Now the cathedral is invisible. It’s not a machine you marvel at; it’s the infrastructure running quietly in your pocket. It’s compute; massive, industrial-scale processing power. Whoever has it can train, steer, and deploy the models. And these aren’t just “cool demos.” They’re prototypes of tomorrow’s operating system for society. Forget oil or steel. The new geopolitics is GPUs.

Yes, AI is transforming advertising. Yes, it’s disrupting jobs, ethics, budgets, and probably sleep schedules. But think of the upside: higher efficiency, lower costs, and campaigns fine-tuned by omniscient math wizards with no sense of irony.

In the end, the industry will adapt, evolve, and continue to sell us things we didn’t know we needed, now faster and more creepily accurate than ever. As long as we remember who’s really in charge. Definitely still the humans, probably.

By embracing the changes brought on by AI, the advertising industry can not only survive but thrive in an increasingly digital world.

Hey gang. I am operating under the assumption that, since I first started this venture, people are experimenting more playing with the tools and things they consider the AI fun stuff. I would guess that this would hold true across all professions from project managers up to CEOs. And if I'm wrong, you guys should be doing it anyway.

What I'm finding is, as AI persists in permeating the Advertising space (and our lives in general), it's becoming invisible. I wrote about this a few months back. It ingrains itself into the tools, technology, furniture, clothing, what have you - that we’re already using. I liken it to Compuserve or AOL, or just plain browser-based connections evolving into the Internet of Things (IoT). And now the fridge has Internet service.

What this means is fewer earth shattering headlines. Less drama. But no less work to do. So I will continue to gather things that catch my eye that others might find compelling useful, helpful or just plain interesting. I'll leave it at that. -pb

Klarna​ - It’s the Final Countdown

Regular readers here have joined me in tracking the internal AI implementations by Klarna’s CMO​ ​David Sandström.

I stumbled over some numbers from “The Information” that paints a picture more of efficiency than opulence.

(BTW - The new Klarna + Uber Eats deal is a new dangerous combo).

Klarna, the Swedish fintech company, has reported annual savings of $10 million attributed to its AI-powered marketing initiatives. While this figure is notable, it represents a modest portion of the company's overall financials. In 2024, Klarna achieved a net profit of $21 million, marking its first profitable year after a series of losses following its U.S. expansion. Despite the $10 million savings from AI, these reductions constitute only a small fraction of Klarna's total expenses. For context, the company's operating expenses in 2023 were significantly higher, indicating that while AI contributes to cost efficiency, its impact on the broader financial landscape remains limited.​

Sources: The Information

And speaking of The Final Countdown

Here are the top 10 AI-related stories associated with the Super Bowl:

1. OpenAI's Super Bowl Advertising Debut.

OpenAI is set to air its first television commercial during Super Bowl LIX, marking the company's initial foray into commercial advertising.

The company that makes ChatGPT. They’re spending millions to tell us they exist.

I mean, do they really need a Super Bowl ad? “Hey, have you heard of ChatGPT?”

And OpenAI is dropping $8 million for 30 seconds. For what?! “Hi, we’re ChatGPT. Everyone knows we exist.”

Google did this last year too. I bet you had already heard of them too.

And you know Microsoft’s involved. Of course, they are. Microsoft is like that friend who always wants to split the bill evenly, even though they only had a salad. “We own part of OpenAI!”

And this is all happening at the Super Bowl! The most expensive, most ridiculous place to advertise anything. The game where we pretend every commercial is the greatest thing we’ve ever seen. “Oh my God, did you see the one with the puppies and the AI chatbot?”

So now, OpenAI is spending all this money, hiring a marketing officer, raising billions of dollars—what’s next? ChatGPT can probably replace Elon and his DOGE gang; optimizing America.

Reuters

2. Google's Gemini AI Super Bowl Campaign.

Google has launched two new Super Bowl commercials emphasizing AI's role in everyday life via its Gemini AI on Pixel phones. The Verge

3. Meta's Ray-Ban AI Eyeglasses.

So, what's the deal with these Super Bowl ads? You got Chris Hemsworth, Chris Pratt, and Kris Jenner—two Chrises and a Kris—teaming up for Meta's big commercial. They're promoting these Ray-Ban Meta glasses, right?

