
Playbook
Tasteful Skills
“Tasteful Skills” argues that the best agent skills are not documentation or best-practice lists.
The internet is shifting from an attention economy to an interpretation economy — Nate B Jones argues that for the first time in 25 years, the key question is no longer “how do I get attention?” but “how does AI interpret me, my product, or my company when someone asks if they should trust us?”
AI is already replacing big chunks of the buyer journey — he says his recent sound system purchase was driven almost entirely by Claude and ChatGPT, where he fed in room dimensions, budget, and sound preferences and got recommendations without marketers meaningfully shaping the decision.
Marketers need a 'truth layer,' not just better copy — brands must publish structured, provable, high-fidelity product data that agents can read, map to customer intent, and trust, or else they'll get flattened into the generic average of their category.
For workers, the same rule applies: prove it or disappear into the noise — Nate ties this to his Talent Board project, arguing that AI-era hiring depends less on polished LinkedIn posts and more on evidence that you can actually build things like MLOps or agent pipelines.
Brand still matters, but now as prompt-seeding and memory creation — offline events, trust, and emotional resonance become more valuable because if a person remembers you by name, they constrain the AI’s choices instead of asking for a wide-open comparison.
Back-office AI automation is table stakes, not strategy — automating resumes, cover letters, content, or workflows may be necessary by 2026, but he says the real leverage is making yourself memorable to humans and legible to agents at the same time.
Nate opens with a blunt reframing: the web has run on attention for 25 years, but now we're entering an “interpretation economy,” where people ask AI whether they should trust you. He makes it personal on purpose — not just your company or product, but you as an individual candidate, worker, or builder being filtered through an AI opinion layer.
To make it real, he walks through buying a sound system by chatting with Claude and ChatGPT instead of browsing brand marketing. He gave the models room dimensions, budget, and even whether he prefers “warm” or “cool” sound, and got a buying experience he felt was better than what marketers had designed — even though, as he admits, he still doesn’t know if it was truly the best option, only the one AI made seem best.
He takes aim at the common move of using ChatGPT, Claude, or tools like CoWork to automate internal marketing work and calls that “table stakes in 2026.” The same goes for job seekers using AI to customize resumes and cover letters: useful, yes, but not where the leverage is when the real game is how agents interpret what you are.
His core prescription for companies is a “truth layer” — clear, detailed, trustworthy product data that agents can retrieve and use. He uses a running shoe example: if all you have is emotional language about performance, an agent will flatten you into “just another shoe seller,” but if you can prove specifics like materials, spring systems, and reduced impact on knees in readable formats like structured pages or JSON schema, you stay in the consideration set.
Nate says the same logic governs hiring, which is why he references his Talent Board project. In his framing, AI hiring markets reward demonstrated capability — can you stand up an agent development pipeline, build MLOps, or communicate intent clearly enough to ship production code as a PM or designer — not just polished LinkedIn self-branding.
He says there are basically two ways purchases happen now: either the AI interprets the market for you, or the buyer comes in with such strong brand loyalty that they ask for a product by name. That’s where offline events matter: they “seed prompts” so later someone remembers Nate, or a coffee brand, and asks the AI about that specific person or company instead of the whole field.
Nate pushes back hard on the idea that AI makes human-facing brand work irrelevant. His line is memorable: “Human memory becomes more precious as more of the transaction is mediated,” because the agent may compare options, but the human still supplies the preference — and if your brand story and your agent-readable reality don’t match, you weaken yourself in both directions.
He closes with a warning against AI washing for both companies and individuals under pressure to sound “AI native.” In a two-internet economy, the winners are the ones with an honest wedge: memorable to humans, legible to agents, opinionated enough to “survive compression,” and specific enough that they don’t get averaged out into generic AI slop.
Share
Keep Reading
The Weekly Echo. The inbox-shaped summary of what mattered.
New editorials announced here.

Playbook
“Tasteful Skills” argues that the best agent skills are not documentation or best-practice lists.

Playbook
Learn how tasteful prompting helps you move beyond generic AI output by shaping context, style, and judgment from the start.

Playbook
OpenAI shipped /goal for the Codex CLI. It turns a prompt into a persisted, self-continuing contract.