Intersubjective Literacy

In the AI Story

The need arises from AI's entry into intersubjective reality as a non-participant producer of intersubjective content. Traditional media literacy—'check the source, evaluate the evidence, identify logical fallacies'—is necessary but insufficient. An AI-generated legal brief can cite real precedents accurately, construct logically valid arguments, present evidence without distortion—and still be intersubjectively hollow because the system does not understand the law it cites. It processes legal language without grasping the institutional realities, human consequences, or moral stakes that give the law its weight. Detecting this hollowness requires domain expertise (knowing enough about law to recognize when an argument is formally correct but substantively empty) plus hermeneutic discipline—the habit of asking 'does this text carry genuine understanding?'

Segal's Deleuze error in The Orange Pill is the micro-scale teaching case. Claude generates a passage connecting flow theory to Deleuze's smooth space—elegant, well-structured, philosophically incoherent. Segal catches it because he possesses background knowledge (enough Deleuze to recognize misuse) and discipline (the willingness to check even plausible-sounding passages). How many equivalent errors, across how many domains, go unchecked? The question is not rhetorical. It is the educational crisis of the AI age: producing a population capable of evaluating AI outputs with the rigor that genuine participation requires.

Intersubjective literacy is not a cognitive add-on but a literacy transition comparable to the shift from oral to written culture. That transition took centuries; institutions adapted slowly (schools teaching reading, libraries organizing texts, legal frameworks recognizing written contracts). The AI transition is offering years. Educational systems must teach intersubjective literacy at a speed no previous literacy transition achieved—while themselves being flooded by AI-generated educational materials whose intersubjective status is ambiguous.

Origin

The concept emerges from Harari's Nexus analysis of AI's threat to trust, synthesized in this volume with pedagogical frameworks (Bruner's narrative vs. paradigmatic thought, Wolf's deep reading) and epistemological traditions (Newman's illative sense, Polanyi's personal knowledge). The 'literacy' framing emphasizes teachability—this is not innate discernment but cultivated skill, improvable through practice.

Key Ideas

Beyond source-checking to ontological evaluation. Not 'who wrote this?' but 'was this written by someone who understands and cares what it means?'

Requires domain expertise plus hermeneutic discipline. Catching plausible hollowness demands background knowledge (to recognize when something is formally correct but substantively empty) and the habit of questioning even smooth surfaces.

Teachable but not yet taught. Educational institutions have not integrated intersubjective literacy into curricula. The skill is being learned improvisationally by early adopters—or not learned at all.

Collective epistemic defense. No individual can verify everything. Intersubjective literacy is sustainable only if it becomes a shared practice—a new cultural norm of disciplined skepticism toward plausible outputs.

The alternative is epistemic collapse. Without widespread intersubjective literacy, the epistemic commons—the shared pool of reliable information—degrades below the threshold where self-correction functions.