The Refresh News : March 17 – AI Agents, Meta's Community Notes & TikTok's Future - AdTechGod Pod

Summary

AdTechGod Pod: The Refresh News – March 17, 2025

Episode Title: The Refresh News: March 17 – AI Agents, Meta's Community Notes & TikTok's Future
Host: Kate, Architecture
Release Date: March 17, 2025

Introduction

In this episode of the AdTechGod Pod, host Kate delves into the latest developments shaping the advertising technology landscape. Covering significant announcements from Scope 3, Meta’s introduction of Community Notes, updates on Coca Cola’s media strategy with Publicist, and the evolving situation surrounding TikTok’s future, Kate provides listeners with comprehensive insights and analyses crucial for industry professionals and enthusiasts alike.

Scope 3's Inaugural Landscape Conference

Key Announcements:

Brand Standards Enhancement
Agentic Advertising Platform

Brand Standards Improvement: Scope 3 unveiled enhancements to their brand safety and suitability measures. These improvements are designed to provide more robust protection and alignment with clients' brand needs.

Agentic Advertising Platform: One of the standout announcements was Scope 3’s agentic advertising platform, slated for release in Q2. This platform leverages custom large language models (LLMs) and autonomous AI agents to revolutionize ad campaign management.

Kate's Analysis: Kate highlights that these announcements are in line with CEO Brian O'Kelly’s recent discussions on LinkedIn, emphasizing the strategic direction towards integrating advanced AI technologies.

“Scope 3 is essentially training large language models like Claude and Gemini on their clients' data to create a customized output” (07:30).

Environmental Considerations: While some critics argue that increased AI utilization could lead to higher carbon emissions, Kate points out that advancements in AI efficiency and supply chain optimization might mitigate these concerns.

“AI systems are becoming more efficient to run, and if you pair that with a more efficient supply chain, we could come out on top with brand standards” (05:45).

Functionality and Impact: The agentic advertising platform allows media ecosystem participants to develop, deploy, and monetize their own AI agents, which can handle various campaign tasks autonomously. This includes audience analysis, competitive research, media execution, ad verification, and data science operations.

“AI agents are being presented to us as a standalone product, but eventually they'll just be a feature or a function within a larger workflow” (19:15).

Job Market Implications: Kate anticipates that the rise of AI agents may lead to the elimination of traditional entry-level media roles. Instead, media associates might transition to overseeing and managing teams of AI agents, ensuring human oversight remains integral.

“As more AI agents enter the market, it's very realistic to expect that our traditional entry-level media roles will be eliminated” (16:50).

Meta's Community Notes Feature

Rollout Strategy: Meta is introducing Community Notes as a beta feature before a nationwide launch. Initially, it will utilize X's open-source algorithms for content rating, with potential adjustments by Meta based on performance.

Functionality: Qualified contributors can add context to social media posts across Meta’s platforms by submitting notes supported by trustworthy sources. These notes undergo peer review but require consensus from diverse viewpoints to be published.

“No matter how many contributors agree on a note, it won't be published unless people who normally disagree decide that it provides enough helpful context” (12:40).

Challenges and Concerns: Kate expresses skepticism about the efficacy of Community Notes, citing research that indicates a significant portion of accurate notes remain unseen by users.

“Researchers found that 74% of accurate notes were never shown to users. That seems to be a pretty big flaw in the system” (13:20).

Implications for Content Moderation: While the concept of diverse contextual input is promising, the practical application poses challenges in ensuring comprehensive and unbiased information dissemination.

“Having a system that allows for additional context from people with opposing values and views seems like a good idea in theory” (13:00).

Coca Cola’s Media Strategy with Publicist

Business Transition: Publicist, Coca Cola’s media division in the US and Canada, has solidified its position as a dominant force in media buying, particularly within performance media. This strategic move consolidates Coca Cola’s media operations, while other aspects like content production remain with WPP and select agencies.

“Publicist is one Coca Cola's media business in the US and Canada. Given the way Publicist has become an absolute powerhouse when it comes to media buying, particularly performance media, this is not that surprising” (14:30).

Strategic Focus: The emphasis on media buying underscores Coca Cola’s commitment to optimizing their advertising expenditures and enhancing campaign performance through specialized expertise.

TikTok’s Future and Project Texas 2.0

Current Status: TikTok faces an impending deadline of April 5 regarding its ownership and operational structure within the US. Oracle emerges as the primary contender to become TikTok’s sole owner under Project Texas 2.0.

Regulatory Challenges: Despite Oracle’s potential acquisition, existing laws mandate that TikTok’s operations, data, and algorithms be controlled by a US-based entity. Additionally, approval from Chinese authorities poses another layer of complexity.

“That law requires that TikTok and any other apps like it have all their operations, data algorithms owned and operated by a US entity” (17:10).

Speculative Outlook: Kate speculates that multiple influential stakeholders may still influence the final outcome, suggesting that the resolution may deviate from current expectations.

“My hypothesis is that there are still several players who have some pretty powerful cards to play. Regardless, I think we're all going to be a little bit surprised at the final outcome” (18:00).

Additional Highlights

NBC’s Mega Contract with the International Olympic Committee: Kate briefly mentions NBC securing a significant contract with the International Olympic Committee for the Summer 2024 Olympic Games. This deal serves as a valuable case study for future media strategy and execution.

“The Summer 2024 Olympic Games are one of my favorite case studies to refer back to. When it comes to thinking about what the future of media strategy and execution will look like, that's all we have time for today” (19:50).

Conclusion

In this episode, Kate provided a thorough examination of pivotal trends and movements within the adtech industry. From Scope 3’s innovative AI-driven solutions and Meta’s experimental Community Notes to Coca Cola’s strategic media consolidation and the unresolved future of TikTok, the discussions encapsulate the dynamic and rapidly evolving nature of advertising technology. Kate’s insightful commentary equips listeners with a nuanced understanding of these developments, preparing them to navigate and leverage these changes effectively in their own endeavors.

For more information or to become a guest on the AdTechGod Pod, visit AdTechGod Pod.

Transcript

McAfee Advertiser (0:00)

Your data is like gold to hackers. They'll sell it to the highest bidder. Are you protected? McAfee helps shield you blocking suspicious texts, malicious emails and fraudulent websites. McAfee Secure VPN lets you browse safely and its AI powered tech scam detector spots threats instantly. You'll also get up to $2 million of award winning antivirus and identity theft protection, all for just $39.99 for your first year. Visit McAfee.com, cancel anytime terms apply.

Unknown (0:29)

Foreign.

Kate (0:34)

