Chris Duffey from Adobe on Generative AI

Chris (00:05):

My hope and hypothesis is the CMO, the marketer is going to be the central source of truth for a business. And so what I mean by that is it is now going to house all of the insights because, with Generative AI and Agentic AI, we have the capability to create faster, more insightful, and have a faster round trip of in marketplace performance and have that data come back to not only inform marketing disciplines but also the product, the sales and so on. And so, I think we're gonna see increased value of marketers going forward because of Generative AI and Agentic AI.

Jim (00:44):

Welcome to What If? So What?, The podcast where we explore what's possible with digital and discover how to make it real in your business. I'm your host, Jim Hertzfeld. We get s**t done by asking digital leaders the right questions: what if, so what, and most importantly, now what? So I'm very excited today to have Chris Duffey on the podcast. Chris leads global partnerships for GenAI for Adobe. Chris, welcome to What If? So What?

Chris (01:07):

Great to be here, Jim. Great to reconnect.

Jim (01:10):

So Chris, give us a background on your focus at Adobe. I know you've done a lot in your career, so what is your focus now, and what brought you to this role within Adobe?

Chris (01:18):

Yeah, it's been quite an exciting journey from a long-term perspective, but even more so from the last 12 months or so. It's been a wonderful journey in terms of the adoption of Generative AI, specifically into the marketing realm. And we're seeing that it truly is unleashing, unlocking infinite possibilities for everyone across the entire content supply chain. Everyone from the marketers to the agencies to the media suppliers. And it's really, I don't say too gently, but it's quite exciting to see the transformation that is happening across the industry. So specifically, my focus is on a newer offering at Adobe called Adobe Firefly. And if you haven't played with it, I know you and your team have, but if some of the listeners haven't really encouraged everyone to take it for a spin, I was on it last night with one of our new beta products, Adobe Firefly video.

And so, it's really a text prompt that you can generate video out of. And it really is quite exciting. I don't think I've been that excited since I first got introduced to Photoshop 25 years or so ago. And so, Firefly you can think of in terms of three or four pillars, the first being a product in of itself. Mm-hmm <affirmative>. If you go to Fireflyadobe.com, you can go there and do some text prompting for image generation video generation. We've got some beta products in terms of 3D generation. The next one is an extension of that. It's, we call it Firefly Custom Models. And that allows customers to upload their own private library of content or data. And so, then the system trains off that data in a firewalls environment. So the outputs are very customized. And then thirdly, more recently is Firefly services. And essentially you can think of it as APIs for that last mile in terms of production capabilities.

And now you can create in the hundreds of thousands of variations with these APIs. We have about 25 to 30 in market right now. And those really unlock the possibility of scalability. And in addition to that, the fourth pillar would be something we call content credentials. It's part of a broader industry-wide initiative, which is the CAI, the content authenticity initiative. And so that is comprised with hardware companies, software companies, and media companies, maybe 3000 or so. You can think of it in terms of almost a digital nutritional label where you can see embed; oh, that's great. Embed into the content that's getting generated, how it was trained, how it was augmented, how it was distributed. And we're getting very close to how it performed as well.

Jim (04:07):

We're into some of that. This is a great time to be in this space, and it's great to hear that you've landed here and especially given more of, you know, your history and the agency world and the media world. Right? So, you mentioned Photoshop 25 years ago. That was a great product back then. And with these opportunities come, new concerns, you know, like I saw this quote not too recently, like the guy who invented the airplane also invented the airplane crash. You know, it's kind of morbid, but yeah, it does create some new responsibilities. I think we'll touch on that and that that lineage is kind of important. I didn't know that was part of the product suite. You know, it kind of brings me to this question and I was talking to an analyst the other day about digital native companies or becoming, you know, transforming or giving birth to AI native companies.

So, I think for a long time, you know, whether it's D2C direct to consumer commerce, you know, and these sort of digital native companies, it's kind of sprung up or digital native banking. But that certainly affected the CMOs role, right? For a long time. And I feel we're on, you know, over a decade into this sort of, I don't know what we're up to now, but maybe a Web 2.0 or age of the customer or customer centricity and digital-first has really dominated, I think a lot of CMO priorities way beyond just sort of the brand or brand marketing and evolved performance marketing as well. I'm intrigued by this idea of like an AI-native business or an AI-native company. Do you see kind of that same sort of transformation, and you know, how do you see that affecting, you know, the marketer's role or the CMOs role within a brand?

