July 2026

II. The New Writing Work Behind Better AI

The evolution of AI might be taking away some of the lowest-level work for writers, but it’s also creating new opportunities. And a big new specialty area is opening up that would be perfect for commercial writers:

Context Management

What do I mean by this? Earlier (think 2023ish), “prompt engineering” was all the rage. How you phrased an AI prompt could significantly impact the results you would receive.

Prompt engineering isn’t gone, but it isn’t nearly as important as it once was. Now, there’s more of a focus on context than specific prompts. And there have been major changes in how these tools and models work even just over the past few months.

Online AI chat interfaces can set up “projects” with their own stable set of instructions and source documents. They can retain memories across chats within those projects.

But the real commercial writing opportunity comes from tools like Codex (OpenAI) and Claude Code / Cowork (Anthropic), and the more recent introduction of AI operating systems (sometimes called AIOS, agent operating systems, or LLM wikis).

What Are AI Operating Systems (& Why Should Writers Care)?

An AI operating system is basically a file system where companies, or individuals, can keep a wide variety of context files their AI tools can use collectively.

In other words, you don’t have to create separate instructions, backgrounders, etc. for each tool. When you’re using these desktop tools (or even local LLMs), they can all pull from what’s called your “single source of truth.”

They follow the same rules, interact with you in similar ways, and should produce more consistent results.

These are not just small batches of files.

For example, I built one for my business that I use with Codex, Cowork, and Gemma 4 (my local LLM of choice — private, local, no data centers involved, nothing sent to the cloud).

My AIOS keeps consistent background on my company, the divisions of it (client services, digital publishing, and book publishing), every individual property I own, their audiences, specific content rules or guidelines for each, etc. My AIOS also doubles as an Obsidian vault for markdown editing, file linking, and more. And it stores outputs like updated idea logs, project outlines, and research material (which you can add yourself but keep it available for AI searching and references).

The AIOS is also where your skills files live. These are the specialized skills your agents use.

For instance, I have agent teams using skills related to everything from sourcing information (like finding job leads for me to review for the curated listings on my job board) to creating specific types of reports when I request them, like weekly SEO reports combining GA4, Google Search Console, and Ahrefs data.

You can have skill files related to specific types of copywriting projects, checking your own drafts against style guidelines (from standard ones to client style guides), and so much more.

Every piece of context is basically a markdown file. So are the skills. And they all need to be not only written, but maintained.

Can’t AI Tools Create Their Own Context Files?

Yes, they can. But…

• they can add inferences that are incorrect;
• they can add dated information from past chat memories that you don’t want influencing future output;
• there are rules they won’t always think to add during an initial AIOS build (like separate validation layers, cleaning up their own run logs and outdated outputs you don’t need, which can really add up).

Also, things change. You might decide to tweak rules after seeing poor results, and the system won’t know to update the AIOS unless you remember to tell it to.

The Market Opportunity

A commercial writer who understands how to use these tools will be in a prime position to help companies and organizations build out their own context systems to get the most out of their AI tools.

This can become an ongoing gig. These systems need periodic audits, manual checks of the most important files, occasional clean-up of auto-saved memories that shouldn’t be applied moving forward.

Writers who already know how to write executive bios, company backgrounders, mission statements, positioning copy, project briefs… they’re in a position to write, edit, or otherwise improve this vital context layer more and more companies are integrating into their AI use.

Companies need this not only because they want better results, but because better context can lead to better optimization and lower AI token usage. You know… a direct, measurable benefit that can save companies a lot of money.

III. A Copywriter’s Eye Never Clocks Out

Understanding copywriting and consumer psychology can just be part of the job as a commercial writer. 

But as a consumer, those skills can sometimes feel like a superpower. 

You can read between the lines when you see ad copy. You might pick up on meaning from what’s not said in a commercial or presentation. It can make you look downright psychic to those who don’t see the same things. 

I was reminded of this recently with design software…

If you’re familiar with design software, you’re likely aware of Adobe. And you probably also know about Canva. 

Well, in 2024, Canva acquired the Affinity software suite from Serif. And they later announced they would release a new version.

While I’m sure they thought they were being cute or clever with their teaser announcement, there was a lot of initial backlash in the design community. 

The teaser focused on the phrase “true creative freedom.”

Immediately, I knew where they were going.

I could feel it from the copy and energy behind the messaging. I understood the concept and value of a loss leader. I knew how Affinity was being marketed to students differently even before V3 was announced and the value in accessibility in this particular market. And I knew the biggest pain points customers of their biggest competitor frequently had.

But many designers jumped to the conclusion that Canva would move Affinity from a one-off purchase to a subscription model. 

That didn’t happen.

When Affinity 3 was finally launched, they made the entire suite free. As-expected.

While you do activate by connecting it to your Canva account, you can do this with a free account. Being a premium Canva user just gives you some Canva AI features that would be new to Affinity anyway.

If you’re interested in learning more about graphic or document design, I highly recommend trying Affinity 3.

The point of the story isn’t about Canva or Affinity though. It’s about the way working in commercial writing can train you to take in and process information differently.

When you write to influence consumers, you become a better-informed consumer yourself. You will fine-tune your internal BS detector. You’ll avoid unnecessary dread after announcements like these. And occasionally, you’ll even get a good chuckle out of it.