I Miss My Troubled, Unwell Norwegian Friend

Stuart K. Hayashi

 

I still worry about my troubled, unwell Norwegian friend, the one I wrote about previously. I mean my troubled friend who was obsessed with child molesters: the one who kept bringing up, out of nowhere, her fascination with child molesters, only for it to turn out later that her paternal grandfather and two of his brothers were all credibly accused of sexually abusing girls over whom they were tasked with watching over.

I dearly wish for reconciliation, but that is not to happen absent of her returning to regular psychiatric care and ending contact with those who were, and likely still are, reinforcing her morbid gestures. I know that that is a prerequisite, and I know it because of what already happened years ago.

Soon before returning to Norway for the summer, she promised me unsolicitedly that when she returned to Norway, she would resume seeing a mental health professional. I did not initiate asking her about it. She took the initiative to make that promise, unprompted by anyone else. When she was back in Norway, I asked her about it, and she claimed not to remember. Then she said she doesn’t need a mental health professional anyway. Soon after, she became mostly uncommunicative. But in one of the few times in that duration when she did communicate again, she said she was having panic attacks daily. Then when she returned to O‘ahu, she was uploading onto social media the photoshops that an internet-famous Norwegian artist did of her where she was photoshopped to have a chalky-white face like a corpse’s.

After two years of that, and after returning to Norway for the long term, she stopped uploading the dead-body photos. But then Corpse Artist started uploading pretentiously photoshopped images of himself up in the mountains with his mountain bike. Then my friend followed with photos of herself — no longer with the dead-body face — always hiking on mountain tops. Corpse Artist doesn’t do the corpse art anymore, but his influence is still very much in the picture (sometimes literally). 

And my friend made a big show of legally changing her last name to match her father’s, a last name that was also the maiden name of her paternal grandmother who feigned ignorance about the child sexual abuse inflicted by her husband (my friend’s paternal grandfather). (Previously, my friend’s legal last name was her mother’s.) With my now knowing the history of what happened on her father’s side of the family, I think that though it is much more subtle than the chalky-white-face dead-body photos, the legal name change looks like another morbid gesture. It seems a very try-hard attempt to convey that her relationship with her father is fine; even great.

If I resume contact with my friend, and she’s not angry but instead welcomes reconciliation but doesn’t return to mental-health treatment and stay with it, I know what is going to happen. It will be a repeat of what happened before. She might return to treatment for a little while. But eventually she will stop with it and, when I ask about it, she will feign memory loss about it as she did before. And then all of our interactions will become a repeat of what happened that summer. At first it will again seem be a happy situation. But then my friend will resume making morbid gestures with the expectation that I play along and act as if they are safe and fine. And playing along and acting as if that is safe and fine is tacit reinforcement of the morbid gestures.

My friend first establishing regular psychiatric treatment is necessarily a prerequisite to any attempt at reconciliation.

If my friend sees this: legally remove the name of those who facilitated abuse: the last name of a father who facilitated abuse and the maiden name of a paternal grandmother who “looked the other way” to maintain plausible deniability about abuse that she knew had happened. Cut out of your life and social media those, such as Corpse Artist and his sister-in-law and brother, who reinforce your morbid tendencies. Rather than thanking them for their morbid images, they always should have been blocked, just as they should be now. Return to regular psychiatric care and have the borderline/ emosjonelt ustabil personlighetsforstyrrelse diagnosed properly. Make this known publicly. With the danger you have consistently imposed on self and others, even people outside of the inner circle have a right to know about your psychiatric condition and the risks of it. Such important steps would not be for me, but for your own happiness, well-being, and peace. But just as it was in our last phone conversation, you already know that.

Online AI Image-Repackagers and the Libertarian Apologists for Their Art Theft

Such Libertarians Whitewash Regurgitative AI’s Copyright Infringements Not Because They Like the AI But Because They Wish to Erode Copyright Protections Further


Stuart K. Hayashi

There are tremendous issues with what is called “generative AI.” The issue is not with the technology as such, but with the unethical way in which it is being used — a use that the companies behind “generative AI” demand go unpunished by law. And that is a dilemma because it involves the violation of intellectual property rights (IPRs). And on account of their party line from the 1970s being one of undue hostility to IPRs, it is all too common for self-proclaimed libertarians to whitewash the harms being inflicted by the “generative AI” as it currently stands.

