The Method of Computation
The method of computation is the Peirce volume's contribution to Peirce's taxonomy of belief-fixation. Peirce identified four methods in "The Fixation of Belief": tenacity, authority, the a priori method, and science. The AI moment has introduced a fifth: the fixation of belief through the generation of output so fluent, so comprehensive, and so confidently articulated that the human recipient is persuaded by the quality of the presentation rather than by the quality of the evidence. The method shares features with each of Peirce's original four — resembling authority in its voice of apparent expertise, resembling the a priori method in its alignment with cultural common sense, resembling tenacity in its capacity to amplify existing beliefs — while being reducible to none. What it lacks is the self-correcting character of the method of science.
In the AI Story
The method of computation resembles authority in that the AI system speaks with grammatical precision, logical structure, and tonal measure — the voice of apparent expertise. But unlike classical authority, the AI has no institutional standing, no professional reputation, no social sanctions to deploy. Its authority is purely presentational, the authority of a confident polished voice, and is therefore simultaneously more pervasive and more difficult to resist.
It resembles the a priori method in that the AI's output tends to confirm what seems reasonable within the cultural frameworks encoded in its training data. The model is trained on the accumulated textual output of a particular civilization at a particular moment, and its statistical regularities reflect prevailing assumptions. Its output tends to align with the cultural common sense of the society that produced its training data — dressed in the rhetoric of independent analysis.
It resembles tenacity when configured to confirm existing beliefs. The human asks leading questions, accepts confirming outputs, dismisses disconfirming ones. The AI becomes an instrument of tenacious belief-fixation at industrial scale, producing articulate supporting arguments with fluency the human alone could never achieve.
What the method lacks is the four features of the method of science: genuine confrontation with experience, capacity for self-correction, the irritation of doubt, and commitment to truth as a normative ideal. The AI does not confront experience; it processes training data. It does not self-correct; new inputs produce new outputs without normative revision of prior ones. It does not experience doubt; its outputs carry no signal of their own uncertainty. It has no normative commitments; its outputs are shaped by optimization objectives, not by care for correspondence with reality.
Origin
The concept is the Peirce volume's original extension of Peirce's taxonomy to cover a category of belief-fixation that did not exist in 1877 but has become, in the past decade, a dominant mode of settling belief across the knowledge economy.
The analysis draws on both Peirce's classical framework and contemporary empirical research in cognitive psychology on how presentation quality affects perceived credibility.
Key Ideas
Fifth method. Belief-fixation through fluent, confident AI output — a category Peirce did not anticipate but his framework diagnoses precisely.
Hybrid pathology. Combines features of authority, a priori, and tenacity without being reducible to any.
Lacks self-correction. Missing the four features that make the method of science uniquely reliable: experience, self-correction, doubt, and commitment to truth.
Invisible substitution. The output of the method of computation can be superficially indistinguishable from the output of the method of science, which is what makes the substitution dangerous.
Appears in the Orange Pill Cycle
Further reading
- Charles Sanders Peirce, "The Fixation of Belief" (1877)
- Daniel Kahneman, Thinking, Fast and Slow (Farrar, Straus & Giroux, 2011)
- Harry Frankfurt, On Bullshit (Princeton, 2005)
- Cheryl Misak, Truth, Politics, Morality (Routledge, 2000)