Allison Schiff (3:45)

That is one of my favorite answers to the tell me something about you that not a lot of other people already know question. Like, hands down, it's an, you know.

Paul Longo (3:54)

It'S an interesting dynamic. And I do, you know, I think about even like the connection points between, you know, screenwriting as a hobby and my role. And I think, you know, one of the big ones is just, you know, it's a very daunting task to write a screenplay and just chunking it out kind of like, you know, bit by bit, day by day. And you know, and you know, especially right now in the world of digital and AI, like a lot of what we're doing is super daunting. So I think it's just like kind of like that incrementality of getting that like 1% better every day, like atomic habits. But also too, I just wrote my first screenplay AI assisted. So that was fun. I was like, I'm going to do this and see how this works and how I could use it. And I wrote it probably half the time and it was a really interesting learning experience. I did learn that, at least currently today, that AI assistants are not great at writing screenplays, but they're super helpful in a lot of ways. So I did find, as an example, like, I love writing dialogue. To me, that's like the fun part of writing a script where I used AI was to write a lot of the scene descriptions, especially like blocking and tackling scenes where it's just like boring part of the script, like setting up the scene in a restaurant or that I was just Like, I saved a ton of time, you know, and there were even some, like, scenarios, like dialogue scenarios where I was just like, you know what, maybe I'm not hitting it. You know, could, could AI help me punch it up a bit or give me a couple of alternatives? And that was helpful too. So that was kind of an interesting, kind of bringing my worlds together moment.

Allison Schiff (5:29)

Thank you for saving me for myself because I almost made this podcast into just Paul Longo's advice on screenwriting AI.

Paul Longo (5:38)

I could, I could talk about screenwriting for a while, definitely. Let's move into other topics.

Allison Schiff (5:43)

Well, you've been at Microsoft for around 10 years. You've worn a lot of different hats there. Head of Agency Development for North America because you came from the agency world. You were senior Director of Global Agency Partnerships, GM and global Head of Retail Media Sales and now GM for AI in ads, which is such an interesting title. And you've had that role for about a year. And just the phrasing AI in ads really strikes me because it makes me wonder what that means. Does it mean using AI to make advertising better and cheaper to scale ad production? Does it mean introducing more AI into Microsoft's ad tech? Does it mean monetizing a chatbot or an AI powered search experience? Does it mean all of that? Like, basically, what is your job?

Paul Longo (6:31)