I'm not mad, I'm just disappointed about Apple Intelligence being delayed. Now we have to wait for a Smarter Siri Daylight savings time took me an entire week to recover from. And the worst one of all, I'm missing marketecture Live. I'm Kate with Architecture and this is the Refresh, your weekly download on Want One down in Advertising today is Monday, March 17th and on this week's episode we're talking about Scope 3's big announcements meta rolling out their Community Notes feature podcast who was this winnings from the Coca Cola business? And finally, we'll do a quick update on where things stand with TikTok. So let's get into it. Scope 3 dropped two big product announcements at their inaugural landscape conference last week. The first was brand standards, an improvement upon brand safety and suitability, and the second was an agentic advertising platform. If you've been following CEO Brian O'Kelly on LinkedIn the past couple of months, neither of these announcements should come as a surprise, least of which the agentic advertising platform Scope 3 is looking to custom large language models which have the ability to reason, and AI agents which have the ability to operate autonomously to fuel their next act. Some have pushed back on these announcements, saying more AI equals more compute equals more carbon emissions. On the flip side, AI systems are becoming more efficient to run, and if you pair that with a more efficient supply chain, we could come on on top with brand standards. Scope three is essentially training large language models like Claude and Gemini on their clients data to create a customized output. These algorithms can be trained based on a brand's needs, their audiences, business objectives, historical campaign performance, and their own unique brand safety and suitability standards. While brand safety and suitability is of course the desired outcome, performance is too. One of the benefits of large language models is that they're essentially in an ongoing, ongoing learning process, using new inputs to make more informed decisions. This ideally creates a more relevant ad experience that reduces the number of touch points needed before a consumer takes a desired action. Alongside this, Scope three Continues to steer towards reducing carbon emissions by removing inefficiencies or undesirable inventory from the supply chain before a bid request is ever even submitted. Alright, let's talk about this Agentic advertising platform. It's not scheduled for release until later in Q2, but it would allow those within the media ecosystem to build, create and even sell their own AI agents. These agents can be set up to complete tasks across the campaign life cycle. Audience analysis, competitive research, media execution, ad checking, and data science tasks like performance measurement and analysis. I actually had a really hard time sorting through all the press surrounding these announcements. A lot of the terminology is being interchanged and what was being referred to as agentic was actually just a custom bidding algorithm built on the back of an LLM. While these systems have a lot in common and they're intertwined, they are not the same thing. When we talk about AI agents, we're talking about something that's able to be assigned an end goal and then it will navigate itself to that goal. It's learning and reasoning and establishing the steps it needs to take, as well as communicating with other agents or you to complete the task that it's been given. It's also able to remember, store, analyze and retrieve previous information as more AI products are released. It's important for us to understand what you mean when you're saying AI. Are you talking about a machine learning algorithm, a custom bidding algorithm built on an LLM, or a true AI agent? Because those things are different. So will AI agents impact our job roles as advertisers and marketers? As more AI agents enter the market, it's very realistic to expect that our traditional entry level media roles will be eliminated. Media associates and operations specialists could find themselves managing teams of AI agents serving as a human input and checkpoint that's still needed. Right now. AI agents are being presented to us as a standalone product, but eventually they'll just be a feature or a function within a larger workflow. In order to maximize AI's value to our organizations, we need to start taking a systems based approach to how we think about it. Generative AI in its current form is imperfect. It's resource intensive and technically complex. But someone has to start making moves towards implementing it into our media processes to ultimately transform the system. Brian O'Kelly and team seem to be taking a really solid first step towards this. All right, let's talk about Meta's community notes. It'll be rolled out as a beta and the Meta will launch it nationally once they're confident that it's performing as it should. Community Notes will initially launch using X's open source algorithms as its ratings basis. However, Mark Zuckerberg has said that they may diverge from these algorithms and adjust them on an as needed basis. So how does it work? Pre selected contributors that have met Meta's qualifications will be able to submit additional context to various social posts across Meta's platforms. They also have to submit a URL, I'm assuming from a trustworthy source that backs up what they're saying in their note. Notes are then rated by other contributors. But just because contributors are rating other people's notes doesn't mean that it's going to be majority rules. Meta clarified by saying no matter how many contributors agree on a note, it won't be published unless people who normally disagree decide that it provides enough helpful context. Humans are inherently flawed and biased, but so are algorithms, and that's going to filter into any system that's in place. Having a system that allows for additional context from people with opposing values and views seems like a good idea in theory. I do worry about the efficacy of this system. Anecdotally, I haven't seen any community notes on X, and researchers found that 74% of accurate notes were never shown to users. That seems to be a pretty big flaw in the system. Alright Publicist and Coca Cola the TLDR is that Publicist is one Coca Cola's media business in the US and Canada. Given the way Publicist has become an absolute powerhouse when it comes to media buying, particularly performance media, this is not that surprising. The rest of the business, including content production, will stay with WPP and a smattering of other agencies. TikTok's April 5 deadline is ticking closer. It sounds like there are four contenders left with the rows, with Oracle being the primary standout. The deal is being billed as a Project Texas 2.0, but even if Oracle wins the bid to be TikTok's sole owner, that's unlikely to be enough to satisfy the law that's currently in place. That law requires that TikTok and any other apps like it have all their operations, data algorithms owned and operated by a US entity. We also have to consider the question of whether China will approve any deal, no matter who it's with. While everyone's maintaining their poker faces, my hypothesis is that there are still several players who have some pretty powerful cards to play. Regardless, I think we're all going to be a little bit surprised at the final outcome. Before we go, I want to give a quick shout out to our friends at NBC who locked in a mega contract with the International Olympic committee. The Summer 2024 Olympic Games are one of my favorite case studies to refer back to. When it comes to thinking about what the future of media strategy and execution will look like, that's all we have time for today. Thanks for listening in and we'll catch you next week.