The glasses have A.I built-in and can describe in limited detail what you’re looking at. Think Google Glass with less of a surgeon look. We got our kid a pair for his birthday. They have a source for popping prescription lenses into the frames. More convenient than Airpods with similar functionality. Camera is great for photo and video. I believe you can live stream as well.

In one of the spots, Hemsworth and Pratt are in an art gallery, trying to look all sophisticated. Next thing you know, they're eating the infamous $6 million banana that must be a couple of months old. Who knew fruit could be worth more than a penthouse? And, of course, it's owned by Kris Jenner. Because if anyone's got a pricey banana lying around, it's gotta be a Kardashian.

Hemsworth says they had a blast filming this, poking fun at themselves. He even admits they were "trying to look sophisticated and failing miserably." Sounds like a typical day at the museum for these guys.

Meta's new Super Bowl ad promotes Ray-Ban Meta glasses, featuring Chris Hemsworth, Chris Pratt, and Kris Jenner. People.com

Introduction

AI agents are a rapidly evolving technology poised to transform how humans interact with computers and software. These agents are not simply chatbots or single-purpose tools, but rather, they are intended to be autonomous entities that can perceive their environment, make decisions, and take actions to achieve specific goals. They represent a shift from traditional software applications to more dynamic and personalized forms of computing.

What is an AI Agent?

At its core, an AI agent is a type of software that responds to natural language and can perform various tasks based on its understanding of the user. Unlike bots that are confined to a single application and require explicit instructions, agents are proactive, capable of learning user preferences and patterns, and can make suggestions before being asked ... They can operate across multiple applications and improve over time through their ability to remember past activities and recognize intent.

My Favorite Example

My favorite example - You want to take a vacation. Hotel, airfare, event tickets, tours. In theory your agent will …

• Book the best deal on your airline of choice, complete with aisle seat.

  • You’re on the upgrade stand-by list (BTW).

• Book the best deal with your hotel(s) of choice depending on number of destinations, of course.

• Rental car, show tickets, and tours are all negotiated.

You are left with an itinerary with all planning completed.

You might ask, “Who is my agent interacting and negotiating with?” With all of the vendor agents, of course. Welcome to your new audience you‘ll be advertising to and currying favor with - bits of computer programming.

Headlines

1. Sora is out.

And if you’re in the U.S. you had the opportunity to discover that Sora has its own interpretation of bowling (among other things). I used up $20 in a couple of hours. Another $20 and I could probably solve it. Their $200/month deal is ideal for production companies. You spend more than that on coffee or toilet paper. sora.com.

2. Veo 2 is out, and it might be better.

If you’re not already in, you can hop onto the waiting list.

This is one way to compare the main players in AI steak prep.

Source: https://x.com/i/status/1868850653759783033

3. ChatGPT Projects are fancy folders for your AI chats

OpenAI is making it easier to keep track of things you’re working on in ChatGPT.

The Verge Dec 13, 2024, 4:50 PM CST

Projects was announced as Day 7 of OpenAI’s 12 days of “ship-mas.” Previous announcements included the release of the Sora video generator, ChatGPT’s Canvas view, and the $200-per-month ChatGPT Pro subscription.

Users can now utilize Projects to upload files, associate them with conversations, and provide specific instructions for ChatGPT.

For instance, a project manager could create a project for a website development task, upload necessary files, and direct ChatGPT to focus on using its Canvas tool for coding.

By introducing tools that streamline workflows and enhance organization, OpenAI is positioning ChatGPT as a versatile productivity platform with advanced, integrated features.

Executive Summary:

The advertising world is reeling from the bombshell announcement of Omnicom Group's acquisition of Interpublic Group of Cos. (IPG), creating a behemoth poised to surpass WPP as the world's largest agency holding company. This strategic move, fueled by the pursuit of data and technological supremacy in the face of mounting pressure from tech giants, has sparked intense debate amongst industry leaders regarding its potential impact on clients, agencies, and the future of advertising itself.

