A primitive is a unit of technological capability. A product is a unit of value. They are not the same, and the gap between them is where most products fail.

I learned this on Amazon's multimedia systems. We had engines that generated virtual worlds, millions of photoreal 3D models, image generation at millions of images for cents. None of it was a product. Creators and brands didn't want science experiments. They wanted tools that finished a job - design, edit, and publish at scale without needing weeks, a PhD, or a Hollywood budget. The work was the translation: turning capability into value.

Why the gap matters more now.

In any prior platform shift, primitives were expensive enough that building one was the work. A new database engine, a mobile SDK, a cloud API - possessing it was an advantage. Translation into a product mattered, but the queue for primitives was long, and most teams were fighting the backlog.

AI has emptied that queue. Building is cheap; choosing what to build costs what it always did, maybe more. When primitives are abundant and cheap, the primary constraint left is judgment. AI slop is what happens when teams skip it.

This is not only my experience. In March 2026, Anthropic published a study of AI's labour-market impact comparing two things for every occupation: tasks a model could theoretically do, and tasks actually being done with AI.

The blue is capability. The red is capability translated into use. In computer and mathematical work, the most exposed category, models could touch 94% of tasks; real coverage sits near a third.

Anthropic reads this as evidence that AI has not yet displaced work at scale. Read another way, it is a map of untranslated capability. Some of the gap will not close. Some tasks are beyond the reach of a model or its data, others held back by law or trust. The rest is untranslated: the capability exists, and no one has turned it into something someone in that job can reliably use.

Take the most exposed category and watch the gap appear inside it. Coding is a language-forward problem and LLMs handle it well. A May 2026 study of 100,000+ GitHub developers (Demirer et al.) found autonomous coding agents raised activity (commits) by 180%, but those gains drop off, falling to 50% for projects and 30% for actual releases. The spike in apps doesn't translate into more downloads, users, or engagement. The authors call it a weak-link problem: output is capped by the slowest step, and the slowest step is still human - review, integration, and above all, translation.

The primitives.

The taxonomy I've drawn below names the units this new capability comes in: models, prompts, tools, skills, agents, and the scaffolding between them. Anthropic's material and the OpenAI cookbook are great resources to delve deeper.

The point is structural: these are atoms; a product is a molecule. A prompt is not a product. A model is not a product. A tool, a skill, an agent - none of these is a product. They are what products are made from, and this essay is not about the atoms, but how they come together.

The translation.

Translating a primitive into a product has five moves. They are not stages to complete, but questions the PM has to keep open for the life of the product. With primitives mutating faster than they used to, translation is not a one-time act. It is the continuous work of holding a moving primitive against a moving workflow and keeping the product still.

The first is inventory. Understanding the full potential of the primitive, what it does, how it fails, what it costs at scale, the workflows it composes into. An agent that handles each step with 95% reliability finishes a five-step task about 77% of the time and a ten-step task about 60%; errors compound. Same agent, different primitive depending on the workflow you put it inside.

The second is anchoring. A primitive does not become a product until it is attached to a journey someone is already trying to complete. The brand manager doesn't want a generative model. She wants a product photograph with the specified lighting and white background her catalogue update needs, in the four hours it allows.

The third is composition. Most outcomes need more than one primitive. The single AI image a creator publishes is the output of a chain: a template feeds a generation model, an edit pass feeds an editing model, a grounding pass keeps it photoreal. The product is the composition with primitives as inputs.

The fourth is control. This is where translation becomes craft. Power without a control surface overwhelms; after the first failed attempt, the user will not return. The right controls don’t constrain capability; they translate it into the user’s needs - the editing layers, the templates, the single button that compresses a workflow into one click. Control builds trust, and a reason to return.

The fifth is the boundary. The deterministic-probabilistic line is the most consequential translation decision in the product. Which steps of the workflow do we let the model decide, and which do we lock down? Customers tolerate variance in the suggestion. They do not tolerate it in the checkout, the price, or their privacy.

What endures.

One of the most useful things I learned about primitives, I learned from one that could not be translated. An engineer on my team built a renderer that could place millions of light sources in a single scene, where most engines stop in the low thousands. No creator wanted a million lights; they wanted four, well placed, with presets and a way to move them. The capability was real. No workflow, then or now, could turn it into something a customer would reach for.

Some principles hold across cycles. The same gap existed when SQL became a primitive, when the browser did, when the mobile SDK did. The primitives change every cycle. The translator stays the same, standing between the customer and technology, refusing to mistake one for the other.

There is another way to read Anthropic's chart. Who gets the value of a primitive depends on who translates it. Until someone does, the capability pools with the technically fluent. The renderer proved it at small scale: as long as the virtual lights had no presets, that power belonged to people who could speak to a renderer. The brand manager had the need, the taste, and the deadline, but not the access. The blue-red gap is that renderer, drawn across every occupation on the chart: not just untranslated capability, but a map of who is locked out.

Sometimes the start is the capability collapsed into one obvious move like removing a background. Sometimes it is invisible: ChatGPT and Claude are not blank text boxes over a model; behind the box sits a system prompt, a tuned model, suggested openings, and tools that fire on their own. Sometimes it is the context, a frame: the primitive set inside a problem someone already has, which is where technology becomes value. The customer knows the problem but not the capability, the engineer knows the capability but not the problem, and neither speaks the other's language. Four lights, well placed, with presets, and a way to move them. That is the whole of it.

Where translation becomes the artefact.

With AI, the translation stops being implicit and becomes a thing you can point to. The harness, the scaffolding of prompts, tools, retrieval, and controls around the model, holds it. The eval, the written test of what good means here, is how you know it works. The boundary now a core part of the product.

This is why the moat in AI is moving where it is. As foundation models commoditise, every team has access to roughly the same primitives; Cursor and a hundred forgettable IDE plug-ins are built on the same handful of frontier models. What separates them is the translation: the thinking behind the product, held in the harness, proven by the eval.

When the translation is right, the technology disappears into the background of what someone is making or doing. What stays visible is the judgment. The translation is the product.

Next: Of Middles and Edges. The translator has a place to stand: as AI takes the middle of the work, the judgement concentrates at the edges.