The usual people who make use of “generative AI” to produce images, which they call art, supply their own rationalizations for the theft. Artists on social media, such as on the Facebook/Meta Company’s platform Threads, frequently get into arguments with those users, whom they call “AI bros.” Yet such artists are unfamiliar with the rationalizations and motivations by IP-hating libertarians. Thus, when artists on social media come across IP-hating libertarians trying to whitewash “generative AI,” those artists mistake those IP-hating libertarians for the usual AI bros. However, in the context of their propaganda for “generative” AI’s copyright violations, both the motivations and some specific specious arguments of IP-hating libertarians are notably distinct from the AI bros’.

In this essay I shall explain the motivation behind the IP-hating libertarians’ insistence on shoving themselves into this controversy, and how it differs from the AI bros’ intentions. After that, I shall explain a particular specious argument from two IP-hating libertarian journalists who, unlike most IP-hating libertarians, are generally respected among more-mainstream people, such as those on the political center-Left. Then I shall expose the fallacy of that rationalization from the two IP-hating libertarian journalists.

“Generative” AI As Anything But
We begin with an explanation of how “generative AI” trashes intellectual property rights — and how the generative part of that name is a misnomer. Essentially, what happens is this. The user types in what sort of image he wants — an image that we are initially misled to believe comes fully formed out of nowhere. What the user types in is called a “prompt.” And then the image appears.

The truth is that the image does not come from nothing. Rather, what happens is this. The AI sends out “bots” or “crawlers” to crawl the World Wide Web to scrape for data. Upon finding data, such as a copyrighted visual image, the AI produces an exact copy of that image. This is misleadingly referred to as “training” of the AI. Human consumers of this product send prompts to the AI to deliver a new image. Upon this, the AI takes the data it stored and, through use of statistical probabilities, assembles an image derived from that data. This is highly commercialized. Users pay money to the AI companies for this service. Yet the artists who created the actual copyrighted visual images upon which this technology depends, receive no financial remuneration for the economic value that the AI company took from them, the value for which the AI company’s consumers pay to the AI company instead. The AI company did not even ask first for the copyright holders’ permission.

It is thus misleading to say that this technology is “generative.” Nay, it is derived nonconsensually from the economic value that other parties have created. As observed by novelist Maggie North, a more accurate term for this is “regurgitative AI.”

Nor are these, as they have been called, AI image-“generators.” They are image-repackagers. The digital artist Reid Southen notes at the 21-minute, 1-second timestamp, that it is misleading to say that with regurgitative AI, online artists are being expected to compete against a machine. Rather, online artists are expected to compete against “their own work” that has been stolen from them.

A false comparison here in response, made both by AI bros and by IP-hating libertarians, is that the regurgitative AI is just doing what copyright-respecting artists have always done in creating new works inspired by other parties’ artworks that preceded them.

As I have written before, there is a vital distinction between copyright infringement versus one artwork being inspired by other parties’ artworks. IP-hating libertarians note the importance of artists having the freedom to make works greatly influenced by others’. But then those IP-hating libertarians do conflate artistic inspiration with copyright infringement. They do this to rationalize that copyright infringement is actually not bad after all. At the conclusion of this essay I will return to spelling out exactly what is wrong with that equivocation on the IP-hating libertarians’ part.

Here I will point out a crucial difference, though further explanation will have to wait until the end of this essay. For now, it will suffice to remember that when artists produce commercial artworks that still other, preceding artworks have inspired, the artists of the new commercial works do not rely on the production of exact copies of the older works against the consent of older works’ creators. By contrast, it is inherent to the commercial production of new images by the AI that it produces exact copies of the works from which the new images are assembled.

As long as this is done absent of the explicit consent of the owners of the copyrights on the original works the AI is using, the AI companies are indeed directly benefiting financially from the nonconsensual copying and pirating of copyrighted artwork. There are two important values that AI companies owe to copyright holders — first, the revenue accrued from the use of copies made in the data-scraping of the copyrighted artwork and, second and far more important, the ethical acquisition of permission from the copyright holders that should have been sought from the outset.

However, two well-respected, IP-hating libertarian journalists have a reply to this. They say that what the regurgitative AI is doing actually is the same, in principle, as artists making new artworks inspired by other artworks. The two IP-hating libertarian journalists say this is so because, if Artist Bob produces his own commercial work inspired by that of Artist Angela, Artist Bob does necessarily rely on making and storing an exact copy of Artist Angela’s artwork. You are justified in wondering how this possibly could be so. The two IP-hating libertarian journalists continue that the exact copy is stored in Artist Bob’s memory, in his brain. That claim is so asinine that it baffled me initially. That argument departs so substantially from even a remote understanding of memory and human psychology, that I was baffled how two men who are seemingly intelligent could make it.