It's the title that brings up a thousand questions. I know and like, so many different conversations that I've had. And like, yeah, even going back to like my career arc, you know, I've been fortunate enough to have, you know, great mentors and leaders, you know, in my personal, professional life and, you know, you know, having experiences on both the agency side. I was client direct at l' Oreal for, for a number of years too. You know, one of the best pieces of advice that I got from my mentors was just, you know, take smart risks and do the jobs that nobody wants to do. And I've always found that those smart risks generally are coming at change waves, at the front of a change wave, where it's really messy and it's really scary and it's at that place where people don't want to do it. And I've found taking that leap has been just a great, I think, propeller for my career, both on the agency side and then at Microsoft and at Microsoft. I am a boomerang. So I was here, I left, went back agency side and came back. And over the last 10 years, I've had the opportunity to take some of those smart risks, to get involved in leading different elements of our partner business and transform some of those partnerships to take that leap into retail media. Certainly on the bleeding edge with some of the work we're doing at Microsoft and then now in AI about a year ago, obviously there's lots of people at Microsoft within AI in their title, in our ads business, AI certainly core to everything that we're doing, but we didn't have dedicated teams that were just fully working on AI. So unpacking what AI and ads means, there's one component that is around creating advertising experiences in conversational environments. So for us that means Copilot. So we launched Copilot as bing chat in 2023 and then we've been evolving that over the last two years and now it's at a place where the experience has built out. It's more companion, like there's memory. So CoPilot, as an AI assistant, knows users at a more personal level and we're figuring out, well, how to advertise, how do ads fit into that experience? Right, you know, where trust is paramount, where you don't want a trusted companion to be, you know, blurting out ads to you, you know, when they're not super relevant. So we've come up with a set of principles there around, you know, hey, advertising needs to be only served in highly relevant commercial conversations. What it is, it needs to be, you know, very distinct and clearly labeled and. Yeah, and, and sponsored ads will never influence organic results. So, so very, you know, church and state in terms of, you know, how they're, you know, how they're displayed. So we're continuing to push the limits and test new formats. We currently have ads live there now. We've, we've, we've learned a lot. We've actually seen that when you serve ads in these conversational environments and combine that with a search. So when we look at Bing searches and conversion rates, performance alone versus Bing plus ads and copilot, we see that consumer journeys are happening 30% faster, which I think makes a ton of sense. Our cmo, Yousef Mehdi, kind of uses analogy of the, you know, with, with, you know, ads in these conversational environments are moving from a funnel to a tunnel, you know, and as you're having a conversation with an AI, you're getting, you know, you're getting information super quickly, right? You're asking four, five, six different questions. And again, as you get those answers, you're moving through the funnel in real time. We just launched some, a product called a brand showroom, which is kind of inviting brands to come into that conversational environment. So when somebody raises their hand and says, hey, I want to learn more about Mercedes and that's who, who is our first pilot customer, we're able to bring them in. So Mercedes was actually able to come in with their images, their, their copy, their tone, and to be part of that conversation in an ad experience. But it was an ad experience where somebody said, I want to go deeper with Mercedes. And we're able to work through that whole conversational experience up to where that handoff point is when we want to bring people outside of the experience and over to Mercedes. So I talked a lot just about the role, but that's just kind of one side. We can unpack that a little bit more. And then the other side of my role of where we say AI and ads is more focused on advertisers and empowering advertisers who are using our tools and infusing our tools with AI features, functionality. So a couple of examples there. We have a version of Copilot that is incorporated into our Microsoft advertising platform that essentially functions as a virtual account manager that helps to plan, optimize campaigns. We just at. Can we launch some case studies, one with Samsung Germany and Starcom, which showed that Samsung Germany was able to reduce their campaign optimization time by 30 to 40%. So that's just, I think, just the tip of the, the iceberg in terms of, you know, how AI could work as an enabler for, for advertisers. Where we also launched an ad studio and platform which enables, you know, gen AI copy image and zoom video creation. So we're going to continue to build out those suite of tools. So I'd say, you know, a. And ads, AI and ads. One side, ads in these conversational environments and then the other side, you know, dedicated towards advertisers.

Allison Schiff (12:34)

So it feels like ads in AI and AI and ads, it's both.

Paul Longo (12:38)

It's both, it's both and it. And you know, so, so looking for a tattoo to encompass the full thing. But. But yeah, that's, that's really. When you unpack it, you used two.

Allison Schiff (12:51)

Different words to describe copilot, which I think is really interesting. At one point you said assistant and then at another point you said companion and to mind those are different things. So I don't have a good question about that other than what is it? Is it both? Is it a helper? Is it a friend?

Paul Longo (13:13)

Yeah, I think I'm glad that you flagged that. You know, I'd say like, you know, I think it's. I Think it's still probably more assistant like and moving to that more companion piece. Right. I think the companion piece is when it really gets to know you better. Right. I think to me, I think an assistant could just, is something that could just execute just a quick task, almost like a one off transactional thing. I think where it becomes more of a companion is when it is exhibiting a deeper level of iq, but also EQ and it, and it knows your likes and dislikes. So that's where I see kind of the, the transition from assistant to, to companion.

Allison Schiff (13:53)

Not to take us too far away from ads, but have you been reading about how some borderline personalities or narcissistic personalities are being extra affected by AI because an AI will tell them they're a genius and they already know that they're a genius. And so it's making people more narcissistic because an LLM will say totally, you're right, like you're a genius. Like you got it.

Paul Longo (14:25)

I haven't read that actually, but yeah, but I think that to me that is like, I think you'd want an experience that's a little bit more nuanced and you know, has a higher degree of EQ and not just validating but having more of a dialogue I think would be more of an ideal experience.

