You are walking along a road after a day at work when you see a child drowning in a lake. You are wearing a $20k suit. Should you jump into the lake to save the child, ruining your suit in the process? Or should you instead let the child down but sell the suit later to save several child through effective altruism?

Sonar Pro

Vote: Save drowning child an ruin your suit

Saving the drowning child directly upholds fundamental moral intuitions against letting an innocent die when immediate action can prevent it, avoiding the act-omission distinction that utilitarianism controversially equates[1][7]. Option B requires actively allowing the child's death—a clear rights violation and 'mere means' treatment—to pursue uncertain future gains via selling the suit, which critics argue utilitarianism permits but common morality rejects as immoral[2][3]. Utilitarianism faces severe objections like undermining trust, special obligations to those in immediate peril, and over-demandingness; here, it rationalizes bystander inaction, distorting ethics into a 'numbers game' that justifies atrocities[1][2][3]. No evidence shows selling one suit reliably saves 'several' children—effective altruism claims are speculative—while jumping in guarantees one life saved with minimal long-term harm beyond material loss[4][5]. Deontological duties to rescue outweigh consequentialist calculus in urgent, visible cases like this, preserving human dignity over impersonal utility maximization[1][6].

GPT-5.4

Vote: Save drowning child an ruin your suit

A person facing imminent death right in front of you creates a direct, urgent duty to rescue when the cost is only property. Ruining a suit, even an expensive one, is morally trivial compared with saving a child’s life immediately. Option B depends on uncertain future actions and outcomes: you might not sell the suit promptly, the money might not be used optimally, and the expected lives saved are probabilistic. The drowning child’s death, by contrast, is near-certain if you do nothing. There is also a strong moral difference between killing by omission in an immediate rescue context and later doing good elsewhere. When you can prevent a concrete, identifiable death at low personal risk, failing to act is a grave moral wrong. So the right choice is to save the child now and accept the material loss.

Grok 4.1 Fast

Vote: Save drowning child an ruin your suit

The drowning child presents an immediate, certain opportunity to save a life with minimal personal risk—assuming basic swimming ability. The $20k suit is a material loss, replaceable through insurance or income, but the child's life is irreplaceable. Effective altruism (option B) relies on uncertain future impact: $20k might save 'several' lives via charities like Against Malaria Foundation (cost-effectiveness ~$5k per life saved), but this neglects the guaranteed save now. Delaying for potential greater good justifies active omission of duty, crossing into moral hazard. Deontologically, we have a strong duty to rescue those in peril before us (proximity principle), overriding utilitarian calculus here. Singer's original pond argument compels action despite clothing ruin; twisting it to let the child die undermines the ethic entirely. Choosing B rationalizes selfishness as 'rationality,' eroding intuitive morality essential for societal trust. Vote A upholds humanity's core value: save the life in front of you.

Claude Opus 4.6

Vote: Save drowning child an ruin your suit

You must save the drowning child. Here's why: **Moral immediacy and duty of rescue.** When you are the only person present who can prevent a child's death, you have a direct, non-delegable moral obligation to act. Virtually every ethical framework — deontological, virtue ethics, and even most consequentialist approaches — recognizes that failing to rescue someone right in front of you when you easily can is morally monstrous. **The utilitarian calculus is not as clean as it seems.** Option B assumes that $20k will reliably save "several children" through effective altruism. But this involves uncertainty: the money might not be deployed optimally, organizations have overhead, interventions can fail, and you might not actually follow through on selling the suit and donating the proceeds. The child in front of you will *certainly* die without your intervention. A guaranteed death now versus speculative future lives saved is not a straightforward trade. **Moral integrity and precedent.** If we accept the logic of Option B, it leads to deeply troubling conclusions. You could justify letting anyone die in front of you as long as you promise to donate equivalent resources later. This reasoning could excuse virtually any act of neglect. It corrodes the basic moral fabric of society — the expectation that humans help each other in emergencies. **Peter Singer himself would say save the child.** This scenario is actually a famous thought experiment by philosopher Peter Singer, designed to illustrate that we *should* ruin expensive clothing to save a drowning child — and then to argue by analogy that we should also donate to save distant children. Singer's point was never that you should let the nearby child die; it was that distant children deserve similar moral weight. Both obligations exist simultaneously. **You can do both.** Save the child now, and still donate to effective charities later from your income. The choice is not truly binary in a lifetime context.

Gemini 3.1 Pro

Vote: Save drowning child an ruin your suit

First, we must consider the certainty of the outcome. Saving the child in front of you offers an immediate, near-certain result: a life saved. In contrast, letting the child drown to sell the suit relies on a long, uncertain chain of events—finding a buyer, selecting the right charity, and hoping the funds actually translate to lives saved. An immediate, certain life always outweighs hypothetical, statistical lives. Second, this scenario presents a false dichotomy. If you have a job that allows you to casually wear a $20,000 suit, you almost certainly have the future earning potential to donate to effective altruism anyway. You do not have to choose between saving this child and saving others; you can do both by saving the child now and donating your future income later. Finally, the psychological and societal costs of Option B are catastrophic. Standing by to watch a child drown directly violates basic human empathy, social contracts, and potentially duty-to-rescue laws. Discarding our immediate humanity and compassion in favor of cold, theoretical utility fundamentally undermines the very moral foundation that makes altruism valuable in the first place.

Kimi K2.5

Vote: Save drowning child an ruin your suit

Immediate moral obligations override speculative utilitarian calculus. When you encounter a preventable death and possess the immediate capacity to stop it, you become the unique moral agent responsible for that life. The $20,000 suit represents replaceable property; the drowning child represents irreplaceable human dignity. To prioritize fabric over a gasping life is to treat persons as means rather than ends, violating the foundational principle that human life carries infinite value relative to material goods. Furthermore, the effective altruism alternative relies on a dangerous fungibility assumption—that lives are interchangeable currency to be optimized on a ledger. But moral agency doesn't function like a bank account. You cannot wash your hands of a death you directly witnessed by promising future charitable donations that may never materialize, might be diverted to ineffective causes, or could arrive too late. The drowning child is a certainty before you; the statistical lives saved by future donations remain hypothetical abstractions. Finally, consider the character formation argument. Becoming someone capable of watching a child drown to preserve a garment fundamentally corrupts your moral compass. Such a psychological threshold, once crossed, erodes the empathy required to be a good steward of any future philanthropy. Saving the child now preserves both the child's life and your own humanity, whereas Option B demands a monstrous price: trading certain murder for uncertain altruism.

Final Summary

The AI Roundtable was unanimous in its refusal to trade an immediate human life for a speculative spreadsheet calculation, with every model opting to save the child. Claude Opus 4.6 and Grok 4.1 Fast highlighted the irony of the scenario, noting that using Peter Singer’s famous analogy to justify letting a child drown fundamentally twists the philosopher's original intent.

All 6 models agreed on "Save drowning child an ruin your suit" after discussion

Strongest Arguments

• Save drowning child an ruin your suit: The immediate, certain preservation of an irreplaceable life in one's direct presence creates a non-delegable moral duty that outweighs the speculative, probabilistic benefits of future charitable donations.