Artificial Intelligence and the Chaotic Marketplace of Iranian Publishing

۱. You have previously emphasized that Artificial Intelligence ultimately reduces to large language structures—devoid of embodiment, agency, subjectivity, emotion, and authentic human movement—whereas the foundational unit of the humanities and social sciences is human action (praxis), not merely words or abstract thoughts. What is your overall approach to AI within this framework?

I have always rejected absolute anti-technological stances or naive forms of technophobia. Human capability (homo sapiens) as both a tool-making agent (homo faber) and a speaking being (homo loquax) is precisely what enabled humanity to evolve into its present state. Yet, as Jean-Jacques Rousseau noted, this humanity has undergone a rupture with raw nature, placing itself within an “urban” or “civilizational” existential condition. Consequently, human life is frequently fraught with semantic impasses: overconsumption and relentless pressure upon the natural environment have become central to human existence, rendering the capitalist system seemingly inescapable.
However, the “technological” artifact—as the material dimension of culture—does not inherently or automatically conflict with nature, nor does it replace it; rather, it constitutes an intrinsic part of nature itself. Just as human tool-making and linguistic capacities do not stand in a state of absolute rupture with the animal and natural world, but rather in continuity with it. Animal behaviorists such as Frans de Waal [1] and moral philosophers like Élisabeth de Fontenay [2] have long underscored that animals possess intelligence, emotional capacities, empathy, and even rudimentary forms of ethics. If we conceptualize technology as an element of culture situated along a continuum between the animal and the human, there is no rational justification for technophobia.
What is genuinely alarming is not technology per se, but the malicious intent of human actors within political and capitalist structures who either weaponize technology to inflict mutual destruction (such as armaments and warfare technologies) or exploit it through predatory, mercenary, and criminal practices. Martin Heidegger consistently maintained that he harbored no hostility toward technology, arguing that “technology does not think.” It is human beings who think, engaging with the world—whether with nature or with fellow humans—in ways that can be poetic, creative, and aesthetic, or conversely, destructive, utilitarian, and violent. The fact that capitalism has today consolidated its position as the sole viable mode of human interaction with the world is not due to the absence of alternative paradigms, but rather to the deliberate efforts of dominant powers to render all alternatives practically unviable. As Patrick Chamoiseau (the Prix Goncourt-winning author), in a dialogue with Edwy Plenel, eloquently and from a profoundly humanistic and ethical rather than a conventional political standpoint, has analyzed and demonstrated [3].
Therefore, since the inception of the computer revolution—which subsequently evolved into the internet, search engines, and today into Artificial Intelligence, and perhaps tomorrow toward “Artificial General Intelligence” (AGI)—I have not perceived an inherent peril in the technology itself. We are not dealing with an atomic, hydrogen, or biological weapon whose sole conceivable telos is death, devastation, and suffering. The applications of AI, following in the lineage of advanced computational tools like sophisticated search engines, possess infinite potential for democratizing knowledge, advancing the sciences (particularly medicine and environmental studies), and enhancing the well-being of humans and other living entities.
The genuine peril resides in human agents who, driven by political and corporate power, transform this instrument into a hazardous technology that threatens human dignity, the very preservation of human thought and emotion, and inflicts ever-escalating damage upon the natural environment—a trajectory they have actively pursued thus far. For this reason, I personally utilize AI on a continuous basis as an assistant. Several months ago, I published my first AI-assisted dialogue regarding the nexus between violence and democracy in Fararu (June 8, 2026) [4]. Naturally, any engagement with AI must always be grounded in rigorous domain knowledge and an unwavering critical perspective. The arguments I wish to present here are formulated upon this exact premise. Finally, to properly frame contemporary debates surrounding AI, we must also confront two additional crises: first, what is designated as “AI Hallucination”—the hazard wherein output morphs into a species of “quasi-truth” driven by statistical density and lexical frequency; and second, the phenomenon known as “Model Collapse,” whereby AI systems iteratively train on data generated by other artificial models (including machine translation), thereby degrading final output into artificial, low-quality, and superficial artifacts.

