be better and more accurate and
timely on expense reports, on timesheets, you know, which generate how we
charge clients for production and pre production, as we talked about earlier, for
writing social posts, as I talkedabout with the citizen example. For that
(11:07):
mean, we're as as you mayknow, we're actually using it to to
do optimization of actual client campaigns tofigure out, you know, how much
to spend where again against what targetsand messages, et cetera, which I
think could really make marketing much moreefficient and successful. But it's happening in
so many different places. My pointof view, and I'd be interested in
(11:31):
your point of view, is Ithink this is one of those things it's
tough to centralize, right because it'sjust happening everywhere. What do you think
is the right approach? Yeah,I think it has to be. That's
what's hard right, Like, rightnow, we think we're creating sort of
that center of excellence for AI,but ultimately it's going to touch every aspect
in every department. It's almost whenyou used to have when you used to
(11:54):
think of an agency is as asyou know, just digital or just social,
and all of a sudden you're like, well, everything's everything lives on
social platforms or everything's digital. Nowthat seems what's happening with AI. Where
it's it is it going to bea department Now it's going to be woven
(12:15):
into every single department in how thatdepartment is going to use AI to you
know, save time, deliver youknow, better quality, take on more
clients or more projects. So it'sgoing to be in kind of across every
service line, but it's going tobe used in different ways. We'll be
(12:39):
back after the break to talk aboutdata protection when it comes to AI and
what we can expect from this quicklyevolving technology in the coming years. Welcome
back to Partners and Possibility. I'mBrett Marshawn. One concern that seems to
(13:03):
come up again and again in discussionsabout AI is the issue of privacy.
But Jeff McDonald thinks that won't bean issue much longer. Everyone is so
concerned about this AI and the datathat they put in it being not protected
or being used in a nefarious way. And I think what's really interesting is
(13:26):
the next stage is people are goingto be rolling their own AIS as agencies
and as companies, and I thinkthat's really going to help loosen up what
people use the AI for. Likejust yesterday, I started rolling my own
stable diffusion of my own computer,and so everything I generate stays on this
computer. It's going to be thesame thing with companies. Everything they generate
stays within their companies, and thatknowledge set's going to stay within the company.
(13:48):
I think that's going to be superimportant. So you're saying that information
is not global, it's local.It stays on a localized machine. So
imagine if you were logged into theVPN for Plus company. We're all using
Plus COGPD, and I have itwrite a creative brief using a bunch of
secretive information from our clients that stayswithin our server in our secure enclave.
(14:11):
No one can see that data generation, but is what is helpful is that
everyone within the network gets to benefitfrom that by the AI getting smarter by
us seeing the responses and the promptsthat they wrote. So that's the next
stage of this is closed AI.Closed AI, which I know is separate
(14:31):
from where I normally preach from,which is this open, democratized network.
But I think so many people haveconcerns around privacy and especially around what the
EU put out with the AI Actthis week proposal. I think that's going
to be the next step. Ithink that's fascinating though, because we're going
to get asked by clients that question, right, are you using the tools?
(14:52):
And if so, what security measuresdo you have on those tools?
And we are thinking about this,by the way, like with Project Catalyst.
One of the reasons that we havedone things like the fifty five Rush
acquisition is because we are going tohave first party data and client data which
is going to be closed, andwe're going to combine that with you know,
(15:15):
the trillions of pieces of data thatare you can access in the you
know, on the Internet, inthe open market, in in things like
chatchyput and mid Journey. But whenyou combine the two of them, you
can actually create something that's you know, not just unique, but actually uniquely
powerful insightful and proprietary. Yeah,so I completely agree with you, Jeff.
(15:39):
I think that really is the future. I'd love to hear your point
of view and discuss a little bitthe difference between AI and Web three,
because you know, Web three inthe metaverse was everywhere. I mean,
companies like Facebook change their name,so and it feels like it's sort of
not gone because I don't get hasgone, but it's definitely waned and AI
(16:03):
has become the topic. Is thisjust a fad? Is how do you
see AI versus Web three? Job? Well, I think one of the
key differences here is that if youread through any of the headlines that were
written about with Web three, allof the applications that we people were talking
about in the news or around Idon't get and I don't want to mix
(16:26):
words because there's a lot of legislationgoing on in the US about this right
now. But speculative assets, let'sjust call it that around the NFTs in
cryptocurrencies, and I think one ofthe issues that we lost sight of in
Web three is we've talked about thoseobjects and we didn't talk about the underlying
technologies that could truly revolutionize the waythat consumers own their data. By having
(16:52):
you have the ability to own yourpersona and your data on the internet,
the ability to have frictionalist transactions wherethere's no intermediary. We lost the ability
to talk smartly about the blockchain andthe ways that the blockchain can impact every
single aspect of way we do business, even what we're doing with AI,
(17:14):
and it became just about that topic. And I think people very much are
willing to push all of Web threeout the door and say, oh,
well, because cryptocurrencies are down andbecause NFTs are down, web three is
just a fat It's not going tocatch on. It is in the sense
that maybe we you know, thenext time you work with a brand or
you are a brand, you maynot be recommending an NFT launch right now.
(17:40):
But I don't think AI is thesame way, in the sense that
blockchain is not also going away.AI is here. It's impacting the way
that we do our work. It'simpacting the way that we're going to be
looking at data, and obviously theway we're going to view images on the
Internet. We're going to have tobe a little bit more critical of whether
or not those images are real orif they've been created by an AI or
(18:02):
at least modified. So my hopeis is that we've learned what we need
to learn from the rise of Webthree and use it as a learning template
for how AI can be more impactfulin the background and be used at an
infrastructure level. And we don't seeit as this, Oh, let's just
(18:22):
generate something with AI and then nevertalk about it ever again. We need
to look at how AI can beimplemented in every part of our roles and
not just as this flashy new thing, because if we do that, it
will be a fad. It won'tbe something that we jump onto and use
as a way to revolutionize the waythat we do work. Yeah, someone
tells me i's not a fad.I think I read some stat that three
(18:45):
percent of the population own an NFTand almost thirty percent have tried AI.