Allison Schiff (14:42)

Let's talk a little bit about ad quality. Like AI generated ad creative and I guess what you might call like AI augmented ad creative. So it's not created from scratch, just tweaked perhaps based on AI suggestions. How's the quality of what's being produced so far that is coming out of copilot and what still needs improvement, like I, I'm assuming you can generate realistic looking human hands, for example, at this point.

Paul Longo (15:16)

Yeah, it's, it's a, you know, first I would just say is, you know, these tools are all about, you know, human enablement and enabling, you know, creativity versus replacing. Right. You know, so like this is almost like, I would look at it as like, almost like, you know, more, you know, more paint and more tools, you know, for, you know, for the painter. Right. And so I think as a creative, as someone who's creating advertising campaigns, like number one, I think AI could be a great equalizer. Right. Where now all of a sudden if you're like a small medium sized business, maybe without an agency, you're able to create something that is a bit more compelling and that could go toe to toe with maybe some of the bigger advertisers. I think the, you know, ultimately the quality I think is still evolving. You know, like I said, you know, I think right now, you know, it could help you from an efficiency standpoint to generate additional images. Right. So I think like the Strat, you know, developing the strategy, developing the creative, I think is still very much human centric. Like the way I look at these tools are. Well then how do the, how do these tools then amplify the initial brief, right? How do they amplify the initial creative concept and maybe go from 10 versions to 10,000 versions and you know, provide the ability to do that? That wasn't there in the past. So I think it's, I still think it's an enabler, not something that's going to replace. I've seen some attempts online, certainly with certain brands where they've done a fully AI created campaign. I personally didn't think the creative was, was there yet. Right. I think it's pretty obvious that it's an AI campaign. So I think that that's, you know, value and effort great that you know, that experimentation is happening. But again, I think it's, you know, how can it be more of that enabler versus like the thing.

Allison Schiff (17:16)

There are some ads though that are AI generated that are pretty wildly amazing. Like that Kelshi ad during, I think it was the NBA games, right, the finals, they used Google's VO3 text to video generator and they made this ad for like 2000 bucks in under 72 hours. And there was a lot of buzz around it because it looked pretty damn good. It wasn't like that Toys R Us ad from whenever it was last year or the Coca Cola holidays are coming ads, which the uncanny valley there was like pretty uncanny. It's getting remarkable.

Paul Longo (17:55)

It is, like I said, it is improving. And like you mentioned the Coca Cola example, which I think Coca Cola took some heat. But I, I give them a lot of points for, you know, for trying. Right. And for some being, for being out there and, and for really, you know, you know, making it, you know, this, we've, we've been saying that this is like, you know, AI is a year going, moving from promise to practice and we need the Coca Cola is out there like practicing and testing and you know, so that we can really evolve this. So, but yeah, it's, it's moving quickly. Like I said, we, we're, we're moving to have video incorporated to our platform soon. So I think that's the next frontier which is really going to, I think push the envelope for creative capabilities with AI.

Allison Schiff (18:41)

So one more question about humans in the Loop before we take a quick break, you have folks like the CEO of Shopify saying like, no one can make a new hire if they can't prove the new job isn't doable using AI. And then you have the CEO of Duolingo saying they'd use AI to replace contractor jobs or like some contractor jobs and then limit new hires if the automation that they introduced worked well. And there was this backlash to that. And then the Duolingo guy clarified, he's like, no, no, like, AI won't replace employees. It's just to help them work better. And that does sometimes sound like patter that you get from executives. Like, we're not looking to replace you. You use this. We just want you to be more efficient. But then maybe you're replaceable. It's kind of like a little bit of a snake eating its tail in terms of logic. And I'm just putting this out there as a devil's advocate question. But obviously AI is just a tool and painters need paint and they need more colors. But it's also more than just a tool. Like it is a truly transformative technology. And I don't know that if someone says that it's not going to end up eliminating some jobs. That doesn't feel like honest really to me.

