There was a time when AI was something we deliberately chose to use.
Open ChatGPT. Ask a question. Generate an image. Try an AI writing tool.
Today, that choice is slowly disappearing.
AI is moving inside the products we already use.
Search engines summarise results using AI. Phones edit photographs using AI. Email applications suggest replies. Office software writes documents. Browsers summarise webpages. Operating systems analyse what we are doing to provide recommendations.
Soon, using technology without AI may become harder than using technology with it.
AI Is Becoming the Default
The interesting change is not that AI is becoming more powerful.
It is that AI is becoming invisible.
Most people may not even realise when an AI model is processing their information.
Take a simple photograph.
Your phone may recognise the people in it, identify objects, improve the image, categorise the location and make it searchable using natural language.
Write an email and AI may analyse the sentence before suggesting what you should write next.
Search for something online and an AI system may decide what information should be summarised before you even visit the original websites.
These features are useful.
But convenience sometimes removes something important: choice.
That same trade-off between convenience and security is something I discussed previously in Convenience Is an Attack Surface. Convenience Is an Attack Surface
What Exactly Are We Agreeing To?
The bigger question is not whether AI is good or bad.
The question is what AI needs access to in order to become useful.
An AI assistant becomes much more useful when it can read your email, calendar, documents, contacts, browsing history and perhaps even your conversations.
But every additional permission expands what the AI system can see and potentially act upon.
This becomes even more important when AI evolves from simply providing suggestions to actually performing tasks.
As discussed in The Real Risk of AI Coding Agents Is Not Intelligence. It Is Permission, the important question may eventually be less about how intelligent an AI is and more about what we allow it to do. The Real Risk of AI Coding Agents Is Not Intelligence. It Is Permission.
The Question Is No Longer Whether We Want AI
AI will probably continue appearing in phones, computers, cars, televisions, browsers and everyday applications.
Some features will be optional.
Others may quietly become part of the product itself.
And that creates an interesting future.
We spent the last few years asking:
“Do we want to use AI?”
Very soon, the better question might be:
“Can we still choose not to?”