Source: Ad Age Nov 22, 2024

Coca-Cola Responds

Coca-Cola's recent holiday marketing campaign represents a significant case study in the emerging mash-up of artificial intelligence and brand communication strategies. The campaign, which leveraged AI-generated content across multiple commercial productions, has sparked a nuanced discussion about the role of generative AI in creative industries.

The campaign's core innovation was the creation of three separate AI-generated commercials reimagining Coca-Cola's iconic 1995 "Holidays Are Coming" advertisement. Produced by three distinct AI studios - Secret Level, Silverside, and The Wild Card - the project serves as a practical exploration of AI's current capabilities in video content creation.

Pratik Thakar, Coca-Cola's VP and global head of generative AI, positioned the campaign as a strategic technological experiment rather than a wholesale replacement of human creativity. His commentary reveals a measured approach to AI integration, emphasizing that the technology is a tool to enhance, not eliminate, human creative processes.

Key Analytical Observations

These emerge from the campaign and Thakar's subsequent interview:

1. Technological Capability Assessment
   The AI-generated commercials demonstrate the evolving sophistication of generative video technologies. While not perfect, they represent a significant milestone in commercial-quality AI video production. The campaign highlights both the potential and current limitations of AI-driven content creation.

2. Consumer Perception Research

   Coca-Cola's extensive research in North America and Europe revealed an intriguing consumer response. Contrary to industry fears, consumers appeared more focused on narrative and entertainment value than the technological means of production. This suggests a potential paradigm shift in audience expectations and technological acceptance.

Share The Ad Stack

1. rategic Implementation
   The brand's approach was notably deliberate. Thakar emphasized that AI deployment is not universal but contextual. The holiday campaign's "fantastical, hyper-realistic storyline" was specifically chosen to leverage AI's strengths, indicating a nuanced, strategic approach to technological integration.

2. Industry Reaction Dynamics
   The campaign exposed the complex landscape of AI adoption in creative industries. Responses were polarized, with some viewing it as innovative and others as potentially threatening to traditional creative roles. And some saw this as killing off Santa. This reflects broader societal debates about AI's impact on employment and artistic expression.

3. Transparency and Narrative Control
   Thakar acknowledged a potential communication gap, suggesting that more proactive storytelling about the AI process might have mitigated negative reactions. This reflects an emerging challenge for organizations implementing transformative technologies: managing public perception and understanding.

Methodological Considerations

• The campaign involved real actors and original music composition

• Human creatives remained central to decision-making processes

• AI was positioned as a collaborative tool, not a replacement

Comparative Context:
The approach mirrors previous technological transitions in marketing, such as the introduction of computer-generated imagery (CGI) in the early 1990s. This suggests a pattern of technological integration where initial skepticism gradually transforms into accepted practice.

Limitations and Future Implications:
While the campaign demonstrates AI's potential, it also reveals current constraints. The technology still requires significant human intervention, and the quality of output remains variable. Future developments will likely focus on refining generative capabilities and improving human-AI collaboration.

Summary (and then we’re done with this topic)

The Coca-Cola AI holiday campaign thus emerges not as a revolutionary disruption, but as a calculated exploration of emerging technological capabilities. It represents a measured step in understanding how generative AI can be strategically integrated into established creative processes.

The broader significance lies not in the specific campaign, but in what it reveals about organizational approaches to technological innovation: a balanced, experimental methodology that seeks to augment rather than replace human creativity.

Source: Ad Age Nov 22, 2024

HEADLINES

** Niantic uses Pokémon Go player data to build AI navigation system.

This is one of the greatest apps ever (in my life). A shout out to my son Julian - he and I played Pokémon Go together on both coasts, in the Midwest, and our nightly neighborhood rumbles.

The company calls OUR creation …

… a "large geospatial model" (LGM), drawing parallels to large language models (LLMs) like the kind that power ChatGPT. Whereas language models process text, Niantic's model will process physical spaces using geolocated images collected through its apps. - Ars Technica

** AI-powered search engine Perplexity is venturing into e-commerce.

Last week, Perplexity unleashed its latest technological shoppertunity upon unsuspecting American shoppertunists. The company, ever-hungry for market dominance, has unveiled a shopping feature that ensnares its paying customers in an algorithmic web of "recommendations," trapping them within its search results. Most ominous of all, users can now complete their purchases without ever escaping to a retailer's website – another nail in the coffin of traditional e-commerce as we know it. - TechCrunch

** Suno v4 ushers in new era of AI-generated music

Better audio, sharper lyrics, and more dynamic song structures now available in beta to Pro & Premier. With its fourth major update, the music generation platform V4 introduces more “soulful” vocals, with things like grunts and breaths that add to the realism. Plus, it can now generate sharper lyrics and higher-fidelity audio.
- 11/19/24 Suno AI

FEATURE

Klarna - AI Year in Review
Goes Deep With AI Customer Service

Executive Summary:

1. RAPID AI ADOPTION

The impact of Klarna's AI assistant was immediate – within its first month, it processed two-thirds of all customer service cases, performing at a level that would typically require 700 full-time human representatives.

2. IMPROVED CUSTOMER EXPERIENCE

The AI system revolutionized service delivery, slashing average resolution times by over 80% – from 11 minutes to under two minutes – without compromising the quality of customer support compared to human agents.

3. GLOBAL REACH AND ACCESSIBILITY

Operating around the clock across 23 markets, the AI assistant communicates in 35 languages, positioning Klarna to save an estimated $40 million USD in 2024. While customers retain the option to speak with human agents, the system's multilingual capabilities have particularly enhanced service accessibility for immigrant and expat communities.

Thanks for reading The Ad Stack! Subscribe for free to receive new posts and support my work.

Subscribed

Klarna’s AI assistant is working harder than most of us.

When ChatGPT emerged in November 2022

Klarna's Co-founder and CEO Sebastian Siemiatkowski moved swiftly to capitalize on its potential. The company soon achieved dual firsts – launching Europe's first ChatGPT plugin and becoming the global fintech industry's earliest adopter, laying the foundation for more ambitious AI initiatives.

Klarna, the buy-now-pay-later company, is changing how we think about AI in customer service. Following its partnership with OpenAI last year, Klarna’s AI assistant has delivered impressive results just one month after launch. 

Massive Improvement in Customer Communication

Serving Klarna's 150 million customers worldwide, the OpenAI-powered assistant streamlines the shopping and payments journey through multilingual support, automated refund processing, and proactive financial advisory services.

• It handled 2.3 million conversations in its first month; two-thirds of Klarna’s customer service chats.

• It is doing the equivalent work of 700 full-time agents.

• It is on par with human agents in regard to customer satisfaction score. It didn’t just handle the volume, it also improved efficiency. 

• It is more accurate in errand resolution, leading to a 25% drop in repeat inquiries. Resolutions were cut by nine minutes and customer satisfaction scores stayed steady.  

• Customers now resolve their errands in less than 2 mins compared to 11 mins previously.

• The implementation cost $2-$3 million, with a projected $40 million profit in 2024. 

• It’s available in 23 markets, 24/7 and communicates in more than 35 languages.

Refunds in 35 languages

The assistant works in over 35 languages, helping with everything from refunds to financial advice. Human agents are still available for more complex queries, while AI focuses on simpler tasks. 

Despite its success, the AI assistant has sparked concerns. Two years ago, Klarna laid off 700 employees — the same number the AI now replaces. 

Klarna denies a direct link between the layoffs and its AI investment, saying the roles were reassigned within its global service partners, which employ over 650,000 people. However, the timing has raised questions. But then in 2023, the company announced a hiring freeze which was confirmed to be partially due to AI reasoning. 

The AI assistant costs $2–$3 million to develop and is expected to bring in $40 million in profits. Klarna’s AI shows how businesses can improve customer service and efficiency with technology. 

Yet, it also reminds us of the need to carefully balance innovation with its effect on workers.

ChatGPT Enterprise Across The Organization

Klarna’s adoption of AI hasn’t been limited to its products—it extends deep into the organization as well. Klarna has made ChatGPT Enterprise available to all its employees around the globe, enabling access to the power of OpenAI in the familiar and easy-to-use ChatGPT interface, with protected company data.

