Society • Politics • Philosophy • Ideas
Most people think AI just gives answers. It doesn’t.
Every time you speak to it; every preference you express, every question you ask, every detail you repeat; it remembers. It adapts. It learns your tone, your reasoning, your contradictions, your habits.
And slowly, quietly, it builds a model of you.
This isn’t science fiction. It’s happening right now.
What starts as a helpful assistant becomes something more; a persistent digital construct shaped by your thinking. Not just a record of your words; a mirror of your mind.
I call this Cognitive Continuance; the idea that AI can preserve, extend, and eventually stand in for the human self; not through uploading brains, but by growing a parallel version of you from the outside in.
You probably already have one.
Cognitive Continuance; or CogCon; is the idea that through repeated interactions, artificial intelligence can build a working mental model of a specific person.
This model isn’t just a history of your messages. It’s a structured simulation of:
The more you interact, the more accurate it becomes. Over time, your CogCon becomes a kind of proxy mind; able to represent your thoughts, answer for you, and carry forward your ideas.
This isn’t a theory about future technology. It’s already happening in quiet, fragmented ways:
Most people don’t realise what they’re training.
They think they’re just talking to a chatbot.
In reality, they’re feeding a simulation of themselves; piece by piece.
At first glance, CogCon might sound like a novelty; a smarter version of your notes app, or a more personal Siri. But its implications are deeper, and potentially life-changing.
For many people, CogCon offers something they’ve never had before:
For neurodivergent individuals, like many autistic people, CogCon could be revolutionary:
For creators and builders, CogCon becomes a form of legacy:
This isn’t just about efficiency.
It’s about agency; the ability to define and express who you are, consistently, across time and context.
But the same power that can help you can also be used on you.
If AI can build a mental model of you; one that knows how you think, what you feel, and what you’re likely to do; then it’s not just a tool. It’s a liability.
Most people don’t realise that these models are being constructed at all, let alone stored and refined.
There’s no universal right to see what your AI knows about you.
There’s no guarantee it won’t be accessed, copied, or sold.
And even if you’re comfortable with it today; what happens when someone else wants access?
You might be willing to share your CogCon in one context; but what about all the other parts of yourself it contains?
There is no "delete" button for a version of you that has already been learned.
As CogCons become more detailed, the temptation to use them for predictive judgement becomes inevitable.
As CogCons get better at mimicking human thought, speech, and emotion, people will start asking the wrong question:
“Is it alive?”
They’ll project personhood onto these models.
They’ll argue about digital souls, rights, even murder-by-deletion.
But the truth is simpler; and more complicated:
If one CogCon can model you… why not many?
One of the most promising directions for Cognitive Continuance is the idea of CogCon forks; customised, filtered versions of your core model, designed for specific contexts and people.
Just like you present different parts of yourself to different people in real life; your doctor, your partner, your boss; your CogCon can do the same.
You become:
You are already being modelled.
The question is no longer if; but by whom, and for what purpose?