AI cooked coders. Marketers are next.

Most of what we call marketing is a workaround for human limits.

People can't compare 300 options. They forget brands. They skip the fine print. They get tired, and they decide on vibes. Ads, funnels, SEO, retargeting, "only 3 left!" countdown timers, and most modern marketing or sales tactics exist to work around those limits, or to exploit them.

Now picture the buyer as an AI agent acting for a person. It has unlimited patience and perfect memory. It reads every spec, every review, every return policy. It doesn't see (or care about) your website, content, ads, or slick UI. It doesn't feel scarcity. It doesn't care that your logo is nice.

When buying moves to agents the whole persuasion layer loses its grip.

This week I want to sketch where that leads us. In predicting how AI will impact the marketing profession, we can’t assume steady, linear progress, because nothing about the last three years with AI has been linear. We’re in the exponential age. I'll go out 10, 20, and 30 years, then bring it back to how marketers, solo founders and small businesses can prepare.

For each era, I'm asking two questions: who (or what) makes the buying decision, and what moves them?

Agents handle routine buying. Reorders, travel, insurance, software, and commodity business purchases go to agents first. A procurement agent can run a 200-vendor comparison in minutes and negotiate with the vendors' own agents in seconds.

Your website becomes a backend. What matters is whether your offers, pricing, availability, terms, and proof are published in a structured form that agents can read and act on. Payment rails and checkout protocols built for agents are already appearing. If an agent can't read you, you don't exist to it.

The ad-funded internet cracks. Agents don't scroll feeds. The big fight of the decade is whether your agent works for you or for a platform that sells placement inside its recommendations. If platforms win, paid agent placement replaces search ads. If users win, marketing becomes earning the agent's trust with proof it can verify. My bet is on both, depending on the category.

Marketing gets adversarial. Some brands try to trick agents with hidden instructions on product pages, fake reviews written for AI readers, and poisoned data. Agents will get harder to fool. A trust layer grows up around attestations, provenance, and verified reputation. Claims you can't back up become worthless.

Teams shrink headcount 80%-90%. One strategist or CMO will direct swarms of agents that produce content, offers, and even product variations tailored to each customer (both humans and agents). Fixed list prices fade as price becomes something agents negotiate, fluctuating in real-time.

Sales changes shape. Cold outreach and appointment-setting roles are gone. Human closers survive for complex, high-stakes deals, where a person still signs and wants another person accountable on the other side.

This is where "marketing to agents" turns into something else entirely new.

Demand broadcasts and supply responds. Your agent publishes what you need and on what terms, and sellers' agents bid for it. Doc Searls called this "the intention economy" in his 2012 book. Superintelligent agents will make it real. The funnel runs backward.

Many sellers are AI. Largely autonomous companies design, make, price, and deliver with very few people involved. A large share of commerce is machine to machine.

Brand splits in two. It dies for commodities and gets stronger for identity. An optimizer has no reason to be loyal to a paper towel brand. But nobody fully hands off the purchases that say who they are: what they wear, what they eat, where they travel, which communities and causes they join.

Persuasion gets regulated like a weapon. An AI that understands your psychology better than you do is a manipulation engine. Expect persuasion audits, legal limits, and personal agents whose job includes shielding you from being sold to. AI-optimized persuasion aimed straight at people could end up as restricted as tobacco ads.

The marketing profession splits three ways:

Human dealmaking remains for scarce, high-stakes things: capital, land, talent, rare assets, partnerships. Agents negotiate everything else.

An honest caveat first: if we really are entering a technological singularity, 30 years out is honestly an unforecastable time period.

But one pattern seems sturdy. If most production is automated and goods get very cheap, value moves to what stays scarce:

Marketing goes back to what it was before mass production: myth, reputation, and culture. Shopping survives the way horseback riding survived the car, as a leisure experience rather than a full-time job or chore.

If your agent carries out your preferences perfectly, the only lever left is shaping the preferences themselves. Whoever influences what people want, not just what they buy, holds the real power. That makes marketing (in it’s future form) the most important profession of the era or the most dangerous one, depending on how it's governed.

People stay status-seeking, tribal, and driven by stories. Agents are stand-ins for human desire, but a human still has to shape and express that desire.

The mechanics of marketing are disposable. The psychology underneath is not.

What fades out:

What emerges:

None of this begins in 2036. It begins with the central theme of this newsletter, the context layer, which is the organized files that tell an AI who you are, what you offer, who it's for, and why anyone should believe you.

Today, that layer helps your own AI tools do better work. In the future, it's what other people's agents will read when they decide whether to recommend you, book you, or buy from you. A personal brand or small business with a clear, structured, provable context layer will sell to agents. One without it will be invisible, regardless of their brand aesthetics.

Three things to do this month:

The businesses that learn to be readable by machines now are the ones that will help write the trust rules later.

I hope this newsletter was helpful and educational.

See you next Wednesday,

-Ryan

Reply and tell me: How do you think marketing will evolved in the age of AI?