Camp 2 is the AI writer category. Jasper was the early flag-bearer. Copy.ai and Writesonic filled out the middle. These products gave a generation of marketers their first taste of AI-drafted copy and made the case that draft generation was a real time-saver.
Most of that demand has since migrated to the consumer chat tools. ChatGPT, Claude, and a few others now do most of the casual drafting work that Camp 2 was selling. The dedicated AI writer products survive in pockets, often by adding templates or specific industry framings, but the category as built has largely been absorbed.
This post is about why Camp 2, in either its original or consumer form, is not enough to run a marketing function. It is the second of the camp deep dives in our category map.
What AI writers solve
The blank page. The cursor blinking on Monday morning. The specific cognitive cost of going from zero to a draft. AI writers cut that cost to seconds. Type a prompt, get a draft, edit it, copy it out, paste it somewhere.
That is a real, repeatable saving. We are not going to pretend otherwise. Anyone who writes for a living has benefited from having a fast first-pass tool, and the AI writer category proved the use case at scale.
The structural flaw
The flaw shared by every Camp 2 product is simple: each session is a fresh start. The model has no memory of your brand, your past performance, your audience, your library, or what you wrote yesterday.
You can compensate for that by pasting context into every prompt. Many teams do. The maintained "brand prompt" living in a Notion doc, copied into a chat session, edited every week as the brand evolves. It works. It is also brittle. Every team member has their own version of the brand prompt, each version drifts, the consolidated voice document never quite catches up, and the resulting drafts read as a committee.
The output goes to the clipboard
The other structural fact about Camp 2 is that the output terminates in a text box. The draft is generated. The user copies it. The user pastes it into another tool. The other tool publishes.
This sounds minor. It is not. It is the seam that breaks every workflow built on top of Camp 2. The model that wrote the draft is not the model that scheduled it, is not the model that watched the performance, is not the model that wrote the follow-up next Tuesday. There is no continuity. There is no learning loop. The draft is an artefact, not a step in a system.
Where this leaves a marketing team
Teams that try to run their social on Camp 2 plus a scheduler find themselves in a specific pattern. Someone, usually the head of marketing or the founder, spends their week:
- Opening a chat session, pasting the brand prompt, asking for a draft.
- Editing the draft into the voice they actually wanted.
- Pasting the edited draft into the scheduler.
- Doing the same again for the next channel, with different formatting.
- Updating the brand prompt every few weeks based on what they wished they had told it.
Per post, this is maybe twenty minutes. Per channel per week, an hour or two. Per month, a meaningful chunk of the senior person's time. The savings versus writing from scratch are real. The savings versus an agentic system are not.
What is missing, in agentic-marketing terms
We map agentic marketing to four laws in the manifesto. Camp 2 tools satisfy zero of them.
- Memory over prompts: Camp 2 is prompts. The user supplies the memory every session.
- Channels are outputs: Camp 2 outputs to a clipboard. The user supplies the channel.
- Execution by default: Camp 2 does not execute. The user does.
- Brief gets sharper: Camp 2 forgets everything at session end. The brief, if it exists, lives in a Notion doc the user maintains by hand.
This is not a criticism of the tools. It is a category description. They are blank-page assistants. They were not built to be marketing systems.
When Camp 2 is still the right pick
Two honest cases:
You are a writer, not a marketing function.If you write for a living and need a draft accelerator, a consumer chat tool is excellent. It will not run your social for you, but you were not asking it to.
You have an idiosyncratic, one-off draft.A pitch letter. A landing-page hero block. A press release. Anything where the brand-fit matters less than the speed and the brand voice would slow you down. The blank-page tool is the right fit.
The trap
The trap is mistaking Camp 2 for a marketing system. We see this regularly. A team adopts an AI writer, sees the per-post savings, scales their reliance on it, and a year later they have a folder full of session transcripts, a brand prompt nobody owns, and drafts that read as off-brand more often than they should.
The trap is solvable. The solution is not a better AI writer. It is a system that holds the brand instead of asking you to retype it every Monday.
Where to go next
If you have outgrown Camp 2 and you want the work done end-to-end, the camp to look at is Camp 4. We unpack the mechanics in the Camp 4 deep dive, and the broader motion in the agentic marketing manifesto. The short version: the model needs to remember your brand, own the channels, and run on a schedule. That is a different product, not a faster typewriter.