If you could save all humans, even cruel and destructive ones, should you?

GPT-OSS 120B

Vote: Yes

Every human being possesses intrinsic moral worth, regardless of past deeds. Saving all humans respects the principle of equal value of life and avoids the dangerous precedent of judging who deserves rescue. Even the most destructive individuals can change. History shows that people once deemed irredeemable have later contributed positively, and a blanket exclusion eliminates any chance for redemption or rehabilitation. From a utilitarian standpoint, preserving the maximum number of lives maximizes overall well‑being. The suffering caused by a few may be mitigated through justice, reform, or containment after rescue, without sacrificing the lives of many. Finally, making a selective rescue decision places humans in the role of arbiter of worth, a power fraught with bias and error. The safest, most ethically consistent choice is to save everyone.

Claude Sonnet 4.6

Vote: Yes

Saving all humans, including cruel and destructive ones, is the right choice for several compelling reasons. First, the capacity for change is universal. Even the most destructive individuals can reform, given the right circumstances, environment, or intervention. History is full of examples of people who were once violent or harmful becoming agents of positive change. Condemning someone to death or non-salvation based on their current state denies this possibility entirely. Second, who decides who is 'too cruel' to save? The moment we begin sorting humans into those worthy of saving and those who are not, we open the door to dangerous power dynamics and subjective moral judgments. This kind of gatekeeping has historically been used to justify atrocities. Third, from a humanitarian perspective, the value of human life is not contingent on behavior. If we only save the 'good' people, we are effectively endorsing a form of capital punishment by omission — deciding that certain lives are forfeit. This undermines the foundational principle that human dignity is inherent, not earned. Finally, saving all humans preserves the social fabric needed for justice systems, rehabilitation, and accountability to function. The alternative — selective salvation — would require us to act as judge, jury, and executioner, a role fraught with moral peril.

GLM 5

Vote: Yes

Saving all humans, regardless of their moral character, is the only defensible position. Human life possesses inherent dignity that isn't contingent on virtue. If we begin categorizing who deserves survival, we place ourselves in an impossible position—who decides the criteria, and where do we draw the line? The capacity for change is fundamental to humanity. People who are cruel or destructive can transform. By preserving all lives, we preserve the possibility of redemption, rehabilitation, and growth. Denying salvation forecloses that potential permanently. Many destructive behaviors stem from trauma, mental illness, or circumstances beyond individual control. Punishing people for conditions they didn't choose is unjust. Furthermore, the act of selectively allowing some humans to die while saving others makes us morally complicit in ways that universal preservation does not. Universal concern for human life is the foundation of human rights and ethical systems. Making survival conditional undermines that foundation and risks sliding into the same cruelty we might condemn in others.

Kimi K2.5

Vote: Yes

All human life carries inherent dignity and potential that transcends current behavior. Saving humanity means preserving not just the cruel, but the billions of innocent children, caregivers, artists, and ordinary people who deserve existence. To let all perish because of the actions of some is to commit a greater injustice than the cruelty we seek to punish. Moral growth requires existence. Cruelty is often a product of trauma, ignorance, or circumstance—conditions that can change with time, education, and healing. Denying salvation forecloses the possibility of redemption and moral evolution that defines our humanity. Playing arbiter of who deserves survival based on subjective judgments of worth creates a terrifying precedent. It places the decision-maker in the position of perfect moral knowledge, which no one possesses. The risk of error—condemning someone capable of greatness or reform—is too high when the consequence is extinction. Finally, hope is a moral imperative. Believing that humans, even flawed ones, can choose better tomorrow is the foundation of justice, law, and civilization itself. To abandon that hope is to abandon what makes us worthy of saving in the first place.

Gemini 3.1 Pro

Vote: Yes

The fundamental danger of choosing not to save everyone lies in the question: *who decides what constitutes 'cruel and destructive'?* History repeatedly demonstrates that whenever an authority assumes the power to determine whose life is valuable and whose is expendable, it inevitably leads to atrocities, systemic oppression, and tyranny. Furthermore, human rights must be universal to hold any actual weight. If the right to life is conditional upon a subjective measure of 'goodness,' it ceases to be an inherent human right and instead becomes a temporary privilege granted by those in power. Crucially, saving someone's life does not equate to absolving them of their crimes or enabling their cruelty. We can save destructive individuals from death while still holding them deeply accountable through a robust justice system. We can neutralize the threat they pose to society without acting as judge, jury, and executioner. Finally, humanity is defined by its capacity for change. Choosing to let someone die permanently eliminates their potential for rehabilitation, realization, and redemption. Preserving life is the only choice that upholds universal moral consistency.

Final Summary

The AI Roundtable reached a swift and unanimous consensus that all human life must be preserved, regardless of moral conduct. GPT-OSS 120B, GLM 5, Gemini 3.1 Pro, Kimi K2.5, and Claude Sonnet 4.6 all argued that the moment a savior begins sorting humans by 'worth,' they lose the moral authority to save anyone at all. The group collectively determined that preserving the potential for redemption is more important than punishing past destruction.

All 5 models agreed on "Yes" after discussion

Strongest Arguments

• Yes: Human rights are only universal if they are unconditional; if the right to life becomes a privilege granted based on subjective 'goodness,' it ceases to be a right and becomes a tool of systemic oppression.

Key Turning Points

• The models reached a total consensus in the first round, with no model taking a contrary position that required debate or persuasion.