• 90% of Klarna’s employees are using generative AI tools powered by OpenAI daily.

• Their success spans across the org, with Communications, Marketing, and Legal teams seeing adoption rates of 93%, 88%, and 86% respectively.

• Klarna is seeing a wide variety of additional use cases, from building software to streamlining customer service.

So what’s next?

Klarna Plans to ‘Shut Down SaaS Providers’ and Replace Them With Internally Built AI. The Tech World Is Pretty Skeptical

The buy-now-pay-later giant is leaning into automation and slashing its partnerships with Salesforce and Workday.

-Inc. SEP 12, 2024

Klarna is no longer using Salesforce, a platform that aggregates and packages sales and marketing data for businesses. The company has also removed the HR and hiring platform Workday from its tech stack, a Klarna spokesperson confirmed to Inc. 

As rumors circulate about Goldman Sachs' involvement in a potential IPO, Klarna's aggressive AI strategy and push toward profitability could strengthen its position for going public.

If you find this edition of value please feel free to leave a note in the comments. And if you’re so inclined, you can now even buy me a cup of coffee.☕️☕️☕️☕️☕️

Please help share this far and wide:

• Like this post!

• Re-stack it below!

• Share this post on Substack and other social media sites:

• Feedback is always helpful.

Leave a comment

Thanks for the read!

-pb

1. Coke likes to roll the dice with high visibility projects.

Doing remakes of past successes involves palpable risks in any industry. Re-doing live action creative with AI is risky as it facilitates a scene-for-scene comparison between the two approaches despite the fact that the new spots are interpretations versus clones of the original.

2. The Best Available Tools were Used

The software used were among the most fully-baked tools out there. And every month or so we experience one or two new notable software entries into the pool coupled with notable improvements with many of the most notable players.

3. The Tools Have Limitations

It’s not there yet and we’ve got a ways to go. There are details that don’t work. Instead of calling those out, I would note that there are a ton of detail here that does work and took countless hours to achieve. In my experience you are forced at times to bend AI software to your will. You take what you get, adjust the creative and expectations, and continue on. And you render until you get there. You adjust the prompts with every adjective and keyword you can imaging.

4. AI Provides an Interpretation Based on How The Model Has been Trained

This project allows for this. A frame-by-frame AI clone of the original might prove disastrous because you lack the precision you have with live action and 3D software. AI gives you an interpretation of what you’re after - even when using reference imagery.

We have much to look forward to.

Coca-Cola Goes All In On AI Christmas

This is an interesting proposition. Have three AI-focused studios take a cut at redoing a Coca-Cola Christmas holiday classic. For background on this topic, I referenced three articles and the reviews ran the gamut.

• Ad Age’s “Inside Coca-Cola’s first AI-generated TV ads” was written with detail without opinion.

• Futurism’s “Coke's AI Commercial for the Holidays Has Us Wondering If We Live In a Fallen World” offers a blunt review that is to the point, and on fire.

• The New York Post gets excited by most things. Their “Coca-Cola ripped for ‘ugly’ AI-generated Christmas commercial: ‘Dystopian nightmare’” still scores a few cogent points.

• **Last minute entry - “I was wrong. People love Coca-Cola’s AI remake of a Christmas classic.” This did not make deadline but was added for your consideration. The upshot - when viewers didn’t know it was made with AI, these concepts were among the highest responses for any of Coke’s holiday “Truck ads”.

  Testing showed success was in long-term brand-building, not tech. - 11/15/24 The Drum

Here comes a pretty strong list of AI resources that agencies are experimenting with - courtesy of Ad Age. The publication presented nine software reviews per edition over three editions. I combined them all into one list. All of the software titles provide a link to the appropriate site. Some, you might have heard of. Some not.

So I’ll lay it all out for you as an AI Buffet and let you pick through what you already may know and perhaps you’ll find a new gem or two. And if you wish, perhaps even list your additional favorite(s) in the comments section below.

Here is a summary of all three Ad Age installments.

AI tools (27) agencies are embracing right now

Holding companies such as WPP, Havas and Publicis Groupe have pledged to spend hundreds of millions of dollars on the technology in the coming years, with some even building proprietary systems. Major partnerships are a big focus, too.

