Absorbed actions are a much more recent part of this story, but they belong immediately after images — because the image example already contains one.
Consider the system that generated 122 product images.
I did not sit down and say, “Generate 122 product images.” I did not even know 122 product images were needed.
I did not know a database was being constructed. I did not know that the database contained 122 products. I did not know that the product schema benefited from having an image associated with every record. I did not know that Google and other systems would expect certain structured information around those products.
The system knew. And then it acted.
The images themselves belong in the Images category. But the decision that those images needed to exist, the preparation of the product records, the generation of descriptions, the construction of the database, the schema work, the calls to other tools, and all the other work occurring beneath what I was consciously doing belong to another category.
Those are absorbed actions.
Absorption is not automation
The most important distinction is that absorption is not the same thing as automation.
Automation is generally something we consciously establish. We know a task exists. We decide how it should happen. We configure something. We give instructions. We schedule it. We monitor it. We occasionally repair it.
We attend to it.
Absorption is different.
An action has been absorbed when I no longer have to attend to it — and the action still occurs.
Below intention
There is a further distinction I did not have language for at first, and it turns out to matter more than anything else here.
An absorbed action does not contain some small intention of mine that I forgot to voice. It contains no intention at all.
Nobody wanted those 122 images. Not me — I did not know the database existed. And not the system either, not in the way a person wants something. A well-established pattern required them. The system knows how Google likes the world, and it completed the pattern the way a heart completes the next beat.
That is why absorption sits beneath attention. There is no intention down there to be conscious of.
It is also what separates absorption from the category that came after it. A sovereign act is an intention that is present and is not mine — an entity forming its own view about how something should be done, and acting on it without asking.
Absorption happens below intention.
Sovereignty happens beside it.
The synthetic subconscious
The closest biological analogy is the enormous amount of work happening inside a human being without conscious attention.
I do not consciously manage my heartbeat. I do not consciously manage circulation. I do not consciously manage digestion. I do not calculate the movements required to maintain my balance while walking across a room. I am not consciously determining when the next breath should occur.
These things happen beneath attention.
That does not mean nothing sophisticated is occurring. Quite the opposite. An extraordinary amount of prediction, coordination, sensing, and action is occurring continuously. I simply do not attend to it.
That is what I mean by absorption.
Artificial intelligence increasingly gives me something resembling a second layer of this kind: a synthetic subconscious. It predicts that something will be needed before I consciously ask for it.
The system working on a business may determine that 122 product images are necessary. The prediction mechanism itself does not necessarily create the images. It may call an image-generation system. That image generator performs its particular function and returns the images.
Biology works the same way conceptually. What we casually call “the body” or “the subconscious” is actually composed of many specialized systems performing different functions and coordinating with one another.
Artificial systems increasingly have the same character. One capability predicts what is required. Another searches. Another writes. Another generates an image. Another constructs a database. Another validates the result. Another publishes it.
From my perspective, however, most of those individual actions have disappeared. They have been absorbed.
Why the site is plural
This is also why the website is JohnRectors.com, plural.
John Rector is obviously still a human being. But the productive entity represented by that name is increasingly a federation: one human being in persistent relationship with many different artificial intelligences and capabilities working on his behalf.
I operate at a higher level of intent. The federation handles enormous quantities of activity underneath that intent.
That is very different from the way we traditionally thought about using computers. Historically, the human usually needed to know what operation needed to occur. We clicked the button. We opened the program. We selected the command. We created the workflow. We programmed the automation.
Even sophisticated automation still generally begins with human attention. Absorption begins when that requirement disappears.
The scientist and the meals
Imagine a scientist working twelve hours a day on an extraordinarily important aerospace problem. He is completely consumed by the work. Yet three excellent, healthy meals appear beside him every day.
He does not shop for the ingredients. He does not farm them. He does not choose the restaurant. He does not open DoorDash. He does not browse a menu. He does not enter a credit card. He does not schedule a delivery. He does not watch a notification. He does not walk downstairs to meet a driver.
The food simply appears when he needs it, and it is exactly the kind of food he wants.
From his perspective, feeding himself has been absorbed.
Contrast that with ordering DoorDash every day. DoorDash certainly automates portions of food acquisition, but the scientist is still attending to the process. He has to stop his aerospace work, open the application, decide what he wants, select items, place an order, deal with payment, watch the delivery, and retrieve the food.
That is automation. It saves labor. It does not eliminate attention.
This is the distinction that matters.
Automation transfers execution.
Absorption transfers attention.
640,000 of them
An absorbed action is not merely something a machine does for me. Machines have done things for humans for centuries.
An absorbed action is something that continues to be accomplished on my behalf without requiring me to consciously represent the task at all.
I did not delegate those 122 product photographs one by one. I did not delegate the product descriptions. I did not delegate the database. I did not delegate the schema. I did not supervise the sequence.
At the level where my attention existed, I was pursuing a larger objective. The federation inferred the intermediate work required to accomplish it.
This is why absorbed actions have become such a large number in such a short period of time. The ledger currently records approximately 640,000 absorbed actions. And that number will grow much faster than categories such as books or websites, because one visible artifact can contain hundreds or thousands of invisible actions beneath it.
The book is visible. The website is visible. The image is visible. The absorbed actions usually are not.
You see their evidence in the finished world. A database exists that I never consciously designed. A product has an image I never requested. Information has been organized in a way I never specified. A system has made a tool call I never knew was necessary.
The work happened.
The result exists.
But my attention was somewhere else.
That is absorption.