AI Tsunami: Are You Falling Behind?
DeepSeek Proves That AI Is Moving Faster Than Marketing Teams Can Adapt
Welcome to the Savvy AI newsletter. Three to four times a month, I explore what the AI Era means for marketing and business. In this space, I also share excerpts from my recently launched book on Marketing with AI. If you find the insights valuable, feel free to share the newsletter with others and encourage them to subscribe.
Earlier this week, I took a quick trip to the IAB Leadership Meeting in Palm Springs to gauge how marketers, agencies, and adtech firms are thinking about AI in 2025. My takeaway? We may still be asking the wrong questions—debating whether ChatGPT can handle market research or assuming DeepSeek will dictate our breakfast choices next week—rather than focusing on how AI will reshape our jobs, drive real leaps in marketing performance, and potentially spark a marketing renaissance.

What makes these summits worthwhile isn’t just the sessions—though hearing Ron Howard and Ed Norton dissect AI in filmmaking was a standout. The real insights come from the off-the-record conversations: a CMO explaining why she’s skipping the Super Bowl this year, another dismissing AI entirely due to ChatGPT’s hallucinations, or a former PepsiCo colleague revealing how The New York Times has developed an LLM-driven contextual targeting tool that analyzes its vast content library to create custom audience segments—helping advertisers reach the right readers without relying on third-party data.
Nominate top AI talent by 3 p.m. ET today! AI Trailblazers is partnering with ADWEEK to launch the AI Trailblazers Power 100, celebrating leaders shaping the future of AI in marketing. If you know someone making a real impact, submit your nomination now and help recognize standout individuals from brands, agencies, and the adtech/martech ecosystem who deserve special recognition.
Making Sense of DeepSeek
One of the greatest fallacies in marketing is the impulse to chase the latest shiny object. Time and again, many of us have fallen into this trap—only to be burned, sometimes even ridiculed by skeptics outside the field and the more cautious among us within it. That said, I don’t believe this is always the case. In fact, I built my career by being a first mover. And when it comes to AI specifically, I’ve argued extensively that its impact on business—particularly in marketing—is under hyped.
However, it’s important to recognize that technological breakthroughs often trigger tsunami like seismic shifts in other business functions before their full effects reach marketing. We need to account for that—which brings me to DeepSeek, the Chinese large-language model whose launch sent Nvidia’s market cap tumbling by $600 billion in a single day. First, why DeepSeek matters to Corporate America—and what marketers should be prioritizing right now.
Why DeepSeek Matters right now
A True Technological Breakthrough
DeepSeek represents a significant leap forward. It has delivered a large language model (LLM) that was developed faster and at a dramatically lower cost than any of the US based competitors like ChatGPT. This challenges fundamental assumptions about the economics of GenAI—time, training, and architecture—all of which shape pricing models for the business world. So much so that Microsoft, Snowflake and Amazon have already begun offering access to it via Azure, Cortex and AWS.Accelerating AI Adoption
Speaking of which, the most compelling impact of DeepSeek is its ability to drive enterprise AI adoption by lowering licensing costs dramatically. It’ll force costs down across the entire AI provider industry. More companies can now deploy LLMs without immediate pressure for massive returns, making AI adoption more accessible. Case in point: Goldman Sachs is already looking to deploy DeepSeek across its workforce once a security screening process is completed. New York Life is just another example of a company exploring its usage too.A Win for Open Source AI
The future of AI—if we want it to be truly People First—requires greater accessibility, transparency and innovation that’s only possible through open-source models (something that I’ve advocated for over the last year). DeepSeek builds upon Meta’s Llama models, the PyTorch ecosystem, and other open-source contributions and, in turn, is making its code available to the broader technology community. This fosters faster innovation, more transparency, less bias, more limited misinformation, and fewer monopolistic pricing practices. Open-source models are positioned to win long-term, and DeepSeek’s launch strengthens that trajectory.The Reality of Globalization
The world is too interconnected—both in the knowledge economy and manufacturing—to isolate innovation. While concerns about a Chinese-controlled LLM being adopted across the world are valid, a reactionary, nationalistic approach won’t address them. The question requires a nuanced, strategic response—one that safeguards national interests while fostering international collaboration. The "China question" in AI is not a binary issue of security versus globalization; it demands a more sophisticated frenemy strategy. In fact, even China is still sorting through its own AI policies.Potential Ethical Concerns
Reports from The Financial Times and others suggest that DeepSeek may have trained on OpenAI’s outputs through knowledge distillation—where a parent model teaches a student—which, while technically prohibited, remains common practice. While OpenAI itself faces legal scrutiny over its training methods, this raises broader ethical questions about how AI models are built and governed. Some may argue OpenAI is getting a taste of its own medicine, and legal experts increasingly believe it has little ground for a case against DeepSeek. Needless to say, the pot can’t call the kettle black.