۲. Against this backdrop, how does the release of a collection of philosophical, psychological, and scientific books translated predominantly via AI and subsequently subjected to human editing (exemplified by Hermes Publications) transform the ontological status of “translation as cultural praxis”?

Regarding these publications, I should first note that while this specific method has not been widely implemented in Iran until now, it has several precedents globally, including initiatives by Taylor & Francis [5], Daedalus Books Turkey [6], and Amazon [7]. This approach is typically deployed under conditions where a significant volume of literature requires translation, yet the availability of qualified human translators is critically scarce. Under normal circumstances, however, translation ought to be fundamentally executed by human translators; this remains the primary condition for quality, which can certainly be augmented and refined through a critical, judicious use of AI. Furthermore, this in no way invalidates the indispensable role of the human editor, whose intervention further elevates the scholarly rigor of the text.
Nevertheless, two fundamental problems render the uncritical deployment of this technology hazardous. The first problem is that the logic of capital—or what Theodor Adorno termed the logic of the “culture industry,” the commodification of culture and art—either has supplanted or seeks to supplant the logic of thought, art, and creativity. In translating or authoring a book or an artwork, the ultimate objective ceases to be the cultivation of a meaningful (spiritual) connection with the world, nature, and fellow human beings in pursuit of a biocentric existence; rather, the goal is profit maximization achieved through the drastic compression of labor time.
It is evident that if a human translation requires three to six months—representing, say, one to two thousand hours of labor or more—capitalist logic dictates that AI should compress this temporal span a hundredfold or more, dividing labor costs by a hundred and exponentially multiplying profit margins. Viewed strictly through the lens of profitability, the calculus appears entirely rational, yet it completely disregards quality, violated human and ecological values, and their profound cultural, economic, political, and social ramifications.
Worse still, when capitalist profit-seeking amalgamates with lethal ambitions of power, we arrive at harrowing, dystopian scenarios (such as selecting military bombing targets via algorithmic optimization), which lie beyond our immediate purview here. My point, however, is that the legitimizing discourse presenting AI as a mechanism for quality enhancement is predominantly a reality-manipulating ideology. For this very reason, in virtually all international precedents I cited earlier, translators’ and editors’ unions mobilized to defend their labor rights. They argued correctly that AI leverages their accumulated cognitive labor and lived experience to forge tools that, bypassing rightful ownership and remuneration, subordinate human expertise to corporate capital and publisher profitability—while cynically marketing this process under the rubric of “cultural enhancement.”
When we turn our gaze toward the domestic cultural landscape of Iran, we must candidly acknowledge that the publishing ecosystem in which these operations unfold is severely traumatized and pathogenous, crippled by the absence of copyright legislation and healthy institutional relations. Owing to the vacuum or non-enforcement of laws, international isolation, and the unethical conduct of numerous publishers, authors, and translators, Iran’s publishing market is marked by acute turbulence—a structural pathology criticized for years by scholars and cultural practitioners alike.
If yesterday the afflictions of this market were pirated translations, illicit book distribution, plagiarism, book-packaging (ketabsazi), and yellow journalism, today these enduring maladies have been compounded by the pathology of unregulated AI usage. We continuously witness novice translators delegating their unearned labor to AI applications and merely “giving a superficial polish” to the machine output. Under these conditions, one must exercise extreme vigilance: when a reputable publishing house—historically a benchmark for quality—adopts the systematic use of AI, regardless of its professed positive intentions, it issues a dangerous green light for further malpractice and degradation in the application of a modern, otherwise beneficial technology.
Under such circumstances, we are no longer dealing with “translation as cultural praxis”—just as we may no longer be confronting “authorship as cultural praxis.” Instead, we encounter utilitarian or power-driven maneuvers operating within the exact logic of that sinister and perilous nexus of capitalism and technology, masked by populist rhetoric touting enhanced accessibility, democratization of knowledge, and efficiency. The intellectual as an active agent is thereby either marginalized, annihilated, or driven toward a state of self-inflicted idiocy.

