Apple’s new CEO John Ternus, who took the job in September 2026, gave an interview to Wired’s Steven Levy, reported by TechCrunch on 31 October 2026. Asked about the view that Apple has fallen behind in AI, he said the company doesn’t “want to ship AI for AI’s sake”. His test for whether Apple got it right: “did we build something that they really love?” He added that when it is done well, people won’t have to think about whether AI or traditional software made something happen.
The question under the headline
The excitement around this story is about whether Apple has lost the AI race. That matters less to a business owner than the question Ternus is asking, because it is the same one I ask about every tool that crosses my desk. Not “what can it do?” but “what problem does this solve that I couldn’t already solve, and is it worth adding?”
Strip the Apple context away and that is a product decision rule. Apple ships features to people who already own the device. A business owner adds tools to a team that already has a way of working. Both face the same cost: every addition has to earn its place, or it sits there as clutter that someone still has to maintain.
What actually changed, and what didn’t
Nothing in the interview changes what Apple can or can’t do. What the interview does is make a posture public. Apple is telling you it will judge AI by whether the product got better, not by whether AI is in it. You can disagree with Apple’s execution. The posture itself is sound, and most businesses have not adopted it.
I see the opposite posture all the time. A shop owner signs up for an AI writing tool, an AI design tool, an AI customer service tool and an AI analytics tool in the same quarter. Six months later three of them are unused, and the one that is used gets used at about a tenth of what it can do. The spend was small. The attention was not.
How I run the test myself
I use Claude for almost everything. That is a deliberate choice, made after watching the rush to personal AI agents, each one promising to handle a different slice of life or work. One tool used deeply has been more efficient and more effective for me than a stack of tools used shallowly, because the tool learns my context and I learn its limits.
Before I add anything, I ask what I would stop doing if I had it. If the honest answer is “nothing, but it would be nice”, it doesn’t get added.
What this looks like for a real operator
Take a clinic with a front desk of two people. A vendor pitches an AI tool that answers patient questions on WhatsApp. The right first question is what the front desk actually does with its day. If both people spend their mornings telling patients what time their appointment is and whether the clinic takes their insurance, there is a real need. The tool takes over a job that exists and that someone is paid to do.
If patients rarely message and the desk mostly handles walk-ins, the AI tool is a solution looking for a problem. It will get set up, answer a handful of messages a week, and be forgotten. Same product, two different answers, and only one of them is worth adding.
This is also why I built AeroChat around what the conversation actually does in your systems, rather than how clever the chat sounds. At the chat level, every AI chatbot is almost the same. The need is a refund processed, an order checked, a booking made, without a person doing it. That is a job that exists.
What to do before adding any AI tool
Run the Apple test on your own business, in this order.
- Name the job the tool would take over. A specific task, done by a specific person, taking a specific amount of their week. If you can’t name one, stop here.
- Ask what you would do with the time or money it frees up. If there is no answer, the need is not real yet.
- Check whether a tool you already pay for can do the job. Most businesses use a fraction of what they already have.
- If it passes all three, add it and go deep on it. Give it the context it needs, learn its limits, and use it fully before you look at the next one.
- If it fails any of the three, put it on a list and revisit it in six months. Needs change. Tools get better. The rule stays the same.
Source: TechCrunch: Apple CEO says the company doesn’t ‘want to ship AI for AI’s sake’