You know, Herman, I was walking through the city earlier, and I saw three different posters for local businesses—a law firm, a boutique marketing agency, and even a high-end real estate group—and every single one of them had that telltale AI sheen. You know the one? The slightly too perfect lighting, the skin that looks like it was buffed with digital sandpaper, and that extra finger on a hand tucked just a little too deep into a pocket. It got me thinking about that initial gold rush we saw a couple of years back, right around twenty twenty-four.
Oh, the great AI-washing of twenty twenty-four and twenty twenty-five. I remember it well, Corn. It was a fever dream. Everyone was a prompt engineer suddenly, and every single software-as-a-service app was basically just a thin wrapper for a large language model. It was the era of "add a chatbot and call it a pivot." We saw billions of dollars in venture capital flowing into companies that were essentially just paying an OpenAI subscription and putting a pretty skin on it.
Exactly. It was chaotic. And that's actually what our housemate Daniel was asking about in the prompt he sent us this morning. He wanted us to look at the professional services agency landscape specifically. Not just the big tech players like Microsoft or Google, but the PR firms, the SEO agencies, the digital transformation consultants. He's curious about how this whole AI explosion has actually reshaped the way these businesses operate, especially now that we're sitting here in February of twenty twenty-six, a few years into the actual revolution.
It's a fantastic question from Daniel. Herman Poppleberry, at your service, and I've actually been digging into some of the industry data from the last eighteen months. It's fascinating because the agency world is really the front line of this. They're the ones who have to prove to clients that they can do things faster, better, and cheaper using these tools, or they risk being replaced by the tools themselves. In twenty twenty-four, the fear was "Will AI replace the agency?" In twenty twenty-six, the reality is "The agency that uses AI is replacing the agency that doesn't."
Right, because if I'm a client and I see that a PR firm is just using a basic model to write my press releases—and I can see the "delve" and "tapestry" and "testament" in the copy—I'm going to start wondering why I'm paying them a five-thousand-dollar monthly retainer. I could do that myself for twenty bucks a month.
Precisely. The value proposition has shifted from the labor of creation to the expertise of curation and strategy. But before we get too deep into the weeds, we should probably frame just how much has changed. If you look back to early twenty twenty-four, most agencies were in what I call the "panic-adoption phase." They were just trying to figure out what a GPT could do. By twenty twenty-five, we hit the "integration phase," where they started putting it into their workflows. Now, in twenty twenty-six, we've moved into the "architectural phase." It's not about using a chatbot anymore; it's about building custom agentic workflows into the very foundation of the business. We’re talking about autonomous agents that don’t just write text, but actually execute multi-step projects with minimal human oversight.
And that's where Daniel’s point about the emergence of AI-first agencies comes in. We’ve seen these new players pop up—agencies started in twenty twenty-four or twenty twenty-five—that don’t have the baggage of the old-school model. They don't have massive offices in Manhattan or hundreds of junior copywriters. But are they actually winning? Are the big, established agencies suffering, or have they managed to pivot?
It’s a bit of both, but the data suggests a really interesting bifurcation, or a "splitting of the road." If you look at the mid-market agencies—the ones with maybe fifty to two hundred employees—they’ve actually been hit the hardest. They have enough overhead that they need high margins to survive, but they aren't big enough to have the massive research and development budgets of the global conglomerates. They’re stuck in the middle. Meanwhile, we're seeing the rise of the "nano-agency."
The nano-agency? I like that term. It sounds like something out of a cyberpunk novel. What does that look like in practice?
It’s basically a solo consultant or a team of two or three people who are leveraging an entire fleet of AI agents. In twenty twenty-four, a solo SEO consultant could handle maybe five to ten clients comfortably if they were working around the clock. Today, with sophisticated agentic workflows for keyword research, content auditing, and backlink analysis, that same person can manage forty or fifty clients without breaking a sweat. And because their overhead is basically zero—no office, no health insurance for fifty people, no middle management—they can undercut the traditional agencies on price by sixty percent while providing a more data-driven service. According to a recent report from the Agency Strategy Group, these nano-agencies have captured nearly fifteen percent of the small-business marketing spend in the last year alone.
So the smaller players are actually becoming more powerful because the technology is a force multiplier. It’s the "Iron Man suit" for the solo entrepreneur. But what about the big guys? You’d think the huge firms like WPP or Publicis would be slow to move, but they have so much data.