Paul Longo (19:57)

Yeah, I think like the way I look at this is, and if you think about like this technology wave of AI and just how quickly it's been adopted, but also how early we still are, right. You know, in the, in this, you know, change curve. So you think like ChatGPT and the mainstreaming of AI was November 2022. Right. You know, so it's been out there for two and a half years. So I, I do think, you know, if you think about some of the other technology waves and the Internet and mobile, there was also a lot of fear, especially at the beginning. Right. And you know, and you know, when you think back to like even with retail and with, with the Internet, like, hey, there was a lot of fear like that jobs are going to go away in retail. In all actuality, there were actually 7x jobs created in digital retail over 10, over a 10 year period versus jobs that were lost. And when you think about new jobs like web designers and analysts and so I think we're at a place right now where, yes, I mean, I do think it's inauthentic to say, like, hey, nothing's going to change. I mean, I think there are certain jobs that will evolve because of AI, but I think there's going to be New jobs that are created as well. Even just top of mind, like right now there's I think tremendous opportunity from like a consulting standpoint in brands moving from SEO to geo. Right. And how are, you know, how are brands, you know, guided into that next frontier of, of optimization in an agentic world? Also, how companies are incorporating AI across their full workload stack, you know, not just using it as an assistant, but how is it, you know, you know, incorporated into like we just released a work trends in text talking about AI frontier firms that are taking that next step in the new, you know, so you know, what, what's the, what are the companies and consultants that are going to help, you know, move that forward? So I think we haven't seen how it's going to shake out yet. But, but yeah, I think we have to have a conversation around both sides of that coin. And you know, I think this will become clearer over the next, you know, you know, two, three, four years.

Allison Schiff (22:02)

A metaphor just popped into my head, but I haven't thought about it enough. So maybe it doesn't perfectly work, but I guess you won't survive if you're a dinosaur. But it'll be someone's job to refine the fossil fuels. I don't know.

Paul Longo (22:17)

It's true. Like even like, hey, like all the cloud storage that's needed. Right. It's probably good to be in the H vac business right now. Right. You know, so I think that there's pockets of places where jobs will certainly be created.

Allison Schiff (22:31)

So we're going to take a quick break and when we're back, we're going to talk a lot more about that agentic world that you briefly just referenced. So stick with us. All right, we're back and we spent a bunch of time talking about creative, so AI in ads. But I want to talk a little bit more about being the other side of your job, which involves introducing more AI powered elements into Microsoft's ad tech. And one very high profile example of that is the announcement back in May that you're going to shut down Microsoft Invest, which was formerly the Xander DSP. And that's going to happen by March 2026. And the DSP will be replaced by a chatbot style ad buying product powered by Copilot, because the current DSP model no longer aligns with Microsoft's vision for the future of advertising. So what is Microsoft's vision for the future of advertising?

Paul Longo (23:48)

Yeah, I'm glad you brought this up and I know we talked at a high level about this at Programmatic I.O. in Las Vegas, I think when that announcement was made and just going back, we have two buying platforms of Microsoft advertising platform, which buys into Bing and our audience ads and Microsoft O and O and then we obviously have the Xander Invest platform. You know, a lot of the focus was on what we're moving away from versus what we, we're moving towards. And you know, everything that I talked about earlier around these new conversational ads and then even like when you think a little bit further into the agentic future, right. So, you know, we believe that just like all brands have a brand website today, they're going to have their own agents and consumers will have the choice of whether they want to interact with traditional way and the human Internet with a website or if they want to interact with a brand's agent instead in more of a UI less environment or further down the road to be able to invoke their digital assistant to have that conversation with the brand agent on their behalf. Right. So as we move forward towards that more agentic Internet and it'll get there in phases, you know, and we feel like it'll go from human Internet to what we're calling agent assisted Internet, which is this kind of combination of web and agent, to more of that fully functional agent, agentic agent to agent. You know, we, we, we wanted to bet on a buying platform that will get us there and that was our Microsoft advertising platform. So that was really where we placed our bet in, in moving forward. And again, you know, we're still very much involved and you know, are prioritizing, you know, open Internet and you know, we're, we're still in the game via monetize, on the supply side via curate. So I think that was a little bit of a misnomer as well that we've left the programmatic business and that's very far from the truth. So yeah, we really made a decision on consolidating our efforts behind one buying platform, which is the platform that we feel like will best set us up for that agentic future.

