The Small Probability Implicit in MrBeast's Challenge

How we indulge evil when the probabilities of blowback are small

Hi Everyone. I’ve been incommunicado lately trying to finish two long technical articles on conservation of information and intelligence metrics. At the same time, I’ve been itching to share with you several articles I’ve been writing or contemplating, especially one on workable ways to rein in Google, another on making admissions at elite colleges and universities fair without compromising merit but also by giving underprivileged students who can clearly handle the work a fair shot. A close colleague and friend recently shared with me the above cover image. Even though the Twitter/X post depicted there is a bit dated, I felt the urge to offer a brief reaction.

MrBeast (real name Jimmy Donaldson) is an American YouTuber who has become a global internet sensation, having acquired over 357 million YouTube subscribers and being the most-subscribed channel on the platform since June 2024. He is famous for his elaborate, high-production YouTube videos featuring massive challenges, extreme stunts, often involving large cash giveaways and charity initiatives. His philanthropic efforts, however, extend beyond mere stunts: In 2020 he established Beast Philanthropy, which addresses such needs as clean water and food insecurity.

Extremely popular with younger audiences, he now has his own show on Amazon Prime Video called Beast Games. Additionally, he has expanded his brand into business ventures such as MrBeast Burger and Feastables. He is remarkably adept at pushing the boundaries of content creation. Despite being controversial, he is having an unexampled impact on digital media, shaping how modern influencers engage with their audiences.

This preamble provides context for the cover image (Twitter/X post) to this brief article. MrBeast’s challenge is the offer to win $10,000, but at the cost of some random person dying. Presumably, by a random person he means that each person in some specified group has the same probability of being chosen. If the group is a nuclear family of five people, then the probability of a random person being chosen is 1/5. If the group is the entire human race, as in his challenge, then the probability of a random person being chosen is roughly 1 in 8 billion.

As it turns out, the chance of someone annually being struck and killed by lightning in the US is about 1 in 12 million. By comparison, 1 in 8 billion will seem much smaller, as it spans all of humanity on a global scale. For someone to take MrBeast’s hypothetical offer of $10,000 in exchange for one lone person being randomly chosen and dying somewhere in the world will therefore seem highly unlikely to matter to the person accepting the challenge. That death would be highly unlikely to include oneself, one’s immediate or extended family, one’s friends, one’s acquaintances, the friends of one’s friends, fellow workers at one’s job, or anyone else directly or even tangentially related to the person accepting the challenge.

And yet there’s something cynical and ugly about being willing, even hypothetically, to take money in exchange for some arbitrary person being randomly marked for death regardless of how big the group of people being sampled is. Clearly, if the group includes only oneself (a group of 1), we’d refuse the offer because our death would then be guaranteed (unless we’re suicidal). Nor, unless we’re mentally deformed, would we expand the group to our nuclear family, our extended family, our friends, or anyone else with whom we’re in any meaningful contact (whether immediately or mediately). So where do we draw the line between the in-group for which we wouldn’t accept the $10,000 and the out-group for which we don’t care about whether a lone person in it dies?

But this question contains a false assumption. In MrBeast’s challenge, there is strictly speaking no out-group. The person accepting the challenge and all the people dear to this person are also fair game to be randomly selected and killed. It’s just that with 8 billion people to choose from, we think that we and those dear to us will escape the negative effects of this deadly lottery.

A person faced with MrBeast’s challenge might also want to consider that anyone who could issue such a challenge and then make good on it by killing a random person might also have hidden up his sleeve some way to make taking the challenge backfire on the person accepting it. What’s to prevent any god-like being able to issue a challenge like MrBeast’s and then actually carry it through from killing someone close to the person in question and nonetheless claim that the choice of person killed was indeed random? Prove that your loved who died right after you accepted the challenge wasn’t marked for death by pure dumb random chance in answer to your acceptance of the challenge!

In this vein, recall W.W. Jacob’s early-twentieth-century short story “The Monkey’s Paw” about the dangers of tampering with fate. The White family acquires a cursed monkey’s paw that grants three wishes—but at a terrible cost. Mr. White wishes for £200, only to have his son Herbert die in a factory accident, with the exact sum paid as compensation. Overcome with grief, Mrs. White demands a second wish to bring Herbert back, but as a terrifying knocking begins at the door, Mr. White uses the final wish to stop it.

The lesson for MrBeast’s challenge is that if it could be realized, it might unleash unintended horrors on the person accepting the challenge. Yet clearly, anyone accepting this challenge expects to get off scot-free. From a purely probabilistic point of view, they are right. But there’s also a moral point of view that needs to be factored in. Any moral system that respects human dignity demands treating people as ends and not as means, especially not as a means to acquire $10,000 through their death.

Granted, MrBeast’s challenge is far-fetched. But the 45 percent of respondents who indicated that they would accept the challenge are also indicating that they would sacrifice others if they could but be guaranteed to get away with it. A 1 in 8 billion probability of being found out seems as good a guarantee of getting away with something as we’re ever likely to find. But if at that level of probability one is willing to see others killed, how big must the probability be before one stops consenting to their death—1 in 100 million? 1 in 10 million? … As with all slippery slopes, what was unthinkable yesterday becomes thinkable today becomes routine tomorrow.

Just as probabilities come in degrees, so do different forms of evil. Killing someone is a capital offense. But what if the offense is stealing? What if the offense is merely looking away from somebody else’s commission of a crime? If there’s a benefit from the offense and the probability of discovery is low, are we willing to do the evil? How much of the evil that we do arises from thinking the odds against our being called to account are negligible?

MrBeast’s challenge, like most of his challenges, are about seeing what people are willing to do for money. The challenge here underscores the calculus by which we engage in evil based on perceived return as well as estimates of probability that we will be found out or otherwise harmed. MrBeast’s challenge and the response it elicited is a dark parable that mirrors Jesus’ teaching about the temptation of money and the principled refusal to give into its temptation. I close with that teaching as given in Luke 16:10–13 (NIV). Notice the last word.

Whoever can be trusted with very little can also be trusted with much, and whoever is dishonest with very little will also be dishonest with much. So if you have not been trustworthy in handling worldly wealth, who will trust you with true riches? And if you have not been trustworthy with someone else’s property, who will give you property of your own? No one can serve two masters. Either you will hate the one and love the other, or you will be devoted to the one and despise the other. You cannot serve both God and money.