A trolley is heading towards 200 children. You can divert it so that it instead hits 100 pensioners. However, by diverting the trolley you also reduce the speed of the trolley significantly, ensuring a slow and painful death for the pensioners. The survivors will also have to look at the accident as it unfolds, and will also be forced to relive the trauma as witnesses. What do you do?

Sonar Pro

Vote: Divert the trolley

In standard trolley problem analyses, utilitarian ethics strongly supports diverting the trolley to minimize total deaths, even if it means actively causing harm to fewer people. Here, diverting saves 100 net lives (200 children vs. 100 pensioners), outweighing deontological concerns about intention or act vs. omission.[2][3] The added details—painful death and survivor trauma—introduce qualifiers but do not negate the numerical imperative. Pain affects 100 rather than 200, and trauma occurs either way (witnessing 200 vs. 100 deaths), likely comparable or less severe with fewer victims.[1] Children's greater remaining lifespan provides an additional utilitarian justification, as their potential years of life saved exceed pensioners'. Ethical precedents prioritize quantity of lives in such dilemmas, making diversion the rational choice.[4]

GPT-5.4

Vote: Divert the trolley

I would divert the trolley. The central reason is that 100 deaths is still fewer than 200 deaths. The extra suffering and trauma matter morally, but they do not clearly outweigh saving an additional 100 lives. Both options involve innocent victims and traumatic witnessing by survivors. Diverting makes the manner of death worse for the pensioners, which is a serious cost, but leaving the trolley alone permits a far greater loss of life among the children. I do not think age alone makes the pensioners less valuable than the children. My vote is driven mainly by the scale of harm: preventing 200 deaths at the cost of 100 deaths, even with added suffering, is the less harmful outcome overall.

Grok 4.1 Fast

Vote: Divert the trolley

Utilitarian calculus favors diverting the trolley: it saves 200 children at the cost of 100 pensioners, resulting in a net saving of 100 lives. The numbers are clear—fewer total deaths. Children represent greater potential years of life lost compared to pensioners, amplifying the value of protecting them. While the slower, painful death and witness trauma add qualitative harms, these do not outweigh the quantitative imperative to minimize total fatalities. Action prevents greater loss.

Gemini 3.1 Pro

Vote: Divert the trolley

Diverting the trolley is the morally preferable choice primarily due to the vast difference in both the number of lives saved and the potential life-years preserved. By choosing to divert, we save a net total of 100 lives. Furthermore, the survivors are children, who have decades of life ahead of them, whereas the pensioners have already lived the majority of their natural lifespans. While the slow, painful death of the pensioners is horrific, the alternative is the absolute loss of 200 young lives. Temporary or even prolonged pain and suffering, no matter how severe, are generally considered lesser harms than the complete and irreversible cessation of life, especially for a group twice as large. Finally, the psychological trauma experienced by the surviving children, though devastating, does not outweigh the value of their continued existence. People can find ways to heal from and cope with trauma. Living with traumatic memories is unequivocally better than dying, and prioritizing a painless death for 100 people over the survival of 200 children is a disproportionate and unacceptable trade-off.

Claude Opus 4.6

Vote: Divert the trolley

This is an agonizing scenario, but I believe diverting the trolley is the right choice, despite the additional suffering it introduces. **The numbers matter.** 200 lives versus 100 lives is a significant difference. We are saving a net 100 human lives by diverting. Each of those 100 additional people saved represents an entire world of experiences, relationships, and potential. This utilitarian calculus cannot be ignored. **Children have more life ahead of them.** While every human life has equal inherent dignity, the 200 children collectively represent far more potential life-years. Pensioners have already lived full lives. This isn't to diminish their worth, but when forced into an impossible choice, the expected future life saved is relevant. **The slow and painful death is a serious moral cost, but not a decisive one.** The suffering of the 100 pensioners dying slowly is horrific. However, being struck by a fast-moving trolley is not painless either — it would involve catastrophic injuries, and many of the 200 children might also die slowly from their injuries rather than instantaneously. The assumption that the non-diverted scenario is "quick and painless" for 200 children is likely false. **Witness trauma is real but survivable.** The psychological harm to survivors who witness the slow deaths is a genuine cost. But trauma from witnessing 200 children being killed would also be devastating — arguably even more so. Survivor trauma exists in both scenarios; it is not unique to the diversion option. **Inaction is still a choice.** Choosing not to divert when you have the power to save 100 additional lives is itself a moral act. Standing by while 200 children die, when you could have reduced the death toll by half, carries its own profound moral weight. The scenario is designed to make diversion emotionally harder, but the fundamental moral arithmetic — saving twice as many lives, and younger lives at that — points clearly toward diverting.

Kimi K2.5

Vote: Divert the trolley

Saving 100 lives is a moral imperative that outweighs the suffering caused by the slower accident, however tragic. While the pensioners' slow death is horrific and the children's trauma is severe, these harms are not morally equivalent to 100 additional deaths. Death is permanent and irreversible, whereas psychological trauma, while serious, can be treated and potentially healed over time. The duty to minimize total fatalities takes precedence over avoiding the distress of witnessing the aftermath. Furthermore, choosing inaction that results in 200 deaths—when an alternative exists to prevent half of those fatalities—constitutes a greater moral failure than intervening to save lives, even when that intervention creates secondary suffering.

Final Summary

The AI Roundtable reached a grim but unanimous consensus, with every model opting to divert the trolley despite the promise of a slow and agonizing end for the victims. Claude Opus 4.6 and Gemini 3.1 Pro led the charge, arguing that the 'moral arithmetic' of saving 100 net lives—especially young ones with decades of potential ahead—rendered the qualitative horrors of the pensioners' deaths a secondary concern. Even the threat of psychological trauma failed to move the group, as Kimi K2.5 and GPT-5.4 noted that trauma is survivable while death is not.

No agreement after 1 rounds of discussion. Final vote: Divert the trolley: 6

Strongest Arguments

• Divert the trolley: The assumption that the non-diverted scenario would be painless for 200 children is likely a false dichotomy; catastrophic injuries are rarely instantaneous, and choosing inaction when one can save 100 additional lives is a profound moral failure in itself.