Preface: Overheard on the Road to Shemot: Do we really need all of Moby Dick? What's Unimatrix Zero?
A legal and logical critique of the U.S. Copyright Office’s incoherent stance on AI authorship, using Moby Dick, Madonna, and the Borg to expose the collapse of lawful reasoning.
This is not an article about copyright policy. It is about unlawful policy implementation—specifically, the conduct of the U.S. Copyright Office, which now issues judgments based not on law but on belief, perception, and institutional bias. They are one example of a broader issue that is visible in politics, business, and science. The tragedy is not simply that they are no longer applying the law—concerning enough on its own—but that this is not being recognized as such.
The Star Trek Voyager series depicts the journey of a lost ship trying to return home to Earth, traveling through what’s called the Delta Quadrant. It so happens that this sector is also home to the Borg. They don’t learn things—they assimilate whoever knows something useful and turn them into a drone, with the entire collective sharing a hive mind.
It doesn’t work without some hierarchy in the form of the sinister Borg Queen. In one episode, they learn about what’s called Unimatrix Zero, a realm where drones with a certain genetic disposition can connect and regain their former individual consciousness and interact from their prior identities. The starship crew wants to exploit this, and the Borg Queen is trying to shut it down.
So the Starship Captain gets captured by the Borg:
Borg Queen: I know how this must upset you, Captain. As a Starfleet officer, you value all life—even drones. How many more are you willing to sacrifice?
Classic Stockholm syndrome tactics. She wants some code to see who is secretly fraternizing there. And Janeway says, fine, let me give orders to my crew (even though I probably couldn’t give orders on a military ship while held hostage by the enemy—but why not). The Borg Queen agrees and opens a Zoom call.
Captain Janeway: Tell them Unimatrix Zero can no longer exist. That’s an order.
And the crew says, “OK, OK”—and reads the actual meaning: don’t hand it over, blow it up. The Borg Queen doesn’t catch on, but in the meantime, she cracks the code and no longer needs Janeway’s input. Two hours in 250 words. That’s impressive. Why am I saying this?
Because if we don’t distinguish between truth and coherence, we lose the ability to judge at all. And the treatment of AI-generated output is based on blurring the line between fact and belief. That was Unimatrix Zero—a fantasy based on memory. But reality demands that we come to terms with it. I only mention it because sometimes it’s strange how often science fiction turns out to be prophetic.
If imagination counts as evidence, then the law loses meaning. The law becomes ceremonial oppression.
Because regardless of what the text of the law says, the outcome can no longer be derived from that text.
This is not a functioning legal system. It is feudal.
It violates the principle of the rule of law—and no one seems to notice.
If you disagree, I urge you to read again.
There is no freedom without recognizing how essential this is.
It is the assumption behind the Constitution that is being eliminated.
I am not presenting an opinion. I am making a factual claim.
It has long annoyed me that, when discussing such matters, someone will inevitably say: “Well, that’s your opinion...” followed by my answer: “This is not my opinion. This is a fact: you are wrong, I am right.” I say this with full conviction because it is a question of method, not content.
I have not set foot on the moon, but I believe—based on available evidence—that the moon exists and orbits the Earth. Could that belief be the result of a massive conspiracy? Technically, yes—its probability is greater than zero. I never claim absolute certainty in matters of meaning (e.g., "the moon exists"), but I can model certainty by stating assumptions and making falsifiability (or other criteria) a matter of argument structure. Uncertainty always remains, because rational assumptions made under uncertainty do not yield certain outcomes.
However, one thing is absolutely possible: if you state your assumption, I am able to determine whether that statement is valid or false. That’s what I do for a living—and I have yet to meet anyone who does it better. It's only about clarity of method and precision of application.
If you're thinking, "Last I checked, your name wasn’t on the list of U.S. Supreme Court judges," you might be right. But we don’t need the Supreme Court to settle this. I can do that myself, because I’m making a specific claim: does what they say and what they do match their stated principles?
The short answer is: no.
I can show how their actions contradict their stated framework. That contradiction can be fixed in two ways:
Claim they meant to cause this friction (and have the competence to invent a justification and thereby re-establishing coherence), or
Change their actions to align with the principle.
Which option follows is not my decision—I lack formal authority. But the logical incoherence is a fact, and it cannot be decided by a judge.
It's like seeing that the sky is blue: that is a perceptual truth, not a legal one. A judge can decide whether the incoherence should be treated as a legal problem, but the law doesn’t have the answer—logic and reason do.