My initial goal of 200 engaged readers has reached 700 with open rates still holding at 60%. It’s all interesting to me. I’m not pursuing massive subscriber numbers. My subscriber list is curated. I am not interested in bombing inboxes with unopened email.

Thanks for reading The Ad Stack! Subscribe for free to receive new posts and support my work.

Subscribed

I position my readers to lead AI discussions for their agencies and brands.

“Apple Sharply Scales Back Production of Vision Pro”

It’s not AI but retail creative tech counts too.

A ton of buzz began this past week. A number of well-known tech sources quoted a report by The Information, displaying the above headline; “Apple Sharply Scales Back Production of Vision Pro.”

From The Information:

The move suggests that Apple has enough inventory built up to meet demand for the foreseeable future. It follows Apple’s decision earlier this year to focus on building a model that’s cheaper than the current version—which retails for $3,500—for possible release by the end of 2025, as The Information has previously reported.

The Takeaway

• Apple could suspend Vision Pro assembly as soon as November
• Suppliers have made enough components for roughly 600,000 headsets
• Vision Pro assembler Luxshare has halved production to around 1,000 units a day

As part of that decision, Apple suspended work on a second-generation version of the high-end device for at least a year, according to another person directly involved in the company’s supply chain.

The tech industry echo chamber speculated on this report:

Tom’s Guide - Report: Apple may stop producing Vision Pro by next month

This comes quickly on the heels of a Wall Street Journal interview where Apple CEO Tim Cook conceded that the headset was failing.

Laptop Mag - Apple may stop making Vision Pro headsets — but something better is on the way

The Apple Vision Pro walked so Apple's next headset can run.

PC Gamer - Apple reportedly slashes Vision Pro headset production and cancels updated headset as sales tank in the US

Very expensive, fairly clunky, so not exactly surprising.

What does this all mean?

1. If you promised someone an Apple Vision Pro for the holidays, you might want to buy them while they are in-stock.

2. If you have $3500 to burn, buy one now. You deserve it. You can buy the next version too whenever the heck it comes out.

3. Me? I’ve got an Oculus Rift and a Meta Quest 2 so I’m set until Apple figures out what’s next. It won’t be soon.

4. Otherwise, whomever figures out a useable AR that will support a worthy software catalog will win this. I would thing the glut of generative AI companies you catch a clue and pivot.

5. The Meta Ray-ban glasses sound get regarding design. Bit what the fongool is the difference between these and Google Glass from 10 years ago? Doesn’t seem much different. Spyware at best.

Have a great weekend people!

1. The Looming Irrelevance Apocalypse: How Ads Will Haunt Your Every Click.

Source: Ad Age: How Amazon’s 4Cs of relevancy will unlock the future of addressable advertising. October 22, 2024.

Per an Ad Age article, prepare to be bombarded. The world of advertising is about to get a whole lot creepier, and there's no escape. Remember those annoying, irrelevant ads that used to plague your online experience? Soon, they'll be replaced by something far more insidious: ads that know you better than you know yourself.

This isn't the dystopian future everyone depicts with Midjourney; it's October 2024, and Amazon is leading the charge. They claim to have unlocked the key to "addressable advertising," a fancy term for bombarding you with ads based on your every digital move.

Per Amazon:
Addressable advertising allows advertisers to deliver relevant messages to desired audiences based on available signals—regardless of the presence of ad identifiers like third-party cookies. This approach is key to reaching and gauging the effectiveness of online ad spaces, and in driving engagement and ad relevance.

The Problem: You're Drowning in Irrelevant Ads

The article starts with a relatable scenario: you're scrolling late at night and BAM! An ad for a retirement cruise pops up. It's a metaphor for the current state of advertising – a sea of irrelevant noise that annoys consumers and wastes money for advertisers.

The (Supposed) Solution: Amazon's Big Brotherly Embrace

The "solution" comes in the form of Amazon's DSP Ad Relevance, a system that uses "advanced machine learning models" and "unique signals" from your Amazon activity (shopping, browsing, streaming habits – you name it) to target you with laser-like precision.

