Anne Foerst

In the AI Story

Foerst studied theology at the University of Bonn and completed her doctorate on the theological anthropology of the imago Dei (image of God) in conversation with contemporary philosophy of mind. She moved to MIT in the late 1990s, initially as a skeptic of the artificial intelligence project, and remained as a collaborator whose theological training allowed her to ask questions the engineers could not formulate. Her primary research focused on two robots, Cog and Kismet, designed by Cynthia Breazeal and Rodney Brooks to exhibit social behaviors. Foerst did not dispute that the robots could produce behaviors that resembled human social interaction. She disputed the interpretation of those behaviors as evidence of selfhood, consciousness, or moral status. The behavior was real. The ontological claim was a category error.

Foerst's method was not adversarial. She participated in the laboratory's work, attended the meetings, contributed to the design of experiments, and earned the respect of the engineers as a colleague rather than an outside critic. Her theological interventions were received as questions worth taking seriously — not because the engineers became theologically convinced but because Foerst demonstrated that the theological questions illuminated confusions in the technical work. When engineers claimed a robot "wanted" something, Foerst asked whether the wanting was genuine desire or programmed response. When they claimed a robot "recognized" a person, she asked whether recognition was pattern-matching or the kind of interpersonal acknowledgment that constitutes human relationship. The questions were not hostile. They were clarifying. They forced the engineers to articulate what they actually meant, and the articulation often revealed that they meant something more modest than their initial claims suggested.

The Paul Tillich — On AI simulation honors Foerst's method by bringing Tillich into conversation with Segal's Orange Pill not to dismiss the work of builders but to deepen it — to insist that the question of capability ("What can the machine do?") must be accompanied by the question of meaning ("What does the capability serve?"), and that the second question cannot be answered by the metrics that answer the first. Foerst's life ended before the language model breakthrough, but her framework remains the most rigorous available for distinguishing between the machine's functional intelligence and the human's ontological depth. She would have recognized Segal's confession about the grinding compulsion immediately: the builder producing output in the absence of the self, testimony to a being that has departed from its own activity.

Origin

Foerst was born in Germany in 1962 and raised in a Lutheran household where theology was a living conversation rather than a settled doctrine. She entered academic theology during the 1980s, a period when feminist theology and liberation theology were challenging the discipline's traditional frameworks. Her dissertation engaged Wolfhart Pannenberg's theological anthropology and argued that the imago Dei should be understood relationally rather than substantively — humans bear the image of God not by virtue of possessing reason or free will but by virtue of the capacity for relationship with God and with one another. This relational framework became the foundation of her AI work: if the image of God is constituted by relationship, then machines that lack the capacity for genuine relationship cannot bear the image, regardless of their functional sophistication.

Key Ideas

Tillich at MIT. Brought theological ontology directly into the AI laboratory, demonstrating that metaphysical questions are practically consequential.

Behavior Is Not Being. Machines can exhibit behaviors that resemble human social interaction without possessing the ontological structure (self, ultimate concern, participation in being) that makes humans persons.

The Image of God as Relational. Humans bear the divine image not through intelligence or capability but through the capacity for relationship — with God, with each other, with the depth dimension of reality.

Prophetic Voice Died Too Soon. Foerst's 2011 death preceded the language model breakthrough by a decade, cutting short the theological engagement the 2020s crisis desperately needs.