Chris (05:31):

Yeah, it's a wonderful insight, which I fully agree in many ways. We've been here before with the introduction of newer technologies over the last number of years. And there is a pattern every 10 or so years we get introduced to a new technology. We have the mobile revolution, we have the social revolution, right? And to your point, we are now already deep into the AI revolution. I would also offer, there's stages of AI. We started out with probabilistic or predictive AI, which quickly gave birth to Generative. Many are now identifying that we're even moving beyond Generative or adding on to the Generative area into a GenAI, which is a more proactive use case for AI. And so, with all that said, I think the implications on marketing, and more specifically, CMOs, is much like those previous technologies; they did slightly change the capabilities and the roles of the CMO and the marketer and the creatives and the agencies.

Much like that, we are seeing an evolution of the CMOs. Specifically, my hope and hypothesis is the CMO, the marketer, is going to be the central source of truth for a business. And so what I mean by that is it is now going to house all of the insights because, with Generative AI and Agentic AI, we have the capability to create faster, more insightful and have a faster round trip of in marketplace performance and have that data come back to not only inform marketing disciplines but also the product, the sales and so on. And so, I think we're gonna see an increased value of marketers going forward because of Generative AI and agent AI.

Jim (07:20):

That's a great point. We've been kind of seeing the same thing as more products become connected. Yeah. You know, whether it's kind of an ambient computing or passive computing and a retailer or it's a shop floor, you know, that's collecting all kinds of data, or it's the product itself that is phoning home to the kind of go back to a really old phrase. Like, it's more data about customer behavior, maybe customer intent. And so how we utilize that to personalize, yeah? You know, what consumers are, what end users are looking for. I remember, you know, talking to other CMOs like, boy, you're gonna have a real content problem because if you're gonna personalize, you have to have a personalized message. And I think this solves a lot of those problems. Right. You know, in terms of production.

Chris (07:59):

Yeah. I think you raised it earlier with the introduction of new technologies also introduces new considerations and contemplations. Yeah. What is the optimal tech stack, which I'm heavily focused on? Mm-hmm <affirmative>? What are the implications of that technology on the current state and future state, uh, processes? And then how does that ultimately inform better go to market strategies, inform the product in of itself, the sales cycle, right? And so on. And so really is having a foundational impact on businesses globally. I've been almost evolving my terminology from transformation to reinvention because it really is reinventing Yeah. Businesses from the ground up going forward.

Jim (08:44):

Well, that's what I'm intrigued by this comment I heard the other day, you know, AI native, you know, I think, okay, I get to start thinking about it that way. I think I was sort of intuitively thinking about it, but you're right, it's sort of stopping, starting over again. This content supply chain or this interaction could be AI-driven instead of we're guessing; we're putting teams together, doing research, hoping that it's gonna pan out, you know, in a few months. And now we have an AI-native company can automate those processes right in seconds potentially. So, the organizations you're talking to, the customers you're talking to, when you sort of rank maybe a couple of the problems that they're trying to solve, sort of in business terms, you know, are they trying to be better? Are they trying to be faster? Are they trying to be cheaper? There's more to life than better, faster, cheaper. But what are you hearing is sort of the reasons why this adoption is taking place? Are they there yet? Are they just still experimenting? Or do you think there are specific business problems or marketing problems that you know, you find customers are looking to solve with some of these solutions?

Chris (09:40):

Absolutely. And obviously there's a spectrum across the industries, but I think things have evolved very quickly over the last 24 months. We almost took a page from the digital transformation where we try to apply it to existing frameworks and processes within the first few months a couple years ago that now has expanded and matured into a more fundamental consideration for the ingestion of AI. And so, what we've seen over the last 12 months or so is the need to map out a three to five-year AI vision and then reverse engineer to get to that vision across all aspects and layers of a business. So, I would say that is probably first and foremost for industries currently. The second big vector of consideration is then, as you pointed out, the need for responsible AI. And that has multiple considerations in terms of how do you use, how do you build different models, whether it be LLMs, whether it be your own proprietary data and so on.