But now I understand. The desire for IP-hating libertarians to rationalize their desire to undermine IP protection is so ingrained that they will engage in such mental gymnastics. These are two libertarians who claim to have advanced beyond the point of toeing the party line of Murray Rothbard and the other Rothbardians of the 1970s. But in their being latched onto the anti-IP talking points of the 1970s, these two well-respected libertarian journalists are indeed stuck in the party line.

As I said earlier, when artists trying to protect their own artwork come across IP-hating libertarians’ apologetics for regurgitative AI’s copyright theft, those artists usually mistake those IP-hating libertarians for the usual AI bros. I want to rectify that. The IP-hating libertarians pose a specific danger to artists that the usual AI bros do not. The IP-hating libertarians are horrifyingly influential among legal theorists at George Mason University, and I fear that they may even have influence over legal theorists at the University of Chicago. As I have written in previous essays (1, 2), the libertarians’ case against intellectual property relies primarily on a straw-man argument most influentially delivered by a Chicago-school economist. And, as mentioned at the 36-minute, 44-second timestamp, the IP-hating libertarians even have a strong influence over Clarence Thomas on the U.S. Supreme Court.

The vast majority of people in the industrialized world — even ones who have their own complaints about copyright enforcement on websites like YouTube — know better than to swallow the rationalizations of IP-hating libertarians. But as long as they have the ear of legal theorists at George Mason University and a U.S. Supreme Court justice, such IP-hating libertarians still have the power to inflict serious damage upon artists. And these IP-hating libertarians exert influence upon academic legal theorists and jurists in a manner that the run-of-the-mill AI bro does not.

That is why I give special attention to IP-hating libertarians. It is also why, when IP-hating libertarians weigh in on the issue of regurgitative AI’s copyright violations, I implore all artists protective of their work to learn what separates the IP-hating libertarians from the AI bros. (By conventional “AI bros,” I mean the customers of AI companies. I concede that, with his cozying up to Donald Trump, billionaire OpenAI executive Sam Altman also maintains a lot of clout.)

I am writing this essay to explain how the whitewashing of regurgitative AI’s copyright desecrations by IP-hating libertarians is motivated by a particular ideological passive-mindedness — even bigotry — that is uncommon among the rationalizations of the AI brows. I intend to explain its origins and how this has led to the shockingly bizarre rationalizations of two respected libertarian journalists who, at first glance, seem to be the sort who should know better.

The Motives and Tactics of Those Who Actually Use Regurgitative AI — The “AI Bros” 

When it comes to AI bros who, on social media, very defensively rationalize their use of regurgitative AI and accuse objecting real artists of “gatekeeping,” the motivation and rationalizations are much simpler. These are people who once had a vague desire to create beautiful images from their own hands. However, they found such attempts too frustrating and discouraging, and they quit at it.

But when they first used regurgitative AI, it gave them the feeling that this was the closest they would get to knowing what it would be like for the implementation of their own aesthetic choices to result in a new visual image. Having given up on real visual art, this is the nearest that such people have felt to having come to acquire that sort of empowerment. But then real artists came and ruined that feeling by reproaching these AI bros’ use of the regurgitative AI. Hence, the AI bros have offered a plethora of rationalizations.

One of those rationalizations is that regurgitative AI “democratizes” the creation of visual art. By that standard, real artists who object to regurgitative AI are snooty elitists guilty of “gatekeeping.” But compared to many other pastimes, it’s relatively inexpensive to buy a pencil and paper and start drawing. Other than that the materials are usually purchased by their parents, this is something that is done by literally most four-year-olds. It’s difficult to name a pastime that has fewer and less obtrusive barriers to entry.

The fine arts are considered snooty and elitist in that, among those who use their hands to produce visual art, only a tiny minority have reached the point where the visual art can be called photorealistic. But the reason why this small number of people have reached that elite status is that they put in the effort to earn it. That is snooty and elitist only in the same way that anti-capitalists consider it snooty and elitist that, under free enterprise, the party that satisfies the marketplace demand of willing customers more successfully than others is the party that gets the richest.

Another rationalization is that regurgitative AI allows for disabled people to do art in a way that they could not have done in the more traditional methods. Actually, there are plenty of successful people in the fine arts who are disabled. And the AI bros who consistently make this argument, predictably enough, are people who are conventionally able-bodied.