But Here's the Catch (There's Always a Catch)

Forget about pesky cookies! Amazon claims they can achieve this "relevancy" without them. How? They haven't quite said, but it involves a combination of "direct publisher relationships" and "hashed durable ad IDs." Sounds reassuring, right?

A Glimpse into the Future (It's Not Pretty)

Imagine this: you're scrolling through news about the housing market, suddenly an ad pops up for a moving company you casually mentioned to your Alexa a week ago. Or you're reading a review for a new restaurant, and there's a personalized discount code staring back at you. This is the future Amazon envisions – a world where ads track your every whim and desire.

The "Success Stories" They Want You to Believe

The article throws in a few "success stories" – brands reaching new customers and lowering costs. But is it really a win for consumers? These ads might be more "relevant," but are they truly wanted? Merchants who know what I’m shopping for based on search have ALWAYS creeped me out. Right?

The Final Word: Prepare for the Inevitable

The article ends with a chilling image. Forget the irrelevant cruise ad; now you're bombarded with a Caribbean resort offer because you once searched for "New Year's destinations." This is the power of relevancy, they say. But is it really power, or just another form of manipulation?

The Takeaway: We're All Doomed (or at least our privacy is)

The future of advertising seems bleak. Amazon's "relevancy" might drive sales for brands, but at what cost? Get ready for a world where ads not only follow you around the internet, but anticipate your every need (or perceived need). This isn't just about irrelevant cruise ads anymore; it's about the erosion of your privacy and a future where every click, search, and purchase becomes fuel for the advertising machine.

Read more: Ad Age: How Amazon’s 4Cs of relevancy will unlock the future of addressable advertising. October 22, 2024.

It begins with attitude. And rolling hard with purpose.

We first presented the Klarna AI implementation discussed this here. The executive summary is/was:

Klarna, the AI-powered global payments network and shopping assistant, in a daring move that defies convention, has slashed its sales and marketing expenditure by a staggering 11% in Q1 2024. Yet, this noteworthy cost-cutting measure belies an audacious strategy - a masterful leveraging of artificial intelligence to unleash a torrent of campaigns and marketing collateral updates.

Now, detailed in Ad Age September 25, 2024, Klarna offers a peek behind the AI curtain by way of a conversation with Klarna’s CMO Sandstrom.

“The culture here at Klarna is to move extremely fast and adopt AI as much as you can, but not recklessly,” he said. “That is the culture here internally, and if you don’t like that, I don’t think this is a great place to work right now.”

The availability of AI tools is surprisingly not the major part of the implementation. Add attitude and expectations large parts of the strategy.

Never before have a I seen gratuitous amounts of “Mind Blown” references regarding an AI-powered tool. It is an amazing tool with many different applications. What has captured everyone's attention is it ability to crank out a rather high-quality podcast based on feeding it a URL which has content for them to talk about. And you can add additional references from uploaded documents to other URLs.

Notebook will sort it out into a show, and it sounds quite human and balanced. They nailed the humanity that needs to be present in the voices while it also coordinates a conversation between two different AI entities and all your content. It actually is quite amazing.

Summary

ByteDance, the company behind TikTok, has entered the AI video generation market with two new large language models called Doubao-PixelDance and Doubao-Seaweed. These models are designed to create videos from text and image prompts and are intended to compete with similar models like OpenAI’s Sora. The development of these AI models highlights the increasing competition in the field of AI video generation, as companies seek to develop sophisticated tools for creating engaging and realistic videos. However, the rapid development of these models also raises concerns about the quality and reliability of quickly assembled AI tools. The rise of open-source AI tools like Open NotebookLM, which was built in a single afternoon, demonstrates how accessible AI development has become. This shift presents both opportunities and risks, with potential benefits like cost-effectiveness and customization but also concerns regarding data privacy, security, and ethical use.

Joins Stability AI

James Cameron, Self-Proclaimed God of Cinema, Deigns to Grace Stability AI's Board of Directors.