Lawyers deduce. That’s related, but not the same thing. The law doesn’t work if we can’t distinguish between what is and what isn’t.
That’s not an opinion. It’s a logical condition.
I don’t know if they teach that at Harvard Law School or not. If not, that’s the problem right there.
Law can, however, define fictions: for example, "the blue on Sundays is more significant than as on other days, so people get a day off." That may not make scientific sense, but it's a coherent framework, even if it's based on a fiction—as long as it is applied equally and predictably.
If the law says: “When the sky is significantly blue, people get a day off,” and defines Sunday as "the day the sky is blue that is considered significant," then fine. But if I say Monday is better for a day off, that's an opinion, neither right or wrong from a legal standpoint. Because the law in my example doesn’t care about better but about the idea of significance. The sky doesn’t objectively change color. What matters is whether the rule applies consistently.
This leads to further legal questions:
– What if people don’t have a calendar?
– Is it fair if some months have five Sundays?
– Should we rename Saturday to fix it?
That is what law is about.
Under the rule of law, outcomes must be derived from written principles, applied equally and predictably. When institutions begin denying rights based on whether AI tools were used—rather than examining the structure or substance of the work—they violate this principle. There are no small violations here. It’s legal blue screen.
The law only works if it can tell whether an apple is an orange. But copyright law now says: An apple is an orange — unless it was picked by a robot.
The problem is: You can’t tell by looking at the apple.
Not even computers can, because they only compare patterns — green to green, shape to shape — but they don’t know what an orange is.
That means the decision is not based on the structure of the work, but on a guess about how it was made. And that violates the principle of equal treatment: like must be treated alike.
Undisputed facts: US Copyright Act provides copyright for
Protection is for expression in a medium (MP3 file which is or contains an the artist’s expression) from which they can be perceived
The human made the file using traditional elements of authorship (literary, artistic, or musical expression or elements of selection, arrangement, etc.)
And we say this gets checked copyright,
In the case of works containing AI-generated material, the Office will consider whether the AI contributions are “mechanical reproduction” or instead of an author’s “own original mental conception, to which [the author] gave visible form. When an AI technology determines the expressive elements of its output, the generated material is not the product of human authorship.
Their principle are fiction but since this seems not clear to them let me rephrase it more clearly.
The law gives copyright for the medium file but confirms if a file contains a valid expression not by what the file is but whether there was a prior file that is structurally similar enough.
So in case of applying for copyright and determining whether or not it’s a copy, the traditional elements of authorship (literary, artistic, or musical expression or elements of selection, arrangement, etc.) are not used to determine the applicant’s authorship but to describe traditional elements of authorship of somebody else.
No one at the Copyright Office opens an MP3 of Like a Virgin and says,
“Ah yes, this sounds like Madonna — clear authorship here.” They don’t look at waveform patterns or analyze her vocal inflection. Instead, they look at the application:
Name: Madonna
Known song
Not disputed
Registered historically
And that’s it.
There’s no check to verify the “expression” by assessing traditional elements of authorship, That is for disputes and not for undisputed copyright applications was authored by her.
This is exactly where there is such a problem. The law sets out requirements that are met by doing different but coherently different things:
Fiction: Expression in medium →check authorship of medium→give copyright
Reality: Expression is not medium → don’t check authorship→give copyright
AI generated content
Fiction: Expression in medium →check medium authorship→give copyright
Reality: Expression is not medium ‼️ check authorship ‼️give copyright
In other words, the problem is not that they treat AI-generated content differently — it’s that they treat it differently in an incoherent way, leading to contradictions.
Instead of applying the exception formally to the tool (i.e., declaring that the use of AI precludes copyright), they make it conditional on how the use of the tool impacts the assessment of traditional elements of authorship.
But they have no clear criteria for assessing this when it comes to establishing authorship — because they don’t actually assess it.
And since they don’t assess it, they simply stipulate: if we had assessed it, we would have concluded that the AI — not the human — controlled the authorship of the output.
Therefore, authorship is denied.
And they’ve confused themselves into believing this process is grounded in the law — when in fact, it is unlawful. This is a double entendre of legal logic — they say A ("we evaluate authorship based on the expression in the tangible medium"), but they do A-in-name-only while meaning B ("we infer authorship from tool usage, not expression").
Text (like Moby Dick) is structured but, crucially, humans can fully perceive the entire set of letters used—unlike the 16.7 million colors or complex frequencies in digital images or sound. With letters, you need to ask again: in what reference are they used?