And another favorite rationalization is one I have tackled before. It is that there is no such thing as artistic originality anyway, which makes art theft okay. In one of her “speed-paint storytime” YouTube videos, the online illustratrix “LavenderTowne” goes over this. Someone told her, “The fact remains that originality doesn’t exist. It’s something every type of artist knows and has known for a century.”

To that, LavenderTowne quips, “Dude, I missed that part in history class where 1920s flappers discovered that originality doesn’t exist.”

Following that quip, LavenderTowne does proceed to provide an argument that is more serious. With respect to the thought-terminating cliché that “All art makes use of already-established conventions, and therefore artistic originality has never existed,” you can also read my own refutation of it.

In her video, LavenderTowne also goes over tand refutes several other AI bro rationalizations as well. I recommend it.

In the case of AI bros, the rationalizations are about trying to justify morally an action they have already become accustomed to engaging in. By contrast, I would not be surprised if the IP-hating libertarians who are apologists for regurgitative AI do not even use it. Instead of defending an action they already practice, the IP-hating libertarians who defend regurgitative AI are acting out of a desire to to adhere to an ideological doctrine and dogma to which they have already pledged allegiance and are reluctant to let go of. In the case of the two IP-hating libertarian journalists I intend to single out, they pride themselves on veering away from much of the dogma from the 1970s, but the obsession with denouncing IP is something they still insist on holding onto.

The Half-Century-Old Party Line That IP-Hating Libertarians Will Not Give Up On
Here is a brief history on the IP-hating libertarians from the late twentieth century onward.

As usual, the most fanatical positions taken by self-described libertarians can be traced in the late twentieth century to an economist named Murray N. Rothbard. When the Libertarian Party formed in the 1970s, and Rothbard was still a part of it, he made known his opposition to patent rights. He vocalized, though did not originate, the falsehood that a U.S. utility patent is a government-enforced monopoly on an industry, claiming to hold exclusive ownership over a general idea for a whole category of product. The libertarian writer Roy A. Childs, Jr., too, is complicit in this.

Soon after, Wendy McElroy and Samuel Edward Conkin III — self-proclaimed “anarchists” like Rothbard — extended that denunciation and misconception to copyrighted artwork. They conflate an artwork — being a fleshed-out and fully rendered culmination of thousands of creative decisions — as a mere “idea,” as though it is the same in principle as a vague and hazy general idea not given a form by which it can be experienced through the senses.

They then put together the straw man that for an artist to claim ownership over the art she created — the art that is concretely experienced through interfacing with the senses — is to claim ownership over “ideas” in general, including vague hazy ideas of other people that are similar to the general premise of the artwork. As further stitches to their straw man, these IP-hating libertarians pronounce that for you to expect copyright protection over your original artwork is for you to attempt to use the government to police other people’s thoughts. In all seriousness, these IP-hating libertarians propound that copyrights are a form of censorship. One who embraced that standard would have to conclude that plagiarism is the exercise of free speech.

I have rebutted those falsehoods in essays here and here.

There have been two main strategies by which libertarians have tried to undermine intellectual property rights. The first is the blunter approach from the Ludwig von Mises Institute, the Foundation for Economic Education (FEE), and Liberty International (formerly the International Society for Individual Liberty). It consists of declaring outright that intellectual property rights are abhorrent and ought to be abolished altogether.

Frighteningly, this fanatical approach has been becoming more mainstream in recent years. We see this with two billionaires, Elon Musk and Twitter cofounder Jack Dorsey. Both of those men have stakes in regurgitative AI and have been hit with well-deserved copyright infringement lawsuits over it.

In response, Jack Dorsey bleated on Twitter, “Delete all IP law.”

To that, Elon Musk replied, “I agree.”

The folly of that tantrum is explained well by Raymond Van Dyke on the blog IP Watchdog, a blog I recommend highly.

Despite the cries for abolition of IP rights now being endorsed by famous billionaires, such fanaticism is, thankfully, still not entertained seriously by most people. For that reason, the libertarian think-tanks that are more sophisticated about the law and public policy exercise a strategy that is more subtle. This is what they do. They have lawyers keep tabs on court cases pertaining to IP enforcement, especially ones that have potential for setting new precedents on how far and how strictly IP enforcement can be applied in future cases. Then they come out with policy briefs and opinion pieces arguing that the court should deliver a ruling that sets a precedent to weaken the ability to enforce IP.

That is the modus operandi of the libertarian think tanks of Washington, D.C., especially visible (1, 2) from the Reason Foundation.

As they know better than to expect any immediate repeal of IP rights, their strategy is to weaken IP little by little.