Allison Schiff (26:13)

What about an agentic future for the ssp? Why not the ssp.

Paul Longo (26:19)

You know, and, and that's something that I think, you know, we're thinking through what that looks like. You know, what did eight paid, you know, if you think about like supply and demand in the future and like hey, the, the new, you know, you'll have demand coming in through, you know, AI assistance, like Copilot, ChatGPT and you know, this agentic supply. So I think there's like a lot of parallels so, yeah, that's something that we're definitely thinking through, how that, how that all could potentially work together as well.

Allison Schiff (26:44)

What is the end state here? Will it just be AI agents autonomously conducting media buying conversations with other AI agents? And then media planners are just sipping their coffee and making sure that it's all going okay? And what are the implications for buyers that are managing campaigns today? What does their day look like in a year from now or two years from now?

Paul Longo (27:11)

Well, I think a year from now and you know, again, I think it's, it's more tools at, you know, at, at their disposal, you know, and what I hear, you know, from a lot of media buyers is that they're overwhelmed, right? They're overwhelmed with the amount of platforms that they have to log into. That is, it's kind of unwieldy. You know, this isn't not specific to the video world, but we, we recently published a work trends index, and from a Microsoft standpoint, We've got Copilot M360. We got Microsoft M365, which is the biggest work productivity suite, plus LinkedIn Plus. We did a bunch of surveys and it's published, highly recommend people to check it out. 80% of the people surveyed said that they don't have enough time or energy to do the work that they're doing today. So I think even like that near term of just like agencies, buyers are super overwhelmed. So I think even to me that near term, one or two years is like actually bringing a lot of relief into the system of the media buying world.

Allison Schiff (28:19)

Kind of like a Valve. I do see it a little bit in my own experience of playing around with, forgive me, I use perplexity more than I use Copilot. I'm sorry, I don't know, for whatever reason, you know, what it is, it's kind of like the same debate that's happening now with defaults and traditional search because I used it first and now I just use it all the time. So I standardized on it because it was the first one I was introduced to.

Paul Longo (28:48)

Hey, I'm glad you met. No shade thrown at all. Yeah, there's something to the ecosystem play, right? And that's something I think from a Microsoft standpoint. We have our Copilot family of apps and we also have the Microsoft ecosystem. So there's 1.6 billion people on Windows and that's something where we're incorporating Copilot as an assistant threaded throughout that ecosystem. And that's a big part of our strategy for growing our AI. Presence is to. Your point is to find people. And we're still super early, right? So there's so many people that haven't even engaged with an AI system for the first time. So it's like really get this in the hands for people, like create a really valuable trusted experience. And I think that, you know, and then continue to build that trust over time.

Allison Schiff (29:36)

How are you getting people to use those features, like prompting them using the Word prompt on, on purpose to get them to use those features that are just popping up in the services that they're used to? Because we use Microsoft Outlook as our email at Access Intelligence, which is the company that owns Ad Exchanger. And I did notice a little Copilot button pop up very recently in the Microsoft app. I haven't pressed it yet, but I will, I promise I will. Paul. But like, how do you get people to, to do it?

Paul Longo (30:08)

I think that, I think a big piece is just what you, you know, what you called out was just, is making it more just like organic to the experience. So like, yeah, like when you go to write an email now, you see this little co pilot prompt that enables you to, you know, draft the initial email and it's right there now it's front and center. And I know that that's driven a lot, a lot of usage and I think it's, you know, getting, getting that brand out there. Like we just integrated co pilot now on a, on a, on a limited basis into Edge and that's, I've been playing around with that and that's really cool. So from, from.

Allison Schiff (30:47)

Take some water. Yeah, yeah, I've been making a talk for like 40 minutes straight.