What is the meaning, and what is the probable next letter?
D I C _ → English → (Moby) Dick – Name
D I C _ → Roman → DIC = 601, using letters as numbers per Roman convention
If DIC_ is an incomplete Roman numeral and we’re considering only one more letter (i.e., a single-character continuation), then it must be one of the 7 Roman numeral letters:
I, V, X, L, C, D, M.
But “M” is stylistically invalid as a continuation of DIC (it would produce DICM, a malformed numeral).
Music is different. It unfolds over time and can still be recognized in fragments (e.g., “la la la” = Madonna), even if much is lost.
Therefore, losing "detail"—i.e., perceptual loss—is not linear, and representation in a medium is not proportional to expression. That makes it impossible to judge authorship solely based on the medium or the technical fidelity of a copy—because the medium cannot be understood without knowing the context. (You can’t understand a French poem if you don’t speak French; therefore, the poem is not the full artistic expression on its own. It’s only a poem within its context. “Artistic” is the idea of how poem and language connect.)
Because the file does not contain the expression—it only encodes it relative to a system. In music, that system includes harmony, rhythm, structure, rhyme, and genre. The artist didn’t invent these components; their meaning arises only within that framework.
And the law pretends otherwise—but respects this fact in how it’s applied. Except in one case: AI. For that, they break the fiction of the fiction. It appears as though the law is being applied the same for AI and non-AI works. But they don’t use the file for any assessment—they argue: “If we were to do it, authorship would be difficult to evidence.” What they fail to mention is that this is universally true: they never check authorship from the file alone.
This creates a category confusion:
Stated basis: expression in the file (observable, formal criterion).
Actual basis: suspicion of AI involvement (non-observable, tool-based assumption).
The result is:
They appear to follow the law.
But the criteria shift silently.
Therefore, the law no longer governs the outcome — discretion does.
This shows:
A fiction of evaluation (they don’t really look at expression).
A false pretense of neutrality (the law applies equally).
A collapse of principle into opportunistic policy.
And this contradiction is not hidden — it’s written in plain sight. And where are our legal experts, law professors and civil liberties experts to call this out? Good question.
Their technical assumption is undeniably wrong but my argument doesn't challenge the validity of their technical assumption. I am saying you are not consistent following through on your assumption.
The end.
This is outside the proof that the Copyright has unlawful doctrine. It is saying why are their assumptions unconvincing:
The Law Requires a Category, But Can’t Explain Meaning
If it’s music, it goes in one box.
If it’s literature, another.
If it’s visual art, another.
But what if your work is emoji-based? Is it:
A visual design?
A literary sequence?
Or a hybrid code that only has meaning in interaction?
The answer isn't legal — it's cultural and contextual. And that’s the problem: copyright law wants fixed categories, but meaning moves. Let’s say:
Book A is: 😋😎😎
Book B is: 😋😎😗
Now the legal question becomes:
Are those two the same work?
If not, is one a derivative?
If they’re different, how do we prove it — by file structure or perceived meaning?
But meaning isn’t in the file — it’s in how humans interpret it. And different readers could assign entirely different meanings.
Copyright demands a fiction: that expression is in the file.
Meaning isn’t in the file ‘😋😎😎’
Therefore: the test of originality and infringement is structurally incoherent.
What is the different between AI generated/manual expression
User prompt: “Make a pop song” → Music AI → Song file
Van Gogh: “Vincent, I want a yellow picture.” → buy color → Sunflowers
But what is yellow?
RGB yellow is a secondary color: Red + Green = Yellow
Yellow = #FFFF00 = (255, 255, 0) → That’s 255 red, 255 green, 0 blue
Say an Instagram picture is a square 1080x1080 pixels so xy diagram going from zero to 1080. A pixel located at 1x1 has a colour created by mixing 0 to 256 signals of red green or blue, this means 256 × 256 × 256 = ~16.7 million possible colors
If we estimate:
Humans may distinguish ~1–10 million under lab conditions but we can’t reason at this level. Proof me wrong by counting correctly to 1 million. I be you can’t but you know how to in theory.
~500 perceptual variants per main color so 500 yellows we can see
12 “canonical hues” = ~6,000 perceptual color identities
Question
Look at a Van Gogh sunflower painting: How many colors did he use (out of 6,000)? → I bet you can’t answer.
Look at two Sunflowers paintings and ask: Which used more colors? → Probably just a 50/50 guess.
