What to ignore

The hardest part of advising on technology right now is not finding what is possible. It is deciding what to leave alone. The volume of claims has grown faster than the ability of any leader to assess them, and most of the pressure people feel is not a considered view that they are behind. It is the accumulated weight of everyone around them sounding certain.

Judgment, in a market like this, is mostly subtraction. The useful skill is not spotting the next thing. It is having a small set of questions that let you set most of the noise down without guilt, so that attention is left for the few things that genuinely matter to your business. Here is how we tend to think about it.

Does it change the economics of something you already do

A real shift usually shows up as a change in the cost, speed, or quality of an activity you already care about. If a capability makes something you do meaningfully cheaper or better, that is worth attention regardless of how it is described. If you cannot connect it to an existing activity and its economics, it is probably interesting rather than important, and interesting is not a reason to act.

Would it matter if you stripped the label off

A quick test for hype is to describe the thing without the word AI attached, in plain terms of what it does and what it costs. Much of what survives that translation was always worth considering and would have been considered anyway. Much of what does not survive it was riding on the label. This is not cynicism about the technology. It is a way of seeing the substance underneath the branding.

Is the claim about capability or about a demonstration

A demonstration shows that something can be done once, under favourable conditions, by people who built it. Capability is whether it works reliably in your environment, with your data, your constraints, and your people, at a cost you would accept. The distance between the two is where most disappointment lives. When you read a striking claim, ask which one it is. Often the honest answer is that nobody knows yet, which is itself useful information about timing.

Who benefits from you believing it

Not every source is disinterested, and that is fine, but it is worth noticing. A vendor, a platform, and a commentator building an audience all have reasons to make a shift sound larger and more urgent than it is. This does not make them wrong. It does mean their certainty is not evidence, and that you are allowed to wait for something more durable than enthusiasm.

The point of ignoring things

None of this is an argument for standing still. It is an argument for spending a scarce resource, which is the attention of your senior people, on the few things that actually bear on where you are trying to go. Every claim you can responsibly set down is capacity returned to the work that matters.

We think a clear view of what to ignore is one of the most valuable things an outside adviser can offer at the moment, precisely because so little of the market is willing to say it. AI is a support, not a goal. Most of the noise is neither.