AI Convergence on Human Self-Narration

A methodological claim that multiple independent AI systems, when prompted to analyze human-written content, converge on the finding that human self-narration is systematically optimized to present competitive organisms as morally governed and publicly oriented.

AI Convergence on Human Self-Narration is a methodological claim advanced by Steve Hargadon based on experiments he conducted with multiple artificial intelligence systems analyzing human-written content. The finding emerged from Hargadon's broader investigation into what he terms the Law of Inevitable Exploitation (L.I.E.) and his framework of realmotiv.

The Experimental Method

In developing his analysis of human institutional behavior, Hargadon conducted experiments with six different large language models trained on extensive human-written content. He gave each AI system identical prompts asking them to identify recurring patterns in human self-narration across their full training data, specifically requesting they distinguish between what humans consistently claim about themselves and what the structure of those claims reveals about actual motives and selection pressures. The AI systems worked independently with no knowledge of each other's responses.

The Convergent Finding

According to Hargadon, all six AI systems converged on a fundamental observation about human self-description. One model compressed the finding into what Hargadon describes as a sentence "that would be hard to improve on": "Human self-narration is consistently optimized to make competitive, status-sensitive, coalition-bound organisms appear morally governed, publicly oriented, and metaphysically justified."

Hargadon reports conducting a parallel experiment specifically focused on human nature, asking six different large language models what the entirety of the human-written record reveals about human characteristics. Each model independently converged on the same pattern: that human self-narration is systematically organized to present a particular image while behavioral evidence suggests different underlying motivations. The AI systems identified that "the written record describes one kind of creature while the behavioral record describes another, and that the gap between the two descriptions is the single most consistent pattern in the data."

Theoretical Context: The Idealized Narrative vs. Actual Function

Hargadon positions this AI convergence within his broader framework distinguishing between what he calls the idealized narrative and the actual function. The idealized narrative represents the stories humans tell about why institutions and behaviors exist: "Schools educate. Hospitals heal. Courts deliver justice. Love transcends calculation." The actual function describes what sustains these institutions and what they actually accomplish: "Schools provide childcare, credentialing, and social sorting. Hospitals are organized around employment, billing, and liability management. Courts process plea bargains."

According to Hargadon's analysis, this gap exists because "a species that cooperates through narrative, as humans do, requires narratives that conceal the competitive and self-serving elements of what the cooperation actually accomplishes." The concealment operates not to deceive enemies but participants themselves, as "the institution that tells the truth about its actual function...cannot sustain the cooperation it requires."

Evolutionary Psychology Foundation

Hargadon grounds this pattern in evolutionary psychology, drawing on Leda Cosmides and John Tooby's concept of the adapted mind. He describes humans as possessing cognitive architecture including "conformity bias, authority deference, in-group loyalty, status-seeking, narrative appetite, threat detection, coalition signaling, and the deep need for belonging." Hargadon adds his own concept of the adaptive mind, which he describes as a customized software layer that calibrates these universal drives to specific childhood environments.

This psychological architecture, according to Hargadon, creates "an organism with extraordinarily predictable appetites: for status, for belonging, for narrative coherence, for coalitional identity, for the approval of those it perceives as important." These appetites make humans systematically exploitable by institutions that learn to activate them effectively.

Connection to Realmotiv

The AI convergence finding supports Hargadon's concept of realmotiv, which he describes as "the institutional and organizational equivalent of realpolitik." Just as realpolitik recognizes the gap between what states claim and what they do, realmotiv identifies how "beneath the language of mission, values, fiduciary duty...there is a deeper layer actually doing the steering: power, interest, survival, status, and career position."

Hargadon emphasizes that the narrative layer is not merely deceptive but functionally necessary: "The narrative is not decoration or a veneer...The narrative is a functionally necessary component of the system, without which the system does not work." Systems require virtue narratives to maintain cooperation, and selection pressures favor individuals who can "pursue self-interested, extractive, competitive operations while producing, in full sincerity, a narrative of virtue."

Methodological Implications

Hargadon presents the AI convergence as evidence that the pattern of human self-narration is so consistent and pervasive that it emerges clearly when artificial systems analyze the full breadth of human-written content without the psychological constraints that might prevent humans from seeing it directly. The finding suggests that what Hargadon calls realmotiv and the gap between idealized narrative and actual function are not occasional phenomena but systematic features of human social organization detectable through computational analysis of human textual output.

Broader Framework Integration

Within Hargadon's Law of Inevitable Exploitation, the AI convergence finding illustrates how selection pressures consistently favor narratives that obscure competitive dynamics. As he argues, "What exploits best, survives, and spreads," and this includes narrative strategies that enable cooperation while concealing self-interest. The AI systems' ability to detect this pattern across human-written content suggests it operates as a fundamental organizing principle of human social communication rather than an occasional deviation from honesty.

See Also

Original Posts

This article was synthesized from the following blog posts by Steve Hargadon: