The Elephant in the Room: Talking About AI May Actually Be Your Reason To Conveniently Look Away.

There is a meeting happening in your company right now about AI — sign of the times, can't be wrong to go with the flow. It has a budget, a slide deck, a vendor shortlist, perhaps a steering committee with a name that ends in "of Excellence." And there is another conversation — the one that quietly decides whether the company still has a reason to exist in five years — and it is on no one's agenda.

The loudest conversation is the one hiding the real one.

The AI convo actually masks out relevance.

We all fall so easily for the drama spiral: will AI take our jobs? Will it finally make us efficient? Should we govern it or unleash it, fear it or deploy it? These questions fill every conference, every feed, every offsite with a whiteboard and a catered lunch. They are not foolish questions. They are real.

But notice what they share. Every one of them keeps the subject of the sentence pointed outward — at the tool, the trend, the arriving force out there. Not one of them turns the sentence around to point at you: at what your company actually does, and at why a customer would still choose it next year over whatever AI is about to make possible for someone else. The conversation is engineered, almost perfectly, to be about everything except the one thing it keeps gesturing toward.

That is not an accident. It is a relief. A conversation about the technology is infinite, fascinating, and safe. A conversation about your own continued relevance is short, uncomfortable, and has your name on it. Given the choice, a room will pick the first one every single time — and congratulate itself on being forward-looking while it does.

Did your customers ever buy your product or service — or just the promise of a solution to their needs and problems?

Here is the lens that cuts through all of it, and it is not mine. Clayton Christensen spent a career on a deceptively simple idea: people don't buy products, they hire them to make progress — to get a job done in their lives. Nobody wants a quarter-inch drill; they want a quarter-inch hole, and they would happily fire the drill the moment something gets them the hole faster, cheaper, or with less mess. Your customers were never loyal to you. They were only ever after the solution and their benefit — or at least the promise of it. Always have been, always will be.

Run the AI question through that lens and it changes shape entirely. The threat was never that a competitor offers a better model than yours — those are bought by everyone, all the time, at various price points. The threat is that the job your customer hires you for becomes something AI can do directly, or makes someone else able to do so much better that the comparison stops being close.

This is not a new way to die. Theodore Levitt named it in 1960: the railroads did not decline because demand for transportation fell — it exploded. They declined because they had decided they were in the railroad business rather than the transportation business, and so they watched their customer walk to the car and the aeroplane without ever feeling addressed. The mistake is sixty-five years old. AI has simply deleted the decades of warning that used to come with it.

And the part nobody in the AI meeting wants to say out loud: owning the technology buys you nothing durable. Nicholas Carr made that argument in 2003 — once a technology is cheap and available to everyone, having it stops being an advantage; the edge migrates permanently to what you build on top. Rita McGrath took the next step: advantage itself is now transient, and the only real edge left is the discipline of asking, continuously, whether the thing you are good at is still the thing the market rewards. AI does not change that logic. It just shortens the fuse on it.

The elephant is not AI — but the question with no vendor: is what we do still worth choosing?

So let me say the thing the meeting is built to avoid, plainly, because somebody in the room has to.

The elephant is not AI. It never was. AI is the most-discussed subject in the building — and nothing that commands this much airtime is being avoided. The elephant is the question AI keeps pointing at and the organisation keeps stepping around: are we still relevant enough to be hired for the job — to whom, and for how much longer?

That is the whole animal, in one sentence. It stood in the room long before the first Copilot licence. AI did not put it there. AI merely turned the lights on and made it harder to keep not-seeing it.

And here is why that question is so reliably dodged: it has no vendor. There is nobody to buy it from, no demo to schedule, no maturity model to slot yourself into, no partner to share the blame with if it goes wrong. "Is our business still the one customers want to hire?" has no procurement process. So it loses — every single time — to a question that does. An AI programme has a budget line, an owner, a deadline, a flattering story, the cover of everyone-else-is-doing-it, and the visible appearance of motion. The relevance question has none of those. Five boxes ticked versus zero. The meeting that should be about whether customers still need you becomes, predictably, a meeting about the tool.

Which is exactly where I come in — and I should be straight about why it's me saying this and not the firm with the fancy high-gloss deck. Most advisors are commercially built to discuss the tool. An AI engagement has a scope, a deliverable, a renewable contract; "your business model has a relevance problem" has none of those, and naming it can end the engagement instead of extending it. I've made the opposite bet. I don't sell the wand and I don't sell the methodology — I'm the one whose only job is to point at the animal everyone else is professionally incentivised to walk around. (I've written about which kind of advisor is structurally able to do that — the Jester, the one with permission to tell the king the truth precisely because he isn't selling the king anything — in The Other Type of Consultant.) Pointing at the elephant is not a side effect of how I work. It is the work.