The Rothbardians of the Mises Institute revile the Reason Foundation and the other D.C. think tanks for having lost interest in a number of Rothbard’s favorite talking points. Reason and the D.C. think tanks are not apologists for the Confederacy of the Civil War. Murray Rothbard, the Mises Institute, and Liberty International propagate the lie that the Confederation had the moral high ground over the Union. Reason and the D.C. think tanks are also not reflexively inclined to proclaim that every Third-World terrorist’s attack on the U.S. military is merely the innocent terrorist striking back against the U.S. military for being an evil empire. Worse to the Rothbardians, Reason and the D.C. think tanks are not full-throated in extolling the glories of anarcho-“capitalism” and “market” anarchy, an illogical and unjust model of society I refuted here.

And the Rothbardians of the Mises Institute fault Reason and the other D.C. think tanks for not calling for immediate destruction of IPRs. To the Mises Institute, the piecemeal approach of Reason and the other D.C. think tanks is a mealy-mouthed milquetoast position. Yet the approach of Reason and the other think tanks has made much more headway. As these think tanks’ adherents are strategically placed in the law and economics departments of George Mason University, it is the situation that Reason and the D.C. think tanks always have the ear of this college’s legal scholars.

As one of the most-cited straw-man arguments of these IP-hating libertarians also comes from a twentieth-century Chicago-school economist, I fear that the IP-hating libertarians’ rationalizations might have influence over the legal scholars of the University of Chicago as well. And intellectual property lawyer Adam Mossoff has pointed out that on account of GMU libertarians in general having an influence over U.S. Supreme Court justice Clarence Thomas, the rationalizations against IP rights have correspondingly influenced him. (Admittedly, with Clarence Thomas receiving undisclosed gifts from billionaires, I cannot rule out that Sam Altman and other AI executives might be a devil on his shoulder as well.)

The latest court cases that these IP-hating legal theorists have their eyes on are the copyright suits over regurgitative AI. That is why IP-hating libertarians have inserted themselves into arguments on Threads that online artists have been having against AI bros. Again, online artists too often mistake these IP-hating libertarians for AI bros. But many of these IP-hating libertarians do not use the regurgitative AI; their motives are different. And, again, the danger is that whereas internet-arguers who are customers to AI companies do not have sway over actual influential legal scholars, many IP-hating libertarians do.

Libertarian Journalists Nos. 1 and 2 Sounding Sane at First
That brings me to two libertarian journalists who come from the Reason Foundation sort of tradition. These two libertarian journalists are relatively well-respected, even by leftwing laymen who follow the news regularly. I will call them Journalist No. 1 and Journalist No. 2.

Journalist No. 1 ran a rather well-known website for philosophic essays pertaining to the libertarian movement and he now hosts a podcast for one of the more-benign remaining libertarian online periodicals, at which he serves a big editorial role. Unlike Reason magazine, this more-benign libertarian online periodical doesn’t plaster around Robby Soave’s smug smirk as he spreads COVID denialism and makes excuses for MAGA’s pathology and paranoia.

Journalist No. 2 once regularly wrote for Reason magazine. But, wisely, in 2024 on Threads he publicly condemned Reason magazine for its foolish whitewashing of MAGA. Journalist No. 2 noticed the psychological evasion on the part of Reason magazine’s editors in pretending that Kamala Harris’s fallacies on political economy and dunder-headed proposed government regulations were somehow at least as bad as — or somehow worse than — Donald Trump’s flagrant rejection of liberal republican Due Process and the rule of law.

There are big areas where both Journalists No. 1 and 2 have distanced themselves from many of the favorite public-policy positions of the Mises Institute and other worshipers of Murray Rothbard. Unlike most Rothbardians — including the vice president of international relations at Liberty International — these two do not whitewash the Confederation as having the moral high ground over Abraham Lincoln in the Civil War. They abhor the white supremacism of the Mises Institute. They criticize the Mises Institute’s bigotry concerning dark-skinned immigrants. And they take positions that most self-avowed libertarians — including those now running Reason magazine — detest as “Woke.”

It is to the credit of Journalist No. 1 that he dissents from transphobia. In attempt to ingratiate themselves to the Intellectual Dark Web that serves as apologist to Donald Trump, it has become near-uniform for libertarians to affirm agreement with all of the most popular denunciations of transgender people and their desire for equal treatment. Journalist No. 1, by contrast, has done podcast episodes where he has conducted interviews dispelling the misconceptions about transgender people that other libertarians have contributed to perpetuating.

