Postcard from the USA! Two worlds about to collide!

I flew from Adweek in New York to MAICON in Cleveland within days. Same country. Same industry. Two completely different universes.

At Adweek—the 20th edition of advertising's flagship event, with over 260 sessions spread across THE PENN DISTRICT. An electric event that had New York buzzing. It was amazing just being in the hotels, cafes and bars in the area hearing all the marketing buzzwords doing the rounds. Hushed whispers speculating if some lost industry company would make a surprise comeback or what various legends might say on their next podcast appearance.

On stage: AI buzzwords, bold visions, and more than a few sales pitches dressed as keynotes. The excitement was genuine—talk of "agentic buying" systems that could revolutionize how media gets purchased, promises of AI transforming creative workflows, glimpses of a transformed advertising ecosystem. Off stage: heated debates about the AdTech war brewing between publishers, SSPs, and DSPs, all fighting over their place in this new order. Conversations about the open web's uncertain future. Questions about what happens to traditional business models when AI can do in seconds what took teams weeks. It was a great event, full of amazing people but on AI at least it was very short of detail.

Then I arrived in Cleveland for MAOICON. The Marketing AI Conference. I was concerned because no one at Adweek had heard of MAICON. "Why are you going to Cleveland?" was my most common question the previous week, normally with a confused and slightly pained expression.

I needn't have worried. MAICON drew 1,100 attendees in 2024 and expects 1,500 in 2025 —a fraction of Adweek's scale. Yet what I witnessed was the most knowledgeable AI audience I've ever encountered. Not people talking about AI. People doing AI. The conference featured sessions on real AI adoption, AI-driven web optimization, and advancing B2B GenAI from pilots to proficiency.

I must have seen 1,000 different agents, and I'm only halfway through my notes. Agents for content creation, campaign optimization, customer segmentation, creative testing, email personalization, data analysis, competitor monitoring. Some were sophisticated enterprise systems; others you could set up in an afternoon. These weren't future promises—they were production tools delivering ROI today.

Two conferences - two worlds. MAICON was everything Adweek wasn't, embracing detail, looking forward to the new world with hope (and only a little bit of fear). Full of detail, full of ambition, full of humility as to what was still left to do.

Why Two Worlds?

MAICON reminded me of the digital conferences 20 years ago—the ones TV executives never bothered attending. Back then, the establishment dismissed digital as small-time, niche, not worthy of serious attention. One part derision, one part terror.

It's the same pattern now, except the pace is 10x faster.

Adweek represents scale, establishment, power—big agencies, legacy brands, established media companies deeply invested in existing systems. Different tribes, different focus, different information networks. MAICON is the frontline: attendance growing from the 300s in its first three years to 700 in 2023. Practitioners who need solutions now, testing and iterating in real-time.

But here's what I'm convinced of after experiencing both: most people reading this are underestimating how fast change is coming.

The Tracks Are Laid

Yes, there may be a financial bubble in AI valuations. But the pace of change won't slow. It may even accelerate.

Why? The infrastructure is already built. Claude Sonnet 4.5 can maintain focus for more than 30 hours on complex, multi-step tasks Introducing Claude Sonnet 4.5 \ Anthropic. Developers are building full SaaS products with AI-generated code in under two weeks without writing a line of code themselves How I built a SaaS product with 100% AI generated code | by Sakky B | Medium. The hundreds of billions already invested aren't going away—they're infrastructure now.

A financial correction might actually accelerate adoption. When money is tight, companies are forced to choose. And increasingly, the choice is obvious: AI-native approaches are faster, cheaper, and more effective.

What IP Is Defensible?

Here's the uncomfortable truth we've recently witnessed: content isn't sacred. AI models have been trained on vast swathes of human creativity—articles, books, code, images, music—often without compensation or consent. Should content be sacred? Absolutely. We believe it should be. But the reality has shown otherwise.

And if content isn't sacred, what else isn't? What IP is truly defensible anymore?

Code: AI can build a SaaS product in under two weeks How I built a SaaS product with 100% AI generated code | by Sakky B | Medium Data: Gets scraped, synthesized, or recreated. This may not be right but right now its harder than ever to both monetise and protect your data simultaneously. Infrastructure: Even networks like Uber's driver base could be replicated by agents talking directly to open source applications. Patents: Protect implementations, not capabilities—and AI finds new implementations faster than lawyers file paperwork

The money involved, the potential returns, the velocity of development—when you can build in weeks what used to take years, what exactly is defensible?

The only moat left: velocity. You don't need to outrun the bear—just outrun everyone else.

The Pace of Change

The pace of change is going to beat anything we have yet seen for three big reasons.

1) Capability. AI means that you can build faster. The constraints of labour no longer apply. Agents can develop whilst your team sleeps.

2) Opportunity. Those that do build fastest will get the best returns. We are seeing in marketing so many land grabs to see who controls the future operating model. Capital is going to flow to those areas where the opportunity exists. That's everywhere where the pace is high.

3) Necessity. Suddenly we are in a situation where moving quickly isn't a luxury any more but a necessity of survival. When you have that necessity everything else follows.

What To Do Monday Morning

The gap between these worlds is closing faster than the gap between digital and TV ever did. Twenty years ago, traditional media had a decade to adapt. Today's establishment has maybe two years, possibly less.

If you're a CMO or marketing leader reading this, here's where to start:

Ask each of your reports for their 3 biggest time sucks or ROI opportunities and an AI plan to resolve them (give them a few weeks). Not theoretical—actual plans with actual tools. Then what Appoint an AI transformation leader internally. Someone to listen to the podcasts, meet the vendors and develop plans. Train your team: Everyone should be able to prompt. Everyone should be able to build a custom GPT. Technical roles should be learning Claude Code, Cursor, or similar tools. This is the biggest unlock to productivity in history. Hire people who are leaning in with eyes open to the future. Let go of fighting the last war. Crucially. Develop a plan for where your business and competitive might be in 1, 2 ,3 years. Have a bias for believing that change is more than you think.

Two Worlds Colliding

AI is becoming infrastructure. You could do almost everything in Google before—search, email, docs, maps, shopping. Now you can do almost everything in Perplexity or Claude. Search, analyse, create, code, shop, research, interact.

Every company needs to answer: What's our role when everyone has at least one AI agent at their disposal?

The future isn't evenly distributed yet. But it's spreading at exponential speed. The applications are multiplying daily. Adoption is inevitable.

The wave is here. Time to ride it.