Science fiction author Cory Doctorow, who in 2022 coined the phrase enshittification, has a new book out, The Reverse Centaur's Guide to Life After AI, in which he argues that when (not if) the AI bubble bursts, employers will spend years trying to recover the process knowledge they let AI replace. He compares it to the Mad Max dystopia of Fury Road. Stuff you still have but no longer know how to make.



For many industries, that's a bleak outlook. For communications, I don't think it lands in the same way because it points at what's actually being lost rather than just declaring the discipline finished. Here's my bastardisation of Doctorow's earlier concept to explain why.

Enshittification, at its core, describes value getting quietly siphoned off a product until the person it was originally built for is the last one being served. Communications from an advisory perspective has run its own version of that trade over the last decade or more. A chunk of the industry stopped selling judgement and started selling volume, because volume was easier to price, easier to scale, and easier to bill by the hour than genuine strategic thinking and outcome-focused execution. Clients got cookie cutter. Agencies got margin. Nobody was maximising for the work actually being good.

So now when I look at what AI is actually displacing inside communications, I don't see deep expertise being hollowed out. I see that volume layer being exposed. The rewrite of a rewrite. The press release built off a template nobody's questioned since 2017. The stakeholder map that's really just a spreadsheet with some ad hoc colour-coding on it. None of it reflects a body of knowledge worth mourning.

The impact of AI on communications isn't leading to the jobs tsunami some are predicting, presumed to hit every white-collar function equally. The wave is less egalitarian than that, and the communications work getting automated first is rightly that which was already commoditised and cheap to produce.

However, what AI can't touch is the stuff our industry has always struggled to effectively sell upstream. Crisis judgement built over decades of watching things go wrong in real time and knowing which lever to pull in response. Diverse stakeholder engagement that depends on cultural fluency and connection. Take for example the social licence work data centre operators must urgently initiate right now to keep expansion plans on track — the gap between community sentiment and corporate assumption is exactly the gap a machine can't close. You can brief AI on community backlash. You can't send it to a town hall to read the room.

So, Doctorow's vision of a post-AI bubble world needs adjustment for our industry. His theory assumes real knowledge gets replaced. In a lot of white-collar work — law, maybe — that's a fair and worrying assumption. In communications, plenty of what's being automated was never expertise to begin with. It was just dressed up as such and going to be exposed eventually. AI has just sped up the timeline and deshittified along the way.

For the people actually doing the work, that's good news, not bad, so long as you're not living off the volume layer. Advisors who understand their clients' commercial reality, who can calmly sit inside a crisis and navigate it without continually referring to an AI-generated script, who read a stakeholder environment as a messy human ecosystem rather than a spreadsheet: that group, and its skill base, will get more valuable, not less. Clients are already noticing who was thinking and who was just producing.

The same logic holds for in-house teams. The teams that do well won't be the ones chasing curiosity at the expense of the process discipline the business runs on. They'll be the ones doing both, treating AI as something to test against well-built protocols rather than something that replaces them. The teams getting this right are running small experiments inside real guardrails and learning where the tool actually helps, while leaving the load-bearing knowledge intact.

Most communications professionals aren't short on expertise, and in a communications era where knowledge and judgement reign supreme again, I think we'll also see new shoots of growth in response — the rise of personal authority in place of synthesised thought leadership, for one.

Because as AI-generated thought leadership floods every feed, people are getting a nose for the human versus the machine output and actively crying out for real perspective instead. For the next generation of communications leaders, the sharpest advantage for the stakeholders they serve will be personal authority: a point of view derived from real-world experience that they can defend in a room, credibility built on a track record other people will vouch for, trust that compounds because it's attached to a specific person rather than a department. That's what buyers are looking for, both inside and outside of generative search.

Doctorow's post AI-bubble warning still holds. But for communications, the uncomfortable truth isn't that AI is going to strip away hard-won knowledge. It's that AI is already making it much harder to hide the fact that a lot of what we called knowledge never quite was in the first place.