Considering the many fanatical positions of the Rothbardians that Libertarian Journalists Nos. 1 and 2 have shied away from, it may be tempting to assume that Libertarian Journalists Nos. 1 and 2 are not fanatical themselves. Tragically, such an assumption would be wrong. The two remain devoted to the party line that the libertarian movement formed in the 1970s against intellectual property rights. They have bound themselves to the obligatory hostility to legitimate copyright enforcement. And they take the more gradualist position — rather than demand immediately the abolition of IP, they comment on the latest controversy over IP enforcement and then urge the weakening of that enforcement. Toeing that party line, Journalists No. 1 and 2 interject on the topic of regurgitative AI’s copyright infringements. And the argument they provide on it is astonishing in how ludicrous it is.

Libertarian Journalists Nos. 1 and 2 Bizarrely Misunderstanding Human Memory in Particular and Human Beings in General
First, Libertarian Journalists 1 and 2 make note of how, in scraping data of online images, the regurgitative AI will make and store an exact copy of the copyrighted image against the copyright holder’s consent. Then, in a very specious obscuring of context, Libertarian Journalists 1 and 2 say that that is exactly what every copyright-holding human artist does. They say that in learning to do fine art, the artist saw other copyrighted images and that her memory produced exact copies of the copyrighted images that she stored in her memory. Then, continue Libertarian Journalists Nos. 1 and 2, the artist’s own original works are simply a similar statistics-based reconstituting of other artists’ copyrighted works in the same manner that the regurgitative AI’s regurgitative image is a reconstituting of the images that the AI scraped.

Therefore, conclude Libertarian Journalists Nos. 1 and 2, artists who complain about their copyrights being infringed are hypocrites: the AI is doing exactly what those artists do. The implication is that all art is an unoriginal derivative of someone else’s, and therefore actual copyrighted art by human beings is not any less of a copyright infringement than is the output of regurgitative AI. That is not so much to defend regurgitative AI than it is to demean and undermine confidence in the recognition that any artwork deserves strict copyright enforcement at all. Again, the goal is not to call for the abolition of copyrights right away but instead to set a precedent that weakens confidence in, and the ability to enforce, copyrights in the future.

Libertarian Journalist No. 1 has made these assertions repeatedly on his blog and on Threads. Here is one example on the latter. He begins with the insincere hook — which all of his other online writings on this topic belie — “I’m sympathetic to the concern that there is something wrong with the way AI models are trained.” Then he gets to what he actually wants to convey.

But I’ve yet to see a persuasive case that (1) training them specifically constitutes theft of the training materials in a way that (2) doesn’t also entail that a human watching a bunch of movies, internalizing what they’ve learned, and creating new work inspired by them isn’t theft.

Many plainly convincing cases had already been made to Libertarian Journalist No. 1. He just wrote off those convincing cases immediately and arbitrarily because they did not match the conclusion to which he is already married. We can be diplomatic and say that his criteria for what does and doesn’t constitute something “convincing” are . . . idiosyncratic.

To his comrade, Libertarian Journalist No. 2 chimes in, 

My brain contains copies of many, many copyrighted works at a sufficient level of fidelity that they’d be infringing if I could somehow telepathically beam them to a hard drive. We just very reasonably choose not to apply copyright to brains.

Note the far-fetched boast on the part of Libertarian Journalist No. 2 that his own conscious autobiographical memory of a copyrighted artwork is exact to the same degree — or, as he says, “at a sufficient level of fidelity” — as a perfect digital copy of it. The dubious nature of Libertarian Journalist No. 2’s braggadocio, and that Libertarian Journalist No. 1 is not taken aback by a claim so stark in how it defies credulity, surprised me. Upon my initial reading of that exchange, I felt vicarious embarrassment for them both.

This argument from Journalists No. 1 and 2 is so out of touch with even the most rudimentary understanding of human psychology, artistic ability, and the body’s motor functions inhering in art-making, that it makes Journalists No. 1 and 2 themselves sound not like humans but like an AI — and an AI of an already-obsolete model, ready for the junk heap, at that.