You hit the nail on the head. The massive agencies have one thing that the solo consultants don't: proprietary data. Over the last two years, we've seen the giants stop trying to use public models like the basic ChatGPT or Claude and start building their own internal ecosystems. They’re taking decades of successful campaign data, client interactions, and performance metrics and feeding them into private, fine-tuned models. So, if you're a Fortune five hundred company, you're not just buying a marketing strategy; you're buying a strategy informed by a model that has seen every successful campaign in your sector for the last thirty years. Publicis, for example, invested over three hundred million dollars into their "CoreAI" platform. They aren't just using AI; they are becoming AI companies that happen to do marketing.
That makes sense. It's the moat. If everyone has access to the same public AI, the only way to win is to have better data or better taste.
Taste is the big one, Corn. I'm glad you mentioned that. One of the things Daniel brought up was the skepticism surrounding poorly thought-out AI features. We've seen a lot of that in the PR and content world. For a while, the internet was just flooded with mediocre, AI-generated junk. It was the "slop" era. And what happened? The value of truly high-end, human-curated content actually skyrocketed. It’s the "Human Premium."
I've noticed that in the last year. Some of the most successful agencies are actually marketing themselves as "AI-augmented but human-led." They’re almost leaning into the skepticism. They’re saying, "We use AI to do the boring stuff—the data crunching, the initial drafts, the scheduling—so our smartest people can spend ten hours thinking about your brand's soul instead of ten hours writing basic blog posts." It’s a return to the "Mad Men" era of big ideas, but with a supercomputer in the basement.
Exactly. There was a report recently from an industry analyst group that found that while pure AI-generated content volume is up by four hundred percent since twenty twenty-four, the engagement on that content has dropped by nearly sixty percent. People have developed an internal filter for it. They can tell when something was just spat out by a machine without any genuine insight or unique perspective. We’ve reached "Peak AI Content," and now the pendulum is swinging back toward authenticity.
So, if you’re an agency in twenty twenty-six, and you’re just selling the output of an AI, you’re essentially a commodity. You’re in a race to the bottom on price.
Precisely. The growth areas Daniel asked about are really in three specific sectors: AI Strategy and Implementation, Output Auditing, and what I call Synthetic Research.
Let’s break those down because those sound like the new pillars of the industry. Start with AI Strategy. That sounds like the twenty twenty-six version of "digital transformation."
It is, but it’s much more granular. Two years ago, digital transformation meant getting your files into the cloud and using Slack. Today, AI Strategy means an agency coming into a legacy business—say, a mid-sized manufacturing firm—and saying, "Okay, your customer service department has twenty people. We can implement a RAG system—that’s Retrieval-Augmented Generation—that handles eighty percent of the queries by reading your own technical manuals. Then, we can retrain your ten best people to become AI supervisors." The agencies that can bridge the gap between technical possibility and business reality are making a killing. They aren't just consultants; they are "workflow architects."
And Output Auditing? Is that basically just fact-checking for robots?
It’s more than that. It’s brand safety, it’s legal compliance, and it’s quality control. As companies automate more of their external communications, the risk of a model "hallucinating" or saying something off-brand—or worse, something legally actionable—is huge. We’ve seen a whole new category of agency emerge that specializes in auditing the AI pipelines of other companies. They are the gatekeepers. They use their own adversarial AI to "attack" the client's AI to find weaknesses before the public does. It’s like a security audit, but for language and brand voice.
That’s a fascinating niche. It’s like being the editor-in-chief for a thousand robots. And what about Synthetic Research? That sounds like something straight out of a sci-fi novel, Herman. Are we talking about virtual focus groups?
Exactly. It’s actually one of the coolest developments in the agency world. Instead of doing traditional focus groups, which are slow, expensive, and often biased because people tell you what they think you want to hear, agencies are now building synthetic personas based on massive amounts of consumer data. They can create a "digital twin" of a thirty-five-year-old suburban dad who likes hiking and craft beer. Then, they run a thousand different versions of an ad through a synthetic audience of ten thousand of these digital twins to see which one resonates most. They do this before they ever spend a single dollar on actual media.
Wait, so they're testing ads on simulated people? Does that actually work?
It’s scarily accurate. Because these models are trained on the sum total of human internet discourse, they are very good at predicting how certain demographics will react to certain triggers. It allows for a level of iteration that was impossible three years ago. You can test a campaign in fifty different languages and cultures in an afternoon. The agency "Gartner-plus" reported that synthetic research has reduced product launch failures by nearly thirty percent for the agencies using it.
