A thing does not disappear from attention because it is unimportant. It disappears because it is predicted well enough.
The strangest implication of the model is not that surprise captures attention.
Most people already know that.
The stranger implication is this:
Important things can disappear from attention.
Your heartbeat is important.
Your breathing is important.
The floor holding your body is important.
The background systems keeping your life intact are important.
And yet, most of the time, they produce no conscious attention-demand.
Why?
Because they are absorbed.
What Absorption Means
Absorption is not unimportance.
Absorption is prediction success.
In the formal model, pure absorption occurs when Reality is perfectly predicted across the attention window.
For all tau in [t-W, t]:
R(tau) = 1Therefore:
S(tau) = 0
C_W(t) = 0
Attptr(t) = 0
D_att(t) = 0No accumulated surprise.
No attention pointer.
No attention-demand.
Threshold Absorption
Human attention is finite. Not every small remainder becomes conscious.
If:
|Attptr(t)| < deltathen the system produces no conscious attention-demand.
This does not mean nothing happened.
It means nothing survived into conscious attention.
Resolved Absorption
There is also a subtler case.
Reality does not need to be perfectly predicted at every microtick for attention to disappear.
Micro-surprises can exist inside the attention window and still integrate toward zero:
C_W(t) ~= 0Positive, negative, and differently oriented surprise components can cancel under integration and renormalization.
This explains why a human can be surrounded by micro-surprises and still experience calm.
The system is not empty.
The remainder is low.
Why Working Systems Disappear
A working system disappears from attention because it generates no unresolved surprise.
A reliable door does not ask to be noticed.
A reliable payment system does not ask to be noticed.
A reliable assistant does not ask to be noticed.
A reliable body process does not ask to be noticed.
The moment the system fails, attention returns.
The hinge squeaks.
The payment fails.
The assistant makes an error.
The tooth hurts.
The breath catches.
Attention appears where prediction breaks.
The Ethical Edge
This gives a sharper language for technology.
Many systems claim to save time while increasing attention-demand.
They create alerts, reviews, dashboards, errors, decisions, ambiguity, and monitoring burdens. They do not return attention. They convert hidden work into conscious residue.
A better system absorbs what can be predicted.
It removes unnecessary remainders.
It lets the human keep attention for what cannot, and should not, be automated away.
Next: Bad AI captures attention. Good AI returns attention.