I was speaking to a client recently. They got in touch, said they were interested in how we could help, and asked us to put a proposal together. We did. We took them through the opportunity, the thinking, the approach, and what it would take to deliver it properly. They agreed with all of it. Then we got to the investment piece and they said: “These seem like pre-AI costs.”

That line stuck with me because it sums up where a lot of people are right now. There is a growing assumption that AI has fundamentally reduced the cost of work across marketing, comms, growth, planning, strategy, events and experiences. And yes, in some areas, it has. But not in the neat, simple way people want to believe. AI has not just cut costs. It has shifted them.

There are absolutely parts of the process that should now be cheaper than they were before. First drafts. Repurposing. Summaries. Certain kinds of research. Basic production. Admin. Workflow tasks. That is real. That is useful. And anyone pretending otherwise is kidding themselves. But that is only one side of it. Because while some costs have come down, other parts of the work have become even more valuable.

Less on making. More on judging what is worth making.

Less on drafting. More on refinement.

Less on production. More on taste, direction and control.

That is the part I think a lot of people are missing. Faster output is not the same as better work. More content is not the same as stronger marketing. More options are not the same as better strategy. AI has made it easier than ever to produce things. It has not made it easier to produce things that matter.

In a lot of categories, it has simply flooded the market with more polished, competent, forgettable work. A cheap car flooding the market does not suddenly make it a better car. It just means there are more of them. A Bentley is still a Bentley. Not because it exists, but because of the thinking, engineering, judgement, craft and detail that sit behind it. The same applies here.

AI made content cheaper. It did not make great marketing cheap.

Once the cost of producing average work falls, the value of distinctiveness goes up. The same is true in comms. AI can help shape messaging, create variants, speed up internal workflows and remove manual effort. But the second the stakes are high, when something is sensitive, nuanced, political or emotionally loaded, the human layer matters even more. Because plausible is not the same as right. And polished is not the same as sharp.

It is also true in relationships, which is really what a lot of good marketing and comms are trying to build. You can automate messages. You can speed up content. You can produce more touchpoints. But judgement still matters. Timing still matters. Reading the room still matters. Knowing when to speak, what to say, what not to say, and how to build trust over time still matters. AI can help with the mechanics. It cannot replace the instinct required to create real connection.

Growth works in much the same way. AI can absolutely make the operating system around growth more efficient. Better targeting. Faster synthesis. More content. More automation. Better support around execution. But growth itself still depends on judgement. It depends on understanding what the buyer actually cares about, what story will land, what offer is strong, what difference is real, and what the market is ready to hear. AI can support that thinking. It cannot replace it.

Then you get into events and experiences, where this becomes even more obvious. Yes, AI can help with the mechanics around planning. Schedules. Draft agendas. Comms. Support tasks. All useful. But it does very little to replace the part that actually creates value in a live experience.

The admin may get cheaper. The craft does not.

It does not replace instinct. It does not replace atmosphere. It does not replace the tension in an idea, the precision in the execution, or the care needed to create something memorable. The same goes for communications done well. The best experiences and the best communications still have an artisanal quality to them. Not in a fluffy way. In the sense that they are shaped carefully, with intent, judgement and craft. They are not valuable because they exist. They are valuable because they have been thought through properly.

The same goes for strategy. AI can help you go wider and faster. It can help you structure, synthesise and explore. But strategy is not the production of options. It is the exercise of judgement. It is knowing which route matters, which truth has weight, which answer is commercially sound, and which one is just plausible nonsense dressed up well.

Execution is getting cheaper. Judgement is getting more valuable.

That is why I think the whole conversation around pre-AI versus post-AI costs needs to mature. The lazy version says AI should make all of this cheaper. The smarter version says AI should remove waste, compress time and reduce low-value labour, while increasing the importance of the work that actually creates differentiation. Not all cost is bad cost. Some cost is drag. Some cost is craft. Some cost is thinking. Some cost is the price of originality, conviction and getting it right.

The risk now is that businesses cut the wrong layer. They strip out the cost that looked expensive, then realise later they have also stripped out the thing that made the work effective. That is how you end up with cheaper work that performs worse. And there is going to be a lot of that.

The businesses that win will not be the ones that simply use AI to make more for less. They will be the ones that understand what should now be faster, what should now be cheaper, and what should be protected at all costs. Because post-AI, the value has moved. Output is easier. Coordination is lighter. Execution is faster. But judgement, taste, distinction and trust matter more than they did before.

That is the real shift.

And the companies that understand that early will not just save money. They will protect the part of the business that is still worth paying for.

‍