Not long ago, if you were writing something important, you were usually focused on getting one version right. Now, AI means the same piece turns into a collection: summary, talking points, Q&A, rewrites for different channels.
It sounds efficient and sometimes it is.
But the bottleneck on content was never drafting speed, so much of what we’re using AI for isn’t solving the right problem.
AI makes it cheaper to produce a piece of content but doesn’t make it any cheaper to decide whether the content is any good. Someone needs to work out what the message actually is, what can be cut and whether the piece should go out at all.
That is a content governance problem, not a technology problem. (Or as Dr Ian Malcolm of Jurassic Park fame said: “Your scientists were so preoccupied with whether they could, they didn’t stop to think if they should.”)
The work is shifting away from writing towards trying to make sure all the different versions say broadly the same thing — the specific point you started off trying to make easily gets diluted as you run out of time to challenge all the rewrites.
That’s part of the reason it feels like AI is intensifying work, not reducing it. It’s not just you. AI is significantly increasing content output from marketing teams, and AI-generated content now surpasses human-written content on the open internet.
The question for any communications team using AI is not how much you can generate, but whether it’s any good.
Once output gets past the point that you’re rushing to keep up, quality thins out. That’s why we get AI writing that looks fine but doesn’t really say much on closer inspection.
There’s no simple answer but a few habits can help.
Resist the temptation to create more formats just because they’re now quick to do.
Ensure an individual is responsible for the final call on every piece — if they aren’t keeping up, maybe you’re producing too much.
And when AI gives you six ideas, pick one or two and make them good.