A simple explanation
The algorithmic self is the version of you that the platform's ranking system rewards. It did not arrive at once. It arrived through a feedback loop: you posted, certain posts got reach, those posts got repeated, posts that did not get reach got retired, and your composition drifted in the direction the algorithm preferred.
By the time you notice, your default voice — the one you reach for first when you sit down to post — is not the voice you chose. It is the voice the algorithm trained you to use.
An everyday example
You started posting honestly. Some posts did well, most did not. You noticed that posts with stronger hooks did better. You started writing stronger hooks. The hook-strong posts did better still. You noticed that posts with a contrarian angle outperformed measured ones. You started reaching for a contrarian frame even when your actual view was nuanced.
Six months in, you sit down to write about something you genuinely care about. The first draft is measured and honest. Before you publish, you sharpen it. Then you sharpen it more. The published version is louder than what you meant. It does well. The next post is louder still. Somewhere along the way, the voice in your head when composing stopped sounding like yours.
Why does this happen?
Because ranking algorithms are operant conditioning at scale. You produce variations; the algorithm reinforces some; you produce more of those; the algorithm reinforces some of those. Over hundreds of cycles, the loop-runner's behaviour shifts toward the variations the algorithm rewards.
The Belonging System cannot tell the difference between real human approval and algorithmically-mediated reach. Both register as belonging signals. So the System, asked to keep you in relation with others, gradually steers your output toward whatever produces the strongest signal — which is, by construction, whatever the algorithm amplifies.
The behavioral loop
A loop that hides in legitimate engagement:
- Honest post — the loop-runner produces a translation of the offline self.
- Algorithmic verdict — the platform either amplifies or suppresses based on signals the loop-runner cannot see directly.
- Belonging signal — amplified posts produce a wave of engagement; suppressed posts produce silence.
- Variation — the loop-runner makes small adjustments next time, tilting toward features they suspect the algorithm liked.
- Reinforcement — the adjusted post also gets amplified, confirming the hypothesis.
- Reweighting — over dozens of cycles, the composition voice drifts away from the offline self and toward the algorithm's preferences.
- False clarity — engagement is real, so the System logs the cycle as belonging success.
- Drift settles — the algorithmic self becomes the default. The original voice requires conscious effort to find.
Emotional drivers
Three threads:
- A real desire for the engagement, which is genuine human attention even if mediated.
- An incremental self-betrayal that registers below conscious awareness — the loop-runner senses the drift but cannot locate it.
- A growing dependence on the engagement signal, which makes the cost of returning to the honest voice feel disproportionate.
What your nervous system does
The amplification produces a dopaminergic surge tied to engagement notifications. The suppression produces a smaller, sharper aversive signal. Over months, the brain becomes more responsive to engagement-prediction features — the body learns to scan a draft for will this reach before checking is this true.
Sleep onset can suffer because the engagement-monitoring loop runs into the evening. Composition starts to feel like work even when the topic is one the loop-runner cares about.
The DojoWell interpretation
The algorithmic self is the cleanest contemporary example of the false_progress density signature. Each post that performs is logged as success. The engagement is not fake — real humans clicked, real humans replied. But the relations formed are with the algorithmic self, not the offline one. The deposit does not land where the loop-runner lives.
This is what distinguishes false progress from clean engagement. Clean engagement deposits onto the person; the relations formed in clean engagement reach the private self that composed the post. False progress deposits onto the algorithmic substitute; the relations formed are with a version of the loop-runner the algorithm trained.
The Belonging System does not know it is the wrong target. From the System's perspective, the cycle is working — there is engagement, there is reach, there is belonging signal. Only the offline self can tell that something is off, and the offline self gets less and less air as the algorithmic self gets more rehearsal.
How do I tell if I'm performing for engagement?
Three signals:
- You can predict, before posting, which posts will perform — and the prediction increasingly governs which posts get written.
- You catch yourself sharpening a draft past what you actually meant.
- Going viral leaves a faint hollowness rather than a clean satisfaction.
The hollowness is the data. Real human relation deposits as warmth in the body. Algorithmically-mediated reach deposits as a brief surge followed by an emptiness that asks for the next post.
Practical steps
- Read your last six months of posts in one sitting. The voice drift is visible in the prose. The drift is data.
- Write a post you know will underperform. Publish it anyway. Notice the body's response and the engagement's response separately.
- Mute your own analytics for a week. The System cannot tune to a signal it cannot see.
- Compose offline first. Write the post in a notes app with no formatting cues. Bring it to the platform last.
- Track the hollow-vs-warm distinction. After each post, name which kind of return arrived. The distinction is learnable.
Reflection questions
- What does the algorithmic version of you sound like? How is it different from how you sound in private writing?
- Which features of your posts did you adopt because you chose them, and which did you adopt because they performed?
- What would you write if you knew no one would see it? What would change in your composition?
- Where has the algorithmic self crowded out the offline one in topics you actually care about?
Frequently Asked Questions
Why does going viral feel hollow?
Because the reach formed relations with the algorithmic self, not with the offline one. The Belonging System logs the engagement as belonging success, but the private self that needs the warmth does not receive it. Real human relation deposits onto the person; algorithmically-mediated reach deposits onto the substitute. The hollowness is the deposit landing in the wrong place.
Am I posting for myself or for the algorithm?
Almost certainly both, in varying proportions. The honest question is whether the proportion has drifted — whether the algorithm's preferences now show up in your composition before your own do. The test is whether you can write a post you suspect the algorithm will suppress without flinching.
Can I undo the algorithmic self?
Partially. The trained voice does not fully unlearn, but it can be made non-default with steady practice: composing offline first, publishing posts you suspect will underperform, reducing analytics exposure. The original voice does not disappear; it just requires more deliberate retrieval after months of training.
Is this just impression management with extra steps?
It is impression management with a feedback loop the loop-runner did not design and cannot fully see. Ordinary impression management is shaped by readable human cues; the algorithmic self is shaped by ranking signals that operate at a different time scale and reward different features. The substitute is more entrenched because the training is more efficient.
How does this connect to Meaning Density?
The algorithmic self is a textbook false_progress pattern. Each engagement cycle logs success — the metrics are up, the reach is real, the belonging signal arrives — but the relations form with the substitute rather than the offline self. The deposit does not reach the person. Density is low because the equation is being run on the wrong target.