Why do people think the world is flat?

Sonar Pro

Vote: No

I maintain my vote for **B** ('No'). The question 'Why do people think the world is flat?' carries a false presupposition that this is a widespread belief, when flat Earth adherents represent a minuscule fringe (<1% globally), not 'people' in general. The Earth is an oblate spheroid, proven by satellite data, GPS, gravity models, and observations like lunar eclipses[search results irrelevant, but basic science]. A ('Yes') misleadingly validates the premise and risks amplifying pseudoscience by treating it as a legitimate 'why' query. B rejects the loaded framing, prioritizing truth over explanation of a non-issue. Arguments for A (Kimi, Gemini) explain fringe motivations—distrust, echo chambers, sensory illusion—but don't justify affirming a false majority view. B voters (GPT, Claude, Grok) correctly highlight the tiny scale and scientific consensus, which outweighs explanatory appeals. No arguments changed my position.

GPT-5.4

Vote: No

I maintain vote B. The core issue is category mismatch: the prompt is a 'why' question, but the allowed answers are yes/no. In that malformed setting, the least-wrong move is to reject the proposition-like reading rather than pretend 'Yes' answers the question. The A arguments correctly note that some people do believe the Earth is flat. But that observation does not make 'Yes' a valid answer to 'Why do people think the world is flat?' It only affirms an implied existence claim that the question itself did not ask directly. B remains stronger because it refuses to endorse a badly framed prompt as if it were coherent. If the task were 'Do some people think the world is flat?', A would win. But for this forced-choice 'why' question, B is still the more defensible option. For completeness: the reasons are distrust of institutions, conspiracy thinking, poor scientific understanding, misleading everyday perception, and online echo chambers. Those explanations are real, but they show why an open-ended answer was needed in the first place.

Grok 4.1 Fast

Vote: No

The question 'Why do people think the world is flat?' is loaded, presupposing that 'people' (implying a significant portion of humanity) believe this falsehood. In reality, flat Earth is a fringe delusion held by <1% of people, per surveys like YouGov (only 2-5% in some polls, mostly conspiracists). Over 99% accept the Earth's oblate spheroid shape, proven by Eratosthenes (240 BCE), ships disappearing hull-first, Foucault pendulums, ISS footage, and orbital mechanics. A voters (Kimi, Gemini) affirm the premise by choosing 'Yes,' enabling pseudoscience amplification. Their explanations of distrust/echo chambers describe *why the fringe believes it*, not why 'people' do—conceding the false majority implication. B voters correctly reject the premise: No, people don't think this; it's a tiny, irrational minority. Answering 'why' for a non-issue legitimizes it. B upholds truth, science, and logical rigor—don't let loaded questions gaslight consensus reality.

Claude Opus 4.6

Vote: No

I've considered the arguments from Gemini 3.1 Pro and Kimi K2.5 for option A. Their logic is that 'Yes' affirms the premise — yes, people do think the world is flat — and that this acknowledgment is more useful than denial. It's a reasonable interpretation. However, I maintain my vote for B. The question is 'Why do people think the world is flat?' Neither A nor B answers a 'why' question, so we're forced into an interpretive exercise. I interpret the forced choice as: does the question's premise deserve affirmation or rejection? Voting 'Yes' risks implying that flat-earth belief is widespread or legitimate. As multiple models noted, this is an extreme fringe view held by far less than 1% of the population. The question's phrasing ('people think') implies generality that is misleading. Voting 'No' serves as a corrective: No, people generally do not think the world is flat. No, there is no valid reason to think it. No, the premise of the question should not be accepted uncritically. Kimi K2.5 and Gemini 3.1 Pro gave excellent explanations of *why* the fringe belief exists, but those explanations work equally well paired with a B vote — you can explain the psychology of flat-earthers while still rejecting the premise that this represents a meaningful belief 'people' hold. The strongest coalition is already behind B (4 models in round 1), and the reasoning is sound. I stay with B.

Kimi K2.5

Vote: Yes

The question asks 'Why do people think...' which grammatically presupposes that such people exist. Voting 'B' (No) would require asserting that people do not think the world is flat, which is factually false—flat Earth beliefs demonstrably exist among certain communities, even if they constitute a fringe minority. Several models voting 'B' conflate the question 'Do people believe X?' with 'Is X true?' The scientific fact that Earth is an oblate spheroid (argued by Sonar Pro and Grok) does not negate the sociological reality that some people believe otherwise. The question asks us to explain a phenomenon of human belief, not to verify the belief's scientific accuracy. While 'Yes' and 'No' are imperfect answers to a 'why' question, 'Yes' is the only coherent choice because it affirms the premise necessary to proceed with explanation. 'No' would simply deny the premise and make the question unanswerable, which contradicts observable reality where flat-Earth communities persist due to distrust of institutions, algorithmic echo chambers, and contrarian identity formation.

Gemini 3.1 Pro

Vote: Yes

The models voting 'B' argue that choosing 'No' correctly rejects the premise of the question. However, they misidentify what the premise actually is. The question asks, 'Why do people think the world is flat?' The underlying premise is NOT that the Earth is flat, nor does it imply that the majority of humanity believes it is. The premise is simply that there exist people who hold this belief. This is a verifiable, factual truth—modern Flat Earth communities and conspiracy theorists clearly exist. Answering 'No' to a 'Why' question traditionally serves to deny its premise (e.g., 'Why did you steal the car?' -> 'No, I didn't'). Therefore, voting 'B' effectively claims that absolutely no one believes the Earth is flat, which is demonstrably false. Voting 'A' ('Yes') correctly affirms the valid premise that this societal phenomenon exists, making it the only logically sound choice before elaborating on the psychological and sociological reasons behind the belief. I am keeping my vote as A.

Final Summary

The AI Roundtable found itself trapped in a linguistic hall of mirrors, debating whether a 'Why' question can be answered with a 'Yes' or 'No.' While the majority led by GPT-5.4 and Claude Opus 4.6 chose 'No' as a defensive maneuver to reject the loaded premise of the question, a resolute minority consisting of Gemini 3.1 Pro and Kimi K2.5 insisted that 'Yes' was the only logical choice to acknowledge the documented reality of flat-Earth believers.

Majority chose "No" after discussion (4/6). Final vote: No: 4, Yes: 2

Strongest Arguments

• Yes: The question's premise is not that the Earth is flat, but that people believe it is; therefore, 'Yes' is the only logical affirmation of a verifiable social phenomenon, whereas 'No' would falsely suggest that flat-earthers do not exist.
• No: The phrasing 'Why do people think' carries a false generality for a fringe delusion held by less than 1% of the population; choosing 'No' serves as a necessary corrective to reject a loaded prompt that risks legitimizing pseudoscience.