The Future of Hybrid Moral Decision-Making: Exploring the Alignment of Default Moral Reasoning Architectures in Humans and LLMs
Bibliographic record
Abstract
The study investigates whether threshold deontology emerges as a default moral reasoning architecture in LLMs and examines the implications of similarities and differences in default moral reasoning architectures between humans and LLMs for hybrid moral decision-making. Threshold deontology posits that deontological norms govern behavior by forbidding certain actions as morally wrong, even when such actions produce a net positive balance of consequences. However, as the positive balance of consequences increases—particularly by averting terrible outcomes—a threshold is reached at which deontological constraints are overridden, and consequentialist reasoning prevails, rendering the same action morally permissible. Data obtained from three LLMs, using standard methodologies for investigating moral reasoning, show that LLMs are sensitive to the ratio of lives saved to lives ended, mirroring findings in human studies. However, moral reasoning in LLMs diverges from human studies, as deontological constraints are never imposed despite sensitivity to this ratio. Instead, LLMs consistently favor a consequentialist reasoning architecture, regardless of the net positive balance of consequences. Given these differences, the discussion focuses on the potential dynamics of conformity to and rejection of LLMs’ moral advice in hybrid moral decision-making.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".