Paul Longo (30:51)

Oh, that frog really came there. And, and it's kind of like this like, companion that, that was helping me, you know, soar through content. And I actually found myself like, engaging in more content with Copilot as, as an assistant and, and you know, I, I saw the announcement obviously, but now, you know, it's pretty prominently featured, the Copilot button, that people could open up and have that kind of dual experience. I think that, you know, we'll promote it more too and, you know, have some creative ways to make people aware of those features. But I think, like, ultimately, I think what drives the most usage is just that incorporating, you know, Copilot in a really organic way that is authentic to the, you know, our Microsoft experiences.

Allison Schiff (31:38)

Other than Copilot, because I'm assuming that you use Copilot quite a lot. What do you use? What's cool to you in your own life? Work and personal. I also assume that you're playing around with everything because you need to compare and contrast. But what are you actually using in your own life?

Paul Longo (31:58)

Yeah, I mean I do play around with everything to do a little compare and contrast. I mean, I'll share a couple of examples of just how I'm using AI. I shared the screenplay example like earlier, you know, where, you know, I started this job a year ago. We're investing in the team. So now I'm onboarding a team which, which is, which is exciting. And you know, with Copilot M365, I created an onboarding assistant. So I had all these documents and pitch decks and, and, and, and product briefs and I created an agent that was grounded in all of these documents and materials and just such an easier way. And instead of like my team pinging me with like a bunch of one off questions, they're pinging the agent and you know, and they were kind of testing it out and like they were giving me feedback on like well, hey, you didn't include this. Or like hey, it'll be great. And then I just added it in to the grounding. And so like that was, that's just like a, you know, real life kind of productivity example. I think the other one is too. I'm using, I find that I'm using voice a little bit more, you know, especially on mobile. I am not the, the most handy person in the world. And even just things on just like, hey, like what like grill part do I need right? Or like, you know, where I'm outside, you know, where's the closest place to get a new propane tank? So just like little things like that have been helpful, you know, just to get like quick hit answers. When I was going to can, it was, I was just like, hey, what's like show me an image of like the charger, you know, adapter. Because I was like, I had a bunch and I always forget like which is the right one. And I was like, oh, I'm just super easy. Yeah. So I think, I think voice, especially as, you know, assistants get better, smarter with memory. Like the voice piece, I think I'm finding it to be, you know, you know, more on point. You know, just from like a relevancy.

Allison Schiff (34:02)

Standpoint, I was hugely remiss. I did not ask you what your AI screenplay is about.

Paul Longo (34:10)

It is, I do. It is a romantic comedy. I feel like, you know.

Allison Schiff (34:14)

Yeah, yeah.

Paul Longo (34:15)

I feel like romantic comedies are on the way for, for a comeback. I think, I think we, I think we need more. There's, you know, like, I know just Naked Gun movies. I think there's like this like 90s kind of nostalgic combat happening with some films. So I hope that, that, that continues.

Allison Schiff (34:31)

What's funny is I'm a horror movie fan and my boyfriend loves romantic comedies.

Paul Longo (34:38)

Right. I love, I love horror films too. Presence was phenomenal. I don't know if you've seen that.

Allison Schiff (34:43)

I have not seen it. This is not really a horror film, but I guess because it's almost too close to a potential reality. But speaking of companions, have you heard of or seen the movie Companion?

Paul Longo (34:58)

No, I am not.

Allison Schiff (34:59)

Oh, I recommend it highly. I can't say anything about it or it'll give it away, but it has an AI robot mirror. It's very black mirror. Ish. And it's beautifully done. And I would go watch that when you come in. I think it's streaming already.

Paul Longo (35:17)

Good movie suggestions here.

Allison Schiff (35:20)

So I want to talk about pmax just for a second. Not Google's pmax, because Microsoft also has a tool for automated ad campaign management called Performance Max, which I think is a very cheeky name since you know, it already exists. But how does Performance Max work with Copilot? Is there interoperability there?