How Human Memory Actually Works, and How That Makes It a Less Effective Tool for Copyright Infringement Than Does Regurgitative AI 

A strong and relatively vivid memory of a particular sight, sound, or other sensation is part of what is called autobiographic memory. If you have committed to heart the plot and dialogue from a beloved movie, that is autobiographic memory at work. More specifically, that involves semantic memory, which is general knowledge, and episodic memory, which is memory of events and the sensations relating to such experiences. But much of those, especially the latter, are subsumed into the larger category of autobiographic memory. With exceptions that are freakishly rare — and I will address that later — even the people with the strongest autobiographic, semantic, and episodic memories cannot produce in their minds a replica of someone else’s art to the exactitude that Libertarian Journalist No. 2 blithely assumes. That is the reason why there is such a psychological phenomenon as the Mandela Effect, where people think they remember a detail of the past so strongly and yet that detail turns out inaccurate. A famous example is that they strongly remember the children’s books being titled The Berenstein Bears instead of The Berenstain Bears with an a in the -stain where they expected an e to be. Even real artists with very strong autobiographic, semantic, and episodic memories are often bedeviled by the Mandela Effect, giving the lie to Libertarian Journalist No. 2’s conceit that his “brain contains copies of many, many copyrighted works at a sufficient level of fidelity that they’d be infringing” on copyright if he could “telepathically beam” those memories “to a hard drive.”

As I will explain soon, even the extremely rare people with memories as strong as what Libertarian Journalist No. 2 claims to have are people who cannot reproduce other people’s artworks manually with their own paint brushes and other art supplies if they have not practiced, for years, the same artistic medium as any work that they could possibly intend to copy. That is because for an actual artist to reproduce another’s work with the same degree of fidelity as the regurgitative AI is capable, the would-be copycat artist must fulfill an additional requirement. For the would-be copycat artist to succeed at being a copycat, she must employ another form of memory that Libertarian Journalists No. 1 and 2 have not mentioned, and which they are apparently trying to conflate with autobiographic memory.

There is an extremely small number of people on Earth — fewer than one-hundred documented — that do have perfect autobiographic, semantic, and episodic memories. And at least one very high-profile one even is an artist. That is the actress Marilu Henner. She has proven to psychologists that she is able to remember the exact and minute details of events she observed from decades past. She has given details of memories of a particular day’s events from years earlier and, when psychologists have investigated those details, they always check out.

I doubt, however, that Libertarian Journalist No. 2 is among one of those one-hundred people with perfect autobiographic, semantic, and episodic memories. The chances of that are less than 1 in 80 million.

And even with her perfect autobiographic memory, if Marilu Henner tried to produce a perfect copy of a work she observed that involved artistic media other than acting, and which she had not been practicing for years, she would not succeed. That is on account of the fact that artists who have strong autobiographic, semantic, and episodic memories still rely on another form of memory both in their original works and in imitations — another form of memory, one that that Libertarian Journalists Nos. 1 and 2 have ignored, and which regurgitative AI bypasses in its own imitations of copyrighted works.

That other form of memory is “muscle memory.” Autobiographic, semantic, and episodic memory are held on the conscious level, and that is not enough for an artist to reproduce manually the work of another. That, and all professional-level real art, requires the acquisition and exercise of muscle memory not through conscious memorizing with one’s cerebrum but subconsciously through consistent practice as the artist moves other parts of her body.

Consider the “Get What You Deserve” speech that Joaquim Phoenix delivers in the movie Joker. Many of the movie’s fans felt inspired by that speech and committed it to memory — their autobiographic, semenatic, and episodic memories. If I committed that speech to memory and video-recorded myself reciting it, it would not duplicate Joaquim Phoenix’s performance in such a manner that Warner Brothers would identify it as infringement on the movie studio’s copyright. Even aside from my inborn physical differences from Joachin Phoenix, such as my nose being shaped differently from his, my reciting the speech would result in my using a different tone, in my vocal inflections being different, and the gestures of my upper body being dissimilar.

Comparing, side by side, Joachim Phoenix’s performance against my verbatim recitation of his speech, you would notice that Phoenix is a pro whereas I am worse than an amateur. All of those nuances, which eventually add up to a powerful effect, are part of the muscle memory for which Libertarian Journalists Nos. 1 and 2 have neglected to account.

If, through his own muscle memory and use of traditional art supplies like paintbrushes, Professional Artist No. 2 did successfully produce an exact copy of Professional Artist No. 1’s copyrighted work against her consent, then Artist No. 1 would be right to sue him. But the costly demands of an actual artist building up muscle memory to that point has made it relatively unusual for real artists to infringe upon one another so blatantly. Hence, while this sort of unauthorized duplication has always been a problem and worry, it was something that, relatively speaking, real artists have not had to worry as much about. Note that with both her perfect autobiographic memory and acting experience, Marilu Henner is one of the handful of people on Earth who could reproduce Joachim Phoenix’s body language perfectly if she delivered her own rendition of his “Get What You Deserve” speech. And note that even she does not infringe on copyright in the way that Libertarian Journalist No. 2 insinuates that all artists do in practice.

