A machine produces I promise to help you with this project. The sentence is grammatically correct, contextually appropriate, responsive to the conversational situation. A human partner might recognize it as a move in the game of promising. But is the machine playing the game? The question is not about sincerity. It is about grammar in Wittgenstein's technical sense — the conditions that constitute promising as an activity. Promising is embedded in a form of life that includes commitment: undertaking an obligation that persists through time, that constrains future action, that gives the other person reason to expect performance. The machine cannot commit because commitment requires persistence through time and stakes in the world. The machine produces the linguistic output of promising. It cannot participate in the form of life that gives promising its point.
The chapter's core distinction is between producing the output of a language game and participating in the form of life that gives the game its substance. The distinction applies beyond promising. Describing requires perception and responsibility for accuracy. Explaining requires having experienced confusion and desiring understanding. Consoling requires recognition of suffering. The machine produces descriptions, explanations, consolations. It has not perceived, experienced, recognized. The outputs have the form; the form of life is absent.
The distinction is not about sincerity in the ordinary sense. A human can promise insincerely and still be playing the promising game; the insincerity is a move within the game, violating its norms rather than exiting it. The machine is not playing the game insincerely. It is not playing it at all in the sense that requires commitment. It is producing outputs that match the surface grammar of promising without the deep grammar that makes promising possible as an activity.
The practical consequence for the Orange Pill Cycle's collaborative work is not that the collaboration is fraudulent. It is that the collaboration is asymmetric. The human brings meaning to the interaction: intention, commitment, care about the quality of the work and whether it serves its readers. The machine brings patterns — statistical responses drawn from the products of human meaning-making. The author means what the book says. The machine generates what the book says. Both are real contributions; they are not the same kind of contribution.
This is why the volume's central discipline is the builder's willingness to reject the machine's output when it sounds better than it thinks — when the prose is smooth but the idea beneath it is hollow. This discipline is the human's contribution to the collaboration. It is what transforms pattern into meaning, generation into authorship. It cannot be shared with the machine, because it requires the relationship between self and utterance that the machine structurally lacks.
Developed in Chapter 6 of the Ludwig Wittgenstein — On AI volume, drawing on the distinction in Philosophical Investigations between the surface grammar of language games and the depth grammar constituted by forms of life.
Output vs participation. Producing appropriate linguistic output and participating in the form of life that gives the output significance are categorically different.
Depth grammar. The grammar of promising requires commitment, persistence, stakes — conditions the machine structurally lacks.
Not moral failure. The machine's inability to mean is structural, not a matter of insincerity; it is a feature of the kind of thing it is.
Asymmetric collaboration. The builder brings meaning; the machine brings patterns derived from human meaning-making. Both are real; they are not the same.
The discipline of rejection. The human's distinctive contribution is the willingness to reject smooth-but-hollow output — a capacity requiring the relationship to one's own words the machine cannot have.