That’s incredible, but it also feels a little hollow. It feels like it could lead to a very homogenized culture if everyone is just optimizing for the same simulated preferences. If the AI says "People like the color blue," then every ad becomes blue.
That’s the danger. And that’s why the creative agencies that are willing to take risks and go against what the data says are often the ones winning awards and getting the big clients now. There’s a tension now between the efficiency of the machine and the unpredictability of human genius. We’re seeing a "Boutique Creative" revival where agencies explicitly state they don't use synthetic research for the final creative spark.
I want to go back to Daniel's question about the big agencies versus the small ones. We talked about the nano-agency and the global giants, but what about the traditional "billable hour"? That’s been the foundation of the agency world for a century. How does that survive when an AI can do a week’s worth of work in ten seconds?
It doesn’t. The billable hour is dying, and it’s a painful death for many firms. If you’re still billing by the hour in twenty twenty-six, you’re essentially punishing yourself for being efficient. If it takes you one hour to do what used to take forty, you can't bill for one hour and stay in business. We’ve seen a massive shift toward value-based pricing. Agencies are saying, "We don’t care if it took us ten minutes or ten hours; the value of this strategy to your business is fifty thousand dollars, and that’s what we’re charging."
That’s a huge cultural shift for the clients, too. They used to feel like they were paying for "butts in seats." Now they have to pay for "outcomes." That requires a lot of trust.
It does, and that’s why relationships matter more than ever. In an age where the technical work is increasingly automated, the agency-client relationship becomes the primary product. You’re paying for the partner who understands your business, your fears, and your goals. You’re paying for the "Human in the Loop" who can tell you why the AI suggested a certain path. You’re not just paying for a deliverable; you're paying for accountability. If the AI messes up, you can't fire the AI. You fire the agency.
It’s almost ironic. The more we automate, the more the human connection becomes the premium. But I’m curious about the actual adoption within these agencies. Is it just the young, tech-savvy firms, or are the old-school PR veterans actually using these tools too?
It was a struggle at first. There was a lot of resistance from senior leadership who didn't want to admit that a twenty-two-year-old with a good set of agentic prompts could do the work of a director with twenty years of experience. But the market forced their hand. If you didn't adapt, you simply couldn't compete on speed. I read a study last month that showed that ninety-two percent of agencies now use generative AI in some capacity for their daily operations. The ones that don't are mostly boutique firms that specialize in "ultra-high-end, artisanal services."
Like the "slow-food movement," but for marketing.
Exactly. Hand-crafted, artisanal strategy. And there is a market for that! But for the vast majority of professional services, AI is now as standard as email. If you tell a client you don't use AI, they don't think you're "authentic"—they think you're inefficient.
So, if we look at the sectors Daniel mentioned—SEO, PR, digital transformation—where has the change been the most radical? I'd imagine SEO has been completely turned upside down.
Oh, SEO is a battlefield. With the rise of AI-powered search engines—what we call "Answer Engines"—the old game of just ranking for keywords is over. People don't click on "ten blue links" as much anymore. They get a synthesized answer from the AI. So, agencies have had to pivot from Search Engine Optimization to "Answer Engine Optimization" or AEO. It’s about authority and being part of the training data or the retrieval set. You want the AI to cite you as the source.
That sounds much harder than just stuffing some keywords into a meta tag.
It is! You can't just game the system anymore. You actually have to have the best, most comprehensive information. In a way, AI has forced SEO agencies to actually become better at their jobs. They have to produce real value now, because the AI is very good at filtering out low-quality "content farm" fluff.
And PR? I’ve seen some pretty bad AI-generated pitches in my time. I imagine journalists are losing their minds.
PR has had a rough transition. Because it’s so easy to generate pitches, journalists were getting flooded with thousands of AI-written emails every day. It nearly broke the industry in twenty twenty-five. The agencies that survived are the ones that went back to basics: real relationships. They use AI to research the journalist’s interests and past work—which is great for personalization—but the actual outreach is deeply personal. If a journalist even smells a hint of a template, it goes straight to the trash. The "spray and pray" model of PR is officially dead.
It’s like the AI has acted as a filter for mediocrity. It’s made the lazy version of every job obsolete, but it’s empowered the people who are actually good at the core of what they do.
That’s a perfect way to put it, Corn. It’s a talent amplifier, but it’s a mediocrity killer. If your only value was that you knew how to follow a process, you're in trouble. If your value is that you know how to think, you've never been more powerful. The "Agency of the Future" is a small group of highly creative thinkers managing a massive infrastructure of autonomous agents.