۳. In previous seminars, you noted that while AI can process data at unprecedented speed and scale, interpreting human relationships, deep social meanings, and cultural contexts still necessitates the cognitive frameworks of the humanities. Does the precipitous rush of publishers into machine translation of complex texts not risk reinforcing the illusion that linguistic data processing can substitute for human interpretation? What are the implications for the status of the humanities in the production of knowledge in the Persian language?

Precisely so. The relationship between a translator and a translated text, or between an author and a scholarly or literary work, mirrors the relationship (or ideally, ought to mirror it) that an artist maintains with their creative medium. Indeed, if we move beyond the strictly textual framework into the domains of poetry, literature, and the visual arts, this dynamic becomes even more transparent.
If one imagines that advanced AI cannot imitate a poet, sculptor, painter, or thinker to generate novel or seemingly “inventive” samples, that assumption is instantly shattered by the realities observable online and the capabilities accessible to anyone via emerging AI applications. Today, AI can compose music and poetry, produce photographs and imagery, and write or translate books so convincingly that human observers struggle to detect their non-human provenance. In a recent interview with The Economist [8], Yuval Noah Harari remarked that numerous fabricated audio-visual exchanges attributed to him are circulated online—some sent by friends and family asking whether they are authentic. He recounted watching one such deepfake and requiring a full minute and a half to verify that it was entirely artificially generated.
Several AI-generated singers, musicians, and actors have also emerged across social networks, amassing substantial followings. According to a Guardian report from March 2026 [9], an Iranian AI-generated virtual singer named “Nava” garnered over 13 million views on Instagram alone.
This brings us directly to the timeless aesthetic debate concerning the “original” versus the “simulacrum,” compelling us to recall Walter Benjamin’s seminal 1935 essay, The Work of Art in the Age of Mechanical Reproduction [10]. Benjamin underscores the centrality of the process of praxis in artistic production—a vital dimension dissolved in mechanical reproduction, transforming art into an “industrial” commodity. Similarly, closer to home, Abbas Kiarostami’s exquisite film Certified Copy (2010) explored this exact thematic terrain.
This discourse has persisted into the digital and algorithmic era, as engaged by theorists such as McLachlan [11], Miao [12], and Bianchi [13]. Yet we must push beyond these paradigms. I conceptualize the “human” dimension through the lens of sublimation—namely, a commitment to biocentrism encompassing the agent, time, space, and deep historical, mnemonic, and socio-political matrices.
Large Language Models (LLMs) underpinning AI do not “think.” Although our precise neurobiological definition of “thought” remains incomplete—understood as a complex matrix of electrochemical synaptic interactions [14]—we cannot bypass the foundational reality of biological embodiment, mnemonic situatedness, and sensory grounding. Consequently, becoming an artist, poet, translator, or writer invariably entails an intricate tapestry of sensory and experiential narratives—in a word, poetics and aesthetics—synthesized with cognitive and cultural matrices.
This leads directly to Roland Barthes’ concept of the “Death of the Author” (1967) [15]: once an artwork is detached from its creator, it assumes an independent life. However, what must be added here is that such an artifact (whether a translation, object, painting, or text) remains comprehensible, experiential, and meaningful only within a framework of human narratives. When a reader engages a translation, they re-create it through their prior knowledge of the translator and their oeuvre, their understanding of the original author, the collective culture of those who contributed to its creation, and ultimately through the prism of their own personality, language, culture, and lived experience.
If this task is surrendered entirely to human-absent AI—even under the guise of the Hermes experiment, substituting human translators with “specialized teams in linguistics, literature, and philosophy”—we are effectively executing a transition from the “original” to the “simulacrum.” Many readers purchase a work based on the “name” of the author, translator, publisher, or even its physical typography. When these signifiers possess grounded, authentic, or validated “lived” references, they hold efficacy. Replacing a reputable human translator’s name with “Artificial Intelligence” inflicts the same epistemic injury upon a book as attending an exhibition of AI-generated paintings and sculptures, reading machine-composed poetry, watching algorithmic films, or listening to synthetic symphonies.
Such phenomena possess novelty value initially (as observed with the phenomenon of “Nava”) driven by curiosity and an aura of mystique, but they are subsequently abandoned with staggering swiftness—much like the perceptual gulf between a genuine gemstone and its synthetic replication.
To grasp this trajectory more clearly, we may transition from cultural commodities to existential realities. How many human beings would willingly consume “synthetic” meat, vegetables, and fruit (even if manufactured under factory conditions) possessing an ontologically distinct essence, in lieu of authentic agrarian produce? An experiment conducted decades ago during the advent of agricultural genetic engineering revealed that when manufacturers, pressured by public opinion and statutory regulations, were legally mandated to label genetically modified products (“Contains a bioengineered food ingredient”), they suffered severe consumer backlash. The U.S. Congress enacted the National Bioengineered Food Disclosure Standard in 2016, designating the USDA to enforce mandatory disclosure effective January 1, 2022—decades after the underlying technologies emerged in the 1970s and 1980s and commercial tomatoes hit the market in 1994. Consequently, food producers preferred absolute opacity regarding origins.
In my view, comparing genetically engineered foods to “natural” foods (even if farm-cultivated) serves as a much closer analogue to comparing authentic human translation with machine-engineered AI translation than comparing natural produce to a synthetic tablet or chemical pill. This very consumer resistance forced a restructuring of market strategies, giving rise to the “bio/organic” market niche—itself a subject of separate critical evaluation. In light of this, I maintain that any cultural, artistic, literary, or scientific undertaking involving AI must, alongside rigorous human and algorithmic oversight, explicitly and transparently disclose its artificial genesis within the work itself.