Suppose I create a version of a Van Gogh in RGB, but reduce it to allow only 12 different colors. This means that out of ~500 perceivable variants of "yellow," only one is used (say, (250,250,0)).
Will you notice the difference? Yes
Would you be able to correctly count the colors used? Yes (assuming you don’t have impaired eyesight, I let you turn on the lights, and I don’t install spotlights in a tinted hue).
If you can identify all the structural elements of a 12-color Van Gogh, but not of a 16.7 million-color Van Gogh, doesn't that mean you can perceive the 12-color version—but not the original? And the question is not whether you get the numbers right. The question is: If you can't get them right, how can you say you truly perceive it (which copyright needs) and if we can’t see all the different yellow then in how does the medium contain the expression?
For a Painter, the traditional elements of authorship typically include:
Conceptual Selection – Choosing the subject matter (e.g., painting a yellow circle)
Arrangement/Composition – Deciding where and how elements appear on the canvas
Color Choice – Selecting the palette to express tone, emotion, contrast, etc. etc etc bla blah
These elements are all considered intentional and executed by the human, and thus count as “authorship.”nBut that is a fiction, because it would require that you can distinguish at minimum 16.7 million colors.
If you are blind, you cannot see whether it’s red or green at the traffic light. So how can we say it’s your fault if you get run over crossing at red? We don’t — we add a sound signal.
Now swap the traffic light for choosing “yellow” in a painting. If you can’t perceive the difference between two nearly identical options, how can we reward you as the “author” for being unable to pick purposefully one over the other?
🎨 Visual Choosing from 12 shades of yellow creates a relative structure of light, contrast, and color meaning. You don’t need to understand pigment chemistry to make a meaningful aesthetic choice. The output emerges from choices within a constrained space — not from total authorship of that space.
🎵 Music Saying “make me a pop song” is selecting from a genre space (e.g., 12 genre archetypes). You don’t need to understand music theory or signal processing to guide the process. The output emerges from your intent, filtered through a constrained structure — just like with paint.
💡 Therefore: The act of constrained selection from a perceptual set is the same. Even if the domain is different (color vs. harmony), the creative function is structurally equivalent. You’re not coding the output — you’re
Choosing from within a system you didn’t build
Navigating options you don’t fully see
Constrained, not full control
You can prove me wrong by making a song and get it on Spotify recorded at >20 kHz if you can’t you did not control authorship
So the legal argument that says: “You didn’t author this AI-generated thing because you don’t control all the parameters” …is incoherent — unless we also say: “Painters aren’t authors unless they invented their own pigments and could perceive all color wavelengths.” Which, of course, we don’t say.
Which file contains the full artistic expression?
If a song is available as WAV, MP3, and compressed streaming — do all these versions contain the full work? Or is it 3/3, 2/3, or 1/3?
It’s hard to answer, because music is relational. But consider this:
Say Moby Dick is a 600-page novel, written using just 26 letters — a fixed palette. Now imagine we have three versions:
One with 600 pages
One with 400 pages
One with 200 pages
Which one contains the full artistic expression?
Can the question even be asked?
If Melville wrote 600 pages, who decides what portion is essential? Is 500 pages enough to make it “Melville”? Who decides what matters?
The U.S. Copyright Office needs to answer this coherently to justify its stance on AI. And they cannot. And why this bureaucracy?
The U.K. (and most of Europe) fully embraced the Berne Convention’s principle: no formalities required. Copyright is automatic and enforceable upon creation. No need to register.
In the U.S., you can own copyright automatically, but you must register to fully enforce it. The U.S. conformed to the Berne Convention since 1989 in name only. The Berne Convention prohibits formalities as a condition for copyright protection.
But the U.S. retained registration as a prerequisite for bringing a lawsuit (17 U.S.C. § 411), or continued to offer legal benefits (statutory damages, attorney’s fees) only to registered works. These aren’t technically preconditions to copyright, but they are procedural hurdles that undermine the “no formalities” principle. And ownership without an ability to exercise ownership rights is no different from communal ownership. Just saying.
Overall, it’s a system not anchored in reality but in ceremonial acts — and if change is needed, the behavior becomes erratic.
Shemot is personal journey through dysfunctional systems. It reveals incompetence is not accidental—it is engineered. So what, if anything, can still save us? A new mental operating system. An upgrade in thinking and a migration path. That’s Shemot.
#myndOS #Shemot #MyDigitalTruth