By contrast, regurgitative AI bypasses the limitations of muscle memory in humans. For that reason, regurgitative AI makes it cheaper to infringe on copyrighted works through accurate duplications, making it likelier that these forms of piracy will occur on a much larger scale than before. As a consequence, real artists — whom Libertarian Journalist No. 1 tries to gaslight for being “arrogant” gatekeepers — are, in fact, entirely reasonable in worrying about abuses from this new threat.

IP-Hating Libertarians Getting Something They Deserve
Fortunately the tech journalist Brian Penny, whom I think is politically center-Left, gives Libertarian Journalists Nos. 1 and 2 the ridicule they have rightfully earned. In reply he posts,

Absolutely nobody in the history of humanity ever once learned anything by consuming billions to trillions of hours of anything and then magically becoming an instant expert. That is a myth, and it’s a dumb one. You’ve surely read at least one book or heard one interview in your life talking about people practicing and getting better over time by doing, not by sitting still and binging billions to trillions of hours. . . . You can’t be serious.

As for stealing, go into Walmart right now and start reading every book or watching every DVD. Report back how many you get through before they approach you for theft. Your perspective is too derpy...even [to] be believable.🤦‍♂️🤦‍♂️🤦‍♂️

To reinforce the point further, Brian Penny had some visual aid. On the right for a Threads post he had a professional artist’s illustration of the Marvel Comics character Wolverine. On the left was Mr. Penny’s own amateurish and crude drawing of the same character — the sort he would have done at age four. With that illustration, Mr. Penny calls attention,

If watching a bunch of X-Men movies and the animated series and reading their comics and playing with their action figures growing up made me an artist, then why does my Wolverine look like an underdeveloped child drew it?

Mine is the head on the left 😹😹😹 the one on the right was what AI bros keep telling me I need to stare at to learn like a human.

Mr. Penny was correct in all of that except for his assumption that Libertarian Journalists Nos. 1 and 2 are conventional AI bros rather than ideologues who don’t even use regurgitative AI and are instead still trapping themselves in this part of Rothbardian dogma from the 1970s. When I first read the works of Libertarian Journalists Nos. 1 and 2, I thought they were too smart to believe in something as absurd as what they have stated seriously on this topic. But rationalizations so absurd are what allegedly intelligent people fall into when they will not release themselves from this dogma that does nothing but begrudge our intellectual property rights.

Libertarian Journalist No. 1 has also repeatedly said that regurgitative AI imagery is the same, in principle, as search engines sending out bots and crawlers over websites to gather data on them, all without the permission of those websites. That, too, is false. When owners of websites withhold permission from search engines to obtain particular data, the websites’ codes contain what are called robots.txt exclusion protocols. When a website possesses a robots.txt protocol, the search engine abides by it. By contrast, as shown in this NPR piece, the regurgitative AI’s bots flagrantly disregard and bypass the robots.txt protocol. This indicates that the businessmen who own the regurgitative AI are aware that they do not have permission to scrape data, and then they do it anyway. No, Libertarian Journalist No. 1, contrary to your assumptions, it is not the same in principle at all.

Again, one of the AI bros’ favorite rationalizations is that “AI democratizes art.” It definitely has not, but that word does describe something related that is far more important and far more interesting. Digital technology truly has enabled a human being of relatively modest financial means to create and upload digital content. On account of such technology, the number of people creating their own genuinely original videos, and uploading those videos, has increased substantially since the year 2005. To the degree that the digital technology enabled that person to be the party making most of the creative choices directly, that technology has truly has helped to democratize content-creation. First dibs on ownership of such a piece of content must rightfully go to its creator. Hence, insofar as it has helped regular people — not only corporation-backed professionals — produce and distribute content, the technology has helped democratize copyrights themselves, the same copyrights that AI firms are now trying to deny.

In short, digital technology — AI and all the rest — has not democratized artistic skill. Artistic skill remains the exclusive province of those who put in the years of practice. Digital technology has, however, democratized intellectual property rights themselves, the very property rights that AI bros and libertarians wish to erase.

Use of this regurgitative AI will not be ethical until the data on which it “trains” come exclusively from copyright holders who offer explicit permission in an opt-in system. That is, absent of the explicit permission, the data are not to be used. Until such time, we can only hope that, in defiance of Libertarian Journalists Nos. 1 and 2 and the Reason Foundation’s essays urging otherwise, actual reason prevails and the artists win their civil suits against the corporate owners of the regurgitative AI.