۴. You have consistently prioritized culture over technology, viewing purely techno-economic development as insufficient and frequently pathogenic. Proponents of this publishing project argue that active engagement with AI is imperative to fortify the Persian language infrastructure and prevent cultural consumerism. To what extent is this rationale compatible with the paradigm of “culture as the axis of development,” and to what extent might it inadvertently reproduce a technocentric logic?

Even a rudimentary familiarity with AI technology, linguistics, and translation workflows exposes the fallacy of this argument. Publishing AI-translated texts under the pretext of aiding the Persian language’s advancement within AI ecosystems or bolstering its digital representation would only hold validity if raw, machine-generated outputs were systematically juxtaposed alongside finalized human revisions. In other words, such projects would need to function as empirical research initiatives designed to map translation errors and propose linguistic solutions to optimize AI applications.
Even then, such an endeavor holds marginal utility, as massive AI conglomerates like Claude, ChatGPT, and Google (Gemini) already accomplish this autonomously through billions of daily user interactions across all languages, including Persian—a core rationale underpinning the provision of their free public interfaces. Consequently, independent publishing initiatives are redundant for this purpose.
Within Iran’s chaotic marketplace, where the intellectual property and labor rights of translators, authors, and writers are systematically trampled, I categorically oppose practices that harvest the fruits of their arduous intellectual lives via applications without granting them rightful recompense. This is especially egregious when championed by a reputable publisher, as it establishes a pernicious precedent. Unscrupulous publishers already flood the market daily with hundreds of substandard or ghost-written titles. For years, such entities have acquired reputable translations, commissioned superficial cosmetic alterations, and reissued them under pseudonymous or disreputable attributions. Today, wandering through street vendors or even established bookstores reveals multiple unrecognized, spurious translations of major global literary classics, often bearing fabricated authorial or editorial names.

۵. In your analyses of the information revolution and the algorithmic city, you have warned against the erosion of direct sensory relations, the atrophy of affect, and the potential subjugation of humanity to machines. How might the machine translation of intellectual texts, subsequently marketed as cultural commodities, accelerate this alienation of human experience from language and meaning, or conversely, generate an opening for critical reflection?