Why Marketers may not be ready
You’re Still Training Your Teams
Many marketing teams have yet to fully explore AI’s potential, let alone determine how it applies to each sub-function and role. Most are still focused on mastering the basics—using AI for customer research, content generation, media targeting, and analytics. If you’re a marketing leader, your priority should be getting your team trained and AI-ready on the fundamentals before letting a few innovation leads experiment with DeepSeek.You’re Still Defining Business Use Cases
Without clear business applications, premature AI investments risk inefficiency and wasted resources. Moving too fast—or too slow—can be equally problematic. If you’re a marketer in Corporate America, you’re likely still determining the best use cases for AI. Even if you’ve identified them, you may still be evaluating tools, fending off eager vendors, and navigating the aggressive sales pitches from agencies pushing their own AI solutions.Your Organizational Structure Isn’t Built for AI Yet
The shift to AI-enabled workflows will bring fundamental changes to roles, responsibilities, and team dynamics. Or, as the President of my alma mater, Dr. Larry Kramer of the London School of Economics & Political Science, put it at a private event I attended last night—it will lead to significant job displacement that we’re not remotely prepared for. Few C-Suite leaders have begun grappling with this hard reality: How do you future-proof your function for such an uncomfortable shift? Even fewer are considering the deeper question—how the nature of work itself will evolve, and what should be left to humans versus AI agents.You’re Already Overwhelmed with AI Tools
Marketers are being inundated with AI-powered tools like Writer, Copy.ai, Jasper, and Anyword while also being told that their existing Martech stack, from Adobe to Salesforce to HubSpot, already has more than enough AI to meet their needs. Meanwhile, CEOs are fielding pitches from enterprise AI vendors like OpenAI, Salesforce, Amazon, ServiceNow, Microsoft, and Google—not to mention a flood of AI-first startups. The real priority isn’t fixating on LLMs or chasing every new tool—it’s understanding what truly adds value and rebuilding your Martech (and Adtech) stack for the AI era.LLMs Are a Technology Team Decision
Large language models are typically procured through cloud providers—a decision owned by CTOs, not CMOs. DeepSeek is now available on Amazon Cloud and several other platforms, as I mentioned earlier, but unless you have well-defined use cases specific to DeepSeek, there’s no urgency to push your CTO for access. Marketers have enough else to worry about right now, from optimizing AI-powered content strategies to navigating shifting privacy regulations. Instead of chasing the latest LLM, focus on integrating AI in ways that drive measurable business impact.The Real AI Winners Will Be Those Solving Business Problems
The biggest barrier to enterprise AI adoption is cost, followed closely by education. DeepSeek’s arrival is driving industry-wide cost reductions—a net positive. But the real question remains: What business problems are you solving and more tactically is your content AI ready? Technology should serve business needs, not the other way around. Identifying the right problems to solve, determining the necessary effort and resources (including whether your brand content is AI ready), leveraging the right people, and choosing the right technology is far more complex than it seems. That’s the real job of a CMO—not rushing to figure out how to use DeepSeek. It’s a responsibility that rarely appears explicitly in CMO job descriptions, but it deserves more focus than ever.
The Bottom Line
DeepSeek is undoubtedly a transformative milestone in AI shaking up the AI industry like a tsunami does, but marketers shouldn’t feel pressured to react prematurely. Instead, prioritize building AI literacy within your teams, defining clear business use cases, and ensuring your organization is truly prepared for the AI era are just as essential. The real winners won’t be those who adopt AI for the sake of it, but those who strategically apply it to drive meaningful business outcomes—and do so with speed and precision.
Where I’ve been
Last week, I had the pleasure of joining Geoff Livingston and Greg Verdino on their No Brainer - An AI Podcast for Business. We discussed the state of AI in business, what it portends for the future and how we call prepare for this new reality. The episode will drop on Monday.
Marketing with AI for Dummies continues to draw attention, sparking speaking engagements and advisory opportunities focused on AI education, inspiration and marketing roadmapping. If you would like to partner —whether for a speaking event, team education, or strategic planning—don’t hesitate to reach out.
What I’m reading
What to know about DeepSeek (Axios)
Gen AI Is Helping The New York Times (Ad Exchanger)
Which AI to use now (One Useful Thing)
Turbocharging Org Learning With GenAI (MIT Sloan Review)
Recent Savvy AI Articles
2025 AI Predictions You Can’t Afford to Miss (January 2025)
Fires & Falsehoods: A Tale of Two Crises (January 2025)
Omnicom’s Big Move, Brand Clarity, & Looking Ahead (December 2024)
Search Wars Part 2. Google May Win (December 2024)
Marketing with AI for Dummies launches! (November 2024)