So if you if you look atthat kind of shows that you know,
the adoption has been much much faster, and that feels like it's going to
(19:07):
be here to stay. Yeah.I mean, we could have a whole
other episode about the intersection of Webthree, especially both the blockchain and DOWS
combined with AI, and then addquantum computing when that comes along, and
you know, that could be awhole new world, But we'll say that
maybe for another episode. How's that. AI is definitely having a moment and
(19:32):
the most cutting edge tools are availablefor basically anyone to use me. AI
Brett is a perfect example. Wejust interview audio and recorded consent. We
were able to create a relatively convincingAI version of my or Brett Marshawn's voice.
So many of the tools and ofcreation and the tools of image generation,
(19:53):
things like that. For years ofbeing gatecap around it being too technical
and it's too hard for you tolearn. Oh you want to become a
video editor because you're really interested inbeing a film director, Well, the
tools are pretty technical. And allthat we've been doing over the last several
decades is making it so we canclose that gap and make it so that
every tool is democratized an easier touse, and AI just does that at
(20:15):
every level. It makes it sothat you can go into Photoshop and type
exactly what you want, or movinginto video editing, you're going to be
able to go into a video editingtool and type exactly what you want.
I just saw a really amazing exampleof a website generation tool where you can
type in exactly what a website isthat you want. This next generation of
(20:36):
creators, they're going to have thisopportunity that no one has had in the
past where if they can think it, they can make it. Yeah,
they don't have to no code,right exactly. I have a friend and
who lives in Silicon Valley who's gottwo younger kids who once chet YPT came
out, he took his kids outof coding classes because he's like, this
is not going to be necessary.Oh yeah, that's amazing. I think.
(20:56):
I think one scary thing about umAI and and technology is and that
we're facing in the the agency levelis you know, there's a there's a
declining sense of belonging in the USUM with close friendships on massive decline intimate
(21:18):
relationships. I think with the proliferationof technology and coupled with the difficulty of
of O RTO and getting people tointeract and socialize. One flag for me
as an agency you know leader isit it's a it's a tool that is
another tool of um you know ina sense isolation where you can kind of
(21:42):
have that do the work for youand it's less of a I'm I'm stuck.
We need we need to crack thistogether. Um, So that's that's
just something in the back of mymind. Yeah, especially when you add
bias in the bias of AI tothat. Right. So I think it's
a very good point and well,hopefully Jeff's right, and what AI is
(22:03):
going to be about is getting peopleoff their computers. So I think it's
spend more time connecting socially, whichis super important for happiness as we all
know, but it's also super importantfor creativity, you know, which is
to talk to other people with differentpoints of views and find intersections, which
I'm not sure AI can do.Frankly, that's the next stage of AI
tools in my opinion, Like ifyou wanted to take a bet and you
(22:26):
were to generate an AI tool today, I would make a social AI platform
where people are using generative AI togetherto make a better idea because you're bringing
in multiple taste the same way thatAI works by bringing in thousands of data
points all at the same time tomake a decision. If you add that
layer and you have twenty team membersat your agency helping drive where that AI
(22:48):
is headed with let's say the nextpitch deck is created using AI. That
makes it that much more powerful becausehe's you have all of those extra inputs,
those better inputs coming from all thosepoints of views. So my hope
business that instead of AI tools beingthis social isolation tool where you're constantly just
chatting with this computer and not areal person to help make your ideas better,
(23:10):
that will actually become an opportunity foryou to be able to have a
second person over your shoulder when you'reusing chat TBT and you can be writing
in your inputs or responses at thesame time. That's where I hope that
it heads next, and as westart to see how things only get better
as we bring in more human elementsinto this artificial intelligence world. Right listen,
this has been really interesting. Iwas already super excited about AI and
(23:36):
what's going on and it's future andas you guys have said, what it
can do for us and for ourclients. But I'm even more excited now.
And I'm glad we have you guyson top of this stuff, by
the way, So congratulations for whatyou've done up until now, and good
luck with all the work that's comingand all the opportunities that are in front
of us. So thanks for joiningme, Thank you for listening to Partners
(24:02):
and Possibility. That was part twoof my discussion with Jason Harris, co
founder and CEO of Mechanism and directorof Social Strategy Jeff McDonald. I hope
this has been useful in helping youcreate and develop your own relationship with AI
and keep an eye out for thatcollaborative AI that Jeff predicted. I think
you might be onto something
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
My Favorite Murder with Karen Kilgariff and Georgia Hardstark
My Favorite Murder is a true crime comedy podcast hosted by Karen Kilgariff and Georgia Hardstark. Each week, Karen and Georgia share compelling true crimes and hometown stories from friends and listeners. Since MFM launched in January of 2016, Karen and Georgia have shared their lifelong interest in true crime and have covered stories of infamous serial killers like the Night Stalker, mysterious cold cases, captivating cults, incredible survivor stories and important events from history like the Tulsa race massacre of 1921. My Favorite Murder is part of the Exactly Right podcast network that provides a platform for bold, creative voices to bring to life provocative, entertaining and relatable stories for audiences everywhere. The Exactly Right roster of podcasts covers a variety of topics including historic true crime, comedic interviews and news, science, pop culture and more. Podcasts on the network include Buried Bones with Kate Winkler Dawson and Paul Holes, That's Messed Up: An SVU Podcast, This Podcast Will Kill You, Bananas and more.