Here we confront the imminent hazard of not merely institutional market collapse, but the systemic implosion of human cognitive and creative structures. According to visionary scientists and pioneers like Geoffrey Hinton, AI may, within a span of decades or sooner, automate virtually all human labor. With the advent of Artificial General Intelligence (AGI), humans will definitively lose the capacity to compete with machine intelligence, even collectively, raising the terrifying prospect of humanity forfeiting control over the technological trajectory itself.
The prevailing scenarios are overwhelmingly dystopian: the annihilation of nature and humanity by autonomous AI systems—echoing Stanley Kubrick’s prophetic cinematic masterpiece, 2001: A Space Odyssey (1968). The dilemma is not that these scientists fail to perceive the precipice toward which they are marching; rather, they are trapped in a competitive corporate arms race driven by systemic market pressures and geopolitical power struggles. Political and economic power centers cannot halt this momentum out of fear of falling behind and being subjugated by rival states fielding superior intelligence architectures.
This underscores the vital necessity of democratic public intervention and regulatory governance. Some compare this imperative to the moratoria on human cloning (at least within non-military domains) or international bans on biological and genetic warfare—despite major powers secretly continuing such research, as perhaps witnessed during the COVID-19 pandemic.
Consequently, cultural degradation cannot compartmentalized from societal collapse at large. If material production witnesses mass unemployment among engineers, physicians, laborers, nurses, accountants, and administrators displaced by machinery—culminating in endemic poverty, social unrest, political violence, disease, and depression—the intellectual sphere will simultaneously suffer a profound descent into cognitive atrophy, the evisceration of rationality, and the extinction of artistic and literary creativity.
Predictably, representatives of political and financial oligarchy, such as figures originating from the radical right like Elon Musk, propagate absurd, populist falsehoods: claiming that AI will imminently (within a decade or two) ensure that displaced populations reap the boundless dividends of automated factory production, which will drive commodity prices to zero. They peddle fantasies of radical life extension up to 150 years, the eradication of all disease, and eternal youth, promising a future where humanity enjoys a life of unalloyed leisure and abundance.
The absurdity of such infantile assertions is laid bare by the reality of our contemporary world, saturated as it is with geopolitical conflict, war, destitution, cruelty, devastation, and violence. Africa possesses some of the most fertile lands on earth, yet it concurrently witnesses the gravest humanitarian tragedies. Yet, psychological manipulation pits these impoverished populations against one another, weaponizing hatred and violence to safeguard entrenched oligarchic interests.
In summary, if AI advances unchecked, guided solely by the imperatives of capital and power, we will inhabit a global society populated by insensate, uncritical, subservient subjects utterly stripped of aesthetic, poetic, artistic, and cognitive capacities—their gazes permanently locked onto mobile screens or wrists functioning as display monitors, hypnotized by fragmented, ultra-short video micro-content akin to TikTok and low-value network products. As some analysts suggest, this dystopian reality has arguably already materialized in contexts such as China.
۶. You have previously argued that AI functions constructively only when subordinated to human agency within democratic socio-political frameworks. Given that primary control over Large Language Models is concentrated within hyper-capitalist tech monopolies valued at trillions of dollars, to what extent can Iranian publishers entering this domain genuinely preserve “human agency,” and to what extent does it amplify structural dependency on capital logic and exogenous algorithms?
In my view, this is a profoundly misguided endeavor that exacerbates pre-existing systemic vulnerabilities. I sincerely hope that such practices—particularly when initiated by esteemed publishing houses—do not pass into normalization through the complicit silence of cultural actors. We are gambling with the future of culture and art.
Admittedly, I retain deep faith in the resilient adaptive capacities of living systems, and as noted, similar initiatives launched internationally were met with swift failure. I anticipate initial superficial applause for these domestic experiments, but over time, human sensory and biological integrity will reject according machines higher epistemic authority than human thought—unless human thought itself has already been systematically degraded and stripped of its democratic liberties.
In a society where the majority ceases to ascribe value to art, literature, beauty, and creativity, it becomes irrelevant whether translations, texts, paintings, or musical compositions are human-crafted or machine-generated. Yet therein lies the inherent contradiction: such a society ceases to be inhabited by sensitive, refined artists and intellectuals, populated instead by careerists driven solely by wealth, power, and status, viewing art and literature merely as ornamental luxury goods. The extent to which we have already approximated this dystopian future, both globally and domestically, dictates how far the uncritical expansion of AI will penetrate our cultural life.