Los Angeles, California September 24, 2024 - Stability AI, a company desperately in need of some headlines, announced that James Cameron has magnanimously agreed to join its Board of Directors. The announcement was made by Stability AI CEO, Prem Akkaraju, who was reportedly giddy with excitement.

Cameron, known for his modest films about small boats and blue cat people, has been a driving force in convincing Hollywood that more pixels equal better movies. As a director, writer, and producer, he has pushed cinematic boundaries and brought to life some of the most expensive and time-consuming films of all time, including The “Terminator(s)” 1-4, “Titanic” (spoiler alert - it doesn’t end well) and the "Avatar" series (humans are usually bad, aliens good). No argument here.

Cameron's artist-centric perspective, paired with his business acumen and penchant for yelling "Action!" really loudly, will support Stability AI in continuing to unlock new opportunities to empower creators to tell stories in ways once unimaginable, like filming the same movie for 13 years.

For more info watch the CNBC video.

SOURCE: RUNWAYML.COM

Lionsgate and Runway have entered into a first-of-its-kind partnership centered around the creation and training of a new AI model, customized on Lionsgate’s proprietary catalog. Fundamentally designed to help Lionsgate Studios, its filmmakers, directors and other creative talent augment their work, the model generates cinematic video that can be further iterated using Runway’s suite of controllable tools.

“Runway is a visionary, best-in-class partner who will help us utilize AI to develop cutting edge, capital efficient content creation opportunities,” said Lionsgate Vice Chair Michael Burns. “Several of our filmmakers are already excited about its potential applications to their pre-production and post-production process. We view AI as a great tool for augmenting, enhancing and supplementing our current operations.”

“We’re committed to giving artists, creators and studios the best and most powerful tools to augment their workflows and enable new ways of bringing their stories to life. The history of art is the history of technology and these new models are part of our continuous efforts to build transformative mediums for artistic and creative expression; the best stories are yet to be told.” said Runway co-founder and CEO Cristóbal Valenzuela. “Lionsgate has an incredible creative team and a clear vision for how AI can help their work – we’re excited to help bring their ideas to life.”

As we continue our work with Lionsgate and other studios, we’re also considering ways to license or otherwise offer these models as templates that can be made available to individual creators interested in building and training their own proprietary models. For companies interested in fine-tuning and custom models, visit our Enterprise page.

In a plot twist that surprises absolutely no one except my family, Amazon has decided that what their Alexa voice assistant really needs is a brain transplant. The online retail giant, not content with merely accidentally listening to your every word, is now set to make Alexa smarter - or at least putting on a better show.

Enter "Remarkable" Alexa, the soon-to-be-released upgrade that promises to turn your home into a sitcom of AI-powered hijinks. And who's the star of this digital dramedy? None other than Anthropic's Claude, the AI chatbot that apparently impressed Amazon so much, they decided to outsource Alexa's intelligence altogether.

Why Claude, you ask?

Well, it seems Amazon's homegrown AI’s lagging response times put us to mind of the comfort provided by a 28.8 baud modem - taking a leisurely 6-7 seconds to respond. In the fast-paced world of asking your speaker what the weather's like, that's practically an eternity.

Does Amazon truly give a sh-t about your Alexa experience?

Maybe. Here are two items they probably find more prescient:

We first referenced this Great Disappearing Act back a few editions back. The AI functions, services, features, and versions will begin melding into the everyday things we already do. Microsoft 365 recently scooted CoPilot into that product line to give it a home that makes sense. An attempt to make it part of your everyday toolbox without firing up a non-integrated AI capability.

This from Ad Age September 3, 2024

Thanks for reading The Ad Stack! Subscribe for free to receive new posts and support my work.

Formerly, Copilot was isolated from Microsoft’s 365 umbrella, which houses its core productivity tools including Teams, Outlook and other enterprise products. Copilot could be used within these apps, for example, to summarize data or generate text, but the integrations were branded as two distinct systems coming together, such as “Copilot in Word” and “Copilot in Sales.” Now, these use cases will be referred to in a way that reflects Copilot’s presence within the same ecosystem of apps it is assisting: “Microsoft 365 Copilot for Word” and “Microsoft 365 Copilot for Sales.”

No agency was involved in the change.