And so, there's a continuous need to optimize for responsible AI. In parallel, I've been very heartened to see that the public and private sector are converging in that gap, which has traditionally been a bit broad, is really coming closer together. There's some great work coming out of Utah. The senator there is building an incubator where unusual use cases that kind of color outside of the lines, he is built a incubator where you can bring those challenges and work together to help create the policy. The senator from Texas has offered a safe space or a safe room almost where you can showcase the use case to them and then he can give some guidance within a 30-day grace period. And so, I think those are really wonderful indications of how the public and private sector are working together. The third aspect then is there's an element of, I don't know what I don't know or mm-hmm <affirmative> I can't imagine what I can't imagine. And so, there is a need for continuous upskilling in terms of what are the current capabilities of the technology and then how do I apply that and absorb that within the organization. So, I would say those are primarily the three things that we're seeing universally across use cases and industries.

Jim (12:11):

That's interesting. I mean, I think one of the things that we've done is we're building, and I, we've seen a couple of instances is just a prompt library. We found that that's just the kind of to your, I don't know what I don't know is kind of introducing people to just fundamental prompts. You know, the most simplistic things we found can really very quickly orient someone to what's possible with Generative ad because there's still this concept, like for example, choosing a headline, right? So read this and tell me a good headline. I usually write the headline, and then I write the copy. You know, so just turning that around sometimes is the only trigger people need to kind of get the thoughts going. And we see that a lot. And I love this idea behind responsible sort of, I don't know what to call, I'm gonna call it a responsibility lab, but that's a great idea. Kind of getting ahead of some of these problems. And I'm kind of drawn to, like, the trolley problem, right? People aren't familiar with that. It's common among autonomous vehicles and robotics. How does the machine choose what to do next? Right? Do they save the passengers or save the CELs on the sidewalk or on the railroad tracks? So, sounds like that's a way to kind of get in front of those things, no pun intended, <laugh>, that's interesting. And kind of a novel way of doing things.

Chris (13:15):

Jim, you bring up a number of wonderful points. The first one really I think touches on maybe one of the big concerns across the marketing and creative industry is the impact on the craft of creativity. Yeah. And I, I think you highlighting the prompting is a wonderful example of how the skillset is being transformed from the craft traditionally over to a new form of creativity. I think creatives, there's going to be a renaissance in terms of language because adjectives can really have a profound effect. The adjectives that you use within the prompt can really dictate the outcome. And so, I think artists are going to hopefully refall in love with the art of language on how it can impact creativity going forward.

Jim (14:02):

So, if you're an English major, maybe you need to rethink your dream to write the great American novel. You just need to start drawing out films and imagery with Firefly. So that's a great point. It's not just thinking about the solution, it's thinking about, you know, how you approach the problem differently. Yeah. So, that's great. Any uh, kind of final thought for folks? And I have an idea on something having listened to this. We love to give people one thing that they can go do when they get back to their phone or safely in front of a screen. But what's one piece of advice? If you're in a creative role or you rely on a creative team, what's one piece of advice or one thing they could do today?

Chris (14:36):

I'd really advocate for this creator mindset. In many ways, technology was still somewhat of a heavy lift in terms of adoption. The technology is there now. Really unleash your creativity, your creator mindset and feel free to get out there and start creating new products, new marketing programs. It really frees up ultimately opportunities. So I think we're entering the golden age of creativity and invention going forward.

Jim (15:06):

That's great to hear. I love that sort of democratization of finding kind of the inner creator. So Chris, it's great. I really appreciate you joining us and one of the things I'm gonna go out and spend a little more time with Firefly, so wonderful. Appreciate you joining. Thanks Chris.

Chris (15:19):

Thanks for having me.

Joe (15:21):

You've been listening to What If? So What? a digital strategy podcast from Perficient with Jim Hertzfeld. We want to thank our Perficient colleagues, JD Norman and Rick Bauer for our music. Subscribe to the podcast and don't miss a single episode. You can find this season along with show notes at perficient.com. Thanks for listening.