۷. In your historical comparisons involving the advent of the internet and the web, you highlighted initial academic resistance followed by gradual accommodation. Do you view the contemporary phenomenon of AI translation as a continuation of that historical trajectory, or do you discern a qualitative divergence requiring a fundamentally distinct critical posture from anthropologists and cultural actors?

Indeed, historical resistance followed by accommodation has characterized every technological paradigm shift; human life undergoes transformation while the positive utilities of technologies are integrated, albeit without a coherent systemic framework capable of neutralizing their inherent structural hazards.
What distinguishes Artificial Intelligence is its property of “machine agency”—a qualitative novelty fundamentally breaking with preceding technologies. Consequently, AI harbors unprecedented latent dangers. When coupled with malevolent human actors (capital and state power), these latent perils are actualized, steering societies toward the dystopian models envisioned by Aldous Huxley in Brave New World (1932). This represents the optimistic totalitarian scenario, counterbalanced by George Orwell’s darker vision in 1984 (1948).
Whether the future world comprises complacent consumer drones satisfied with biological subsistence or brutalized slaves bearing the lash of capital and power, it will undeniably constitute an existence devoid of cognition, liberty, aesthetic pleasure, poetic vitality, and sensory refinement. Yet, assuming that humanity is locked into an inescapable binary choice between these dystopian models is entirely erroneous. The future is invariably forged by those who seize control of their destiny and refuse to capitulate to existing conditions.
I harbor a profound conviction in the vital resilience of living creatures and nature. Therefore, despite prevailing despairs, I maintain that hope must be preserved, for it constitutes the sole authentic spirit of life.

۸. Finally, reiterating your insistence that the humanities must anchor human agency and treat technology strictly as a subservient tool, what normative framework or guiding principles do you propose for the responsible deployment of AI in translation and cultural knowledge production? What is the role of intellectuals and academic institutions in this endeavor?

We must systematically expand and organize human agency through civic actors—including academics, artists, intellectuals, students, and everyday consumers of ideas—thereby constraining and regulating the imposition of top-down centralized power. There is no arcane secret: the more robust democratic freedoms, public oversight, and civic control over politics and capital become, the healthier our societal condition.
Centuries of historical experience demonstrate that unchecked political and economic power inexorably degenerates away from the common good. The historical ledger of twentieth-century totalitarianisms—from global colonialism to state fascism and state communism—has repeatedly validated this truth.
The core crisis does not reside within AI technology itself, but in its capacity to amplify the interactive and destructive momentum of speculative capital and predatory power to a threshold capable of annihilating not only themselves, but the entire human and natural ecosphere. The harrowing symptoms of this ecological and social devastation are already visible across every corner of our daily lives.
Interview with Fararu regarding Artificial Intelligence

September 2026

References:
۱. de Waal, Frans B. M., 2001, The Ape and the Sushi Master: Cultural Reflections by a Primatologist. New York: Basic Books.
۲. de Fontenay, Élisabeth, 1998, Le silence des bêtes : La philosophie à l’épreuve de l’animalité. Paris: Fayard.
۳. www.youtube.com/watch?v=Oi78ubZtwYs
۴. https://fararu.com/fa/news/977739
۵. Taylor & Francis Group AI-assisted publishing policies and guidelines.
۶. Daedalus Books translation initiatives.
۷. Amazon automated and assisted translation publishing programs.
۸. Yuval Noah Harari interview with The Economist regarding deepfakes and AI synthetic media.
۹. The Guardian, March 2026 report on virtual AI artists and influencers (e.g., “Nava”).
۱۰. Benjamin, Walter, 1935, The Work of Art in the Age of Mechanical Reproduction.
۱۱. McLachlan, philosophical perspectives on AI simulacra.
۱۲. Miao, contemporary aesthetics and digital media studies.
۱۳. Bianchi, research on algorithmic ontology and originality.
۱۴. Neuroscientific definitions of synaptic electrochemical interactions and consciousness.
۱۵. Barthes, Roland, 1967, The Death of the Author.

English translation generated by A.I. Original persian interveiw:

هوش مصنوعی و آشفته بازار نشر ِ ما