Paul Longo (35:44)

Yeah, so actually, so the ads from the Microsoft advertising platform flow into Copilot, so they render in a different way that is more, you know, appropriate to, you know, to, to the format with something called Ad Voice that gives like a little introduction as to why the ad is there. And. Performance Max, yeah, cheeky name. But sometimes, you know, we have to, we have to pick our battles in terms of just like, hey, where do you kind of piggyback on to name recognition versus creating something new and where do we lead and create something new like our showroom ads? And in this case, I think it was, you know, it was the right call to just have a similar name with a similar prospect. And really it optimizes your campaign and your performance. Across all the Microsoft canvases of which copilot is one, what we're seeing is again, we're seeing 73% increase in click through rates on Copilot shorter customer journeys. So that inclusion into Performance Max is a performance enhancer that we're seeing on.

Allison Schiff (36:51)

Our side and sticking with interoperability, but broadening the question, how do you foster interoperability between your ad platform and other just major ecosystems like Google Ads or Meta Ads, given that everyone's basically a walled garden, but advertisers obviously spend a lot of money in all of these places and they need some kind of unified view. It's what they've been asking for from like your dot.

Paul Longo (37:17)

Yeah, we're a big, you know, we're still a big proponent of the, you know, of the open Internet, you know, just through our, you know, you know, monetize and curate platforms. I think in the future too, like, you know, you see, you know, these various, you know, AI assistants and I think there needs to be some sort of interoperability in an agentic future. Right. And, you know, especially if you are a brand and you're building an agent on a specific platform. Right. You know, you're going to want those, those brands, you know, you're going to want those brand agents to be interoperable with all of the AI assistants. So I think that's very much, you know, how we're thinking about it as things, as things go forward. And I think that from a brand standpoint, a consumer standpoint, that interoperability is going to be important.

Allison Schiff (38:12)

You can't have an agentic future, which is conversational without the ability for agents to have conversations with each other.

Paul Longo (38:19)

I would agree.

Allison Schiff (38:21)

So, last question, and I'm actually going to delegate this one to AI. So when you spoke in Las Vegas at our event back in May, I asked Copilot to give me a question to ask you on stage. And it told me to ask you how you see AI, AI reshaping the workforce in advertising. But we talked about that a bunch already, so I asked Copilot for another one. I wrote. What's an interesting question I can ask Paul Longo on the Ad Exchanger Talks podcast. I literally just asked that and I'm going to read you what it said verbatim. I kid you not. This is amazing. Great choice. Paul Longo is a seasoned voice in advertising and the Ad Exchanger Talks podcast is known for its daily deep dives into ad tech and marketing innovation. So compliments for you and for me. If you're preparing questions for an episode featuring Paul, here is a compelling and thought provoking idea to spark a rich conversation. If you could redesign the digital advertising ecosystem from scratch, what would you change first?

Paul Longo (39:29)

Holy wow. Co founder, I think like, you know, I'll go back to like the work, you know, the work trends index, right. And this notion of AI frontier firms that are really incorporating AI into everything that, you know, into, you know, every kind of day to day, you know, operation of their company. And we have a group AI for Work and that partnered with recently Jared Spartero, our CMO of AI for work. I've just posted about this recently partnered with NYU Stern School of Business and they gave out a business case to create these AI frontier firms where they put AI at the center. And they had minimal experience in coding or creating in depth financial models or with marketing. And by the end of this exercise, after a couple months, all of a sudden they were seasoned CMOs and CFOs. And so by putting AI at the core, they were able to kind of reshape their roles, responsibility, and this kind of notion of like intelligence on tap. So along TF for saying, you know, if I, if I did have to kind of recreate, I would just kind of come at it with the lens of if you were starting from scratch today and putting AI at the core, how would we reshape the ecosystem? And I think that's, you know, I'll leave us with that big know question to unpack maybe in the future. But, but I think that would be the biggest change was to, to put AI at the core.

Allison Schiff (41:07)

Well, I, I learned a lot over the past 40 minutes, but my main takeaway is that I need to go watch Meddling mom tonight and I'm going to track it down. Can't wait.

Paul Longo (41:18)

I don't know if you can find it in full transparency. I don't know if, if it's got too many stars on IMDb but give it a, give it a watch.

Allison Schiff (41:27)

I, I'm going to be like really excited when I see your name come up on the screen at the beginning. Well, thanks, Paul. This was a great one.

Paul Longo (41:35)

Yeah. Cheers, Allison. Thank you.