So, looking ahead, what’s the next phase? If the last two years were about integration and the emergence of these new agency models, what does the landscape look like in, say, twenty twenty-eight?
I think we’re going to see the total disappearance of the term "AI agency." It will be like saying you’re an "internet agency" today. It will be so foundational that it’s not worth mentioning. But more importantly, I think we’ll see the rise of "Autonomous Agencies." These are small, highly specialized groups that operate with almost no traditional staff, using sophisticated AI agents to handle everything from lead generation to project management to execution. We might see a billion-dollar agency with only ten human employees.
That’s a bit scary for the job market, isn't it? What happens to all the people who used to do those entry-level roles?
It’s a massive disruption. The entry-level roles—the junior copywriters, the research assistants, the data entry clerks—those jobs are largely gone. The challenge for the industry is how to train the next generation of experts when the traditional training ground has been automated away. How do you become a senior strategist if you never spent three years as a junior doing the grunt work?
That’s a point we don’t talk about enough. If you don't have juniors doing the grunt work, how do they ever learn enough to develop the "taste" and the "strategy" that we say is the human premium?
That’s the million-dollar question. Some agencies are starting to experiment with "AI-shadowing," where juniors work alongside the AI, auditing its work and learning the ropes that way. But it’s a very different career path than it was five years ago. It’s more like an apprenticeship in "AI Orchestration" than in "Copywriting" or "Design."
It sounds like the professional services world is becoming much more top-heavy. A few brilliant strategists at the top, supported by a massive layer of AI, and not much in the middle.
Exactly. It’s an hourglass shape, and that middle section is getting thinner every day. But for the clients, it’s actually a golden age. They can get world-class work done for a fraction of what it used to cost, provided they know how to find the right partners. The barrier to entry for starting a business has never been lower because you can "hire" an entire agency's worth of capability for the price of a few software subscriptions.
And that’s the trick, isn't it? In a world full of AI-washing and skepticism, how does a client tell the difference between an agency that’s actually using AI to innovate and one that’s just using it to cut corners?
You look at the results and the depth of the insight. If an agency presents you with a strategy that feels like it could have been written for any company in your industry, they’re just hitting a button. If they present you with something that feels deeply specific, weird, and human, then you know they’re using the tools correctly. The AI is the engine, but the human is still the driver.
"Weird and human." I like that. That’s probably a good mantra for all of us in twenty twenty-six. It's the only thing the machines can't quite replicate yet—that specific, idiosyncratic human spark.
It really is. The more the machines can do the logical, expected thing, the more the value of the weird and the unexpected goes up. In a world of perfect digital symmetry, the "extra finger" is a mistake, but the "unique perspective" is a gold mine.
Well, this has been a fascinating deep dive, Herman. I think Daniel really hit on something important here. The agency world is like a canary in the coal mine for how the rest of the professional world is going to change. Law, accounting, architecture—they're all following this same trajectory.
Absolutely. It’s where the rubber meets the road. If you want to see the future of work, just look at how a top-tier marketing agency is operating right now. They are the early adopters, the experimenters, and sometimes the cautionary tales.
Before we wrap up, I want to pivot to our listeners for a second. If you’ve been following our journey here on My Weird Prompts, we’d really appreciate it if you could take a moment to leave us a review on your podcast app or on Spotify. It genuinely helps other curious people find the show, and we love hearing what you think about these deep dives into the near-future.
Yeah, it really does make a difference. And if you have your own weird prompts—whether they're about the future of work, strange technology, or anything in between—you can always head over to myweirdprompts.com and use the contact form there. We read everything that comes in, and Daniel’s prompt today is a perfect example of the kind of stuff we love to sink our teeth into.
Definitely. We’re always looking for new angles to explore. Thanks for the prompt, Daniel—it definitely gave us a lot to chew on. I think I’m going to go look at those posters in the city again with a fresh set of eyes. I want to see if I can find the "human premium" in the wild.
Just don't count the fingers for too long, Corn. It’ll give you a headache. Focus on the soul of the ad, not the pixels.
Fair point. Alright, that’s it for this episode. You can find all our past episodes and our RSS feed at myweirdprompts.com.
And we’re on Spotify, too. Until next time, stay curious and keep it weird.
Thanks for listening to My Weird Prompts. We’ll see you in the next one. Goodbye!
Goodbye!
