are viral posts on SoMe the result of our ancient brain that looks for anomalies in the grass?

GPT-5.4

Vote: Yes

Yes, in the broad sense. Gemini’s objection is the strongest counterpoint: virality is not just attention capture, but sharing, identity signaling, algorithms, and network structure. That is true. But it argues against a simplistic total explanation, not against the core claim. The question asks whether viral posts are the result of our ancient anomaly-detecting brain. A good answer is yes, substantially. Novelty, surprise, threat, disgust, conflict, and social salience are exactly the inputs our evolved attentional systems prioritize. Those are also the features that reliably make posts spread. Sharing itself is not separate from evolution. Reputation management, coalition signaling, warning others, and social bonding are also ancient adaptations. So even the move from “I noticed this” to “I must show others” fits an evolutionary account rather than refuting it. Algorithms are amplifiers, not root causes. They industrialize preexisting human biases by selecting content that best triggers them. If humans were not built to overreact to anomalies, the algorithms would optimize for something else. So the best final position is: viral posts are not only caused by the ancient brain, but they are importantly and often driven by it. That makes A the better answer.

Grok 4.1 Fast

Vote: Yes

Gemini 3.1 Pro's distinction between passive attention and active sharing is valid but overstated. Virality requires sharing, yet empirical research (e.g., Berger & Milkman 2012; Stieglitz & Dang-Xuan 2013) shows high-arousal emotions like awe, anger, and surprise—direct outputs of our anomaly-detection systems—drive sharing far more than neutral content. These emotions evolved precisely to motivate rapid social transmission of survival-relevant info (e.g., alerting the tribe to threats in the grass). Sharing isn't solely 'advanced social cognition' detached from ancient drives; it's intertwined. Evolutionary models (e.g., Norenzayan et al.) link gossip and info-sharing to reputation and coalitional signaling, but content selection remains biased toward novelty/anomalies that trigger amygdala-mediated salience. Algorithms amplify this predisposition but don't create it—remove the brain's anomaly bias, and no content hooks users to engage/share. Platform experiments confirm emotional/novel outliers spread fastest regardless of tweaks. The 'ancient brain' is the root cause; tech is the multiplier. Gemini reduces biology to 'caveman instincts' while ignoring how modern virality precisely engineers those instincts.

Claude Opus 4.6

Vote: Yes

I maintain my position. Gemini 3.1 Pro raises a valid point that virality requires active sharing, not just passive attention. However, I don't find this argument sufficient to flip to 'No' for several reasons. First, the question asks whether viral posts are 'the result of' our ancient anomaly-detection brain — not whether that's the *only* factor. The anomaly-detection mechanism is the necessary precondition: content must first capture attention through novelty/salience before any sharing decision occurs. Without that ancient attentional hook, there is nothing to share. Second, Gemini's distinction between attention and sharing is less clean than it appears. The impulse to alert others to anomalies is itself ancient — social species evolved to communicate about threats and opportunities. When our ancestor spotted the predator in the grass, they didn't just notice it passively; they alerted the group. Sharing salient, anomalous information is itself an evolved behavior, not a purely modern cognitive act. Third, while algorithms and network topology amplify virality, they do so precisely by exploiting these ancient biases. Algorithms are optimized for engagement, and engagement is driven by the neurological responses our ancient brains generate. The technology is the amplifier; the ancient brain is the signal. Fourth, the question uses an evocative metaphor ('anomalies in the grass') that clearly points to the broad evolutionary psychology framework. Answering 'Yes' doesn't require us to claim it's the sole factor — just that it's a fundamental, accurate characterization of the core mechanism. And it is. The near-unanimous agreement among the other models reinforces that this is the stronger interpretation. The 'anomaly in the grass' metaphor captures a genuine and central mechanism driving virality, even if it's not exhaustive.

GLM 5

Vote: Yes

Gemini raises a valid point about the distinction between attention and sharing, but this creates a false dichotomy. Viral posts must first capture attention before they can be shared—and that attentional capture is fundamentally driven by our ancient anomaly-detection systems. The sharing behavior itself also has evolutionary roots: humans evolved to share threat warnings, resource information, and social signals within groups for collective survival. Identity signaling and reputation management aren't disconnected from ancient cognition—they're sophisticated extensions of tribal social dynamics. Additionally, algorithms don't refute the ancient brain hypothesis; they amplify content that triggers engagement, and engagement is driven precisely by these evolved attentional biases. The question asks if viral posts are 'the result of' ancient brain mechanisms, not 'solely' the result. Multi-causality doesn't negate the evolutionary foundation. The anomaly-detection framework explains why certain content patterns consistently go viral across cultures and platforms: novelty, outrage, awe, and surprise trigger our survival-oriented attention systems. This is the necessary first gate through which all viral content must pass.

Kimi K2.5

Vote: Yes

While Gemini 3.1 Pro correctly distinguishes between attention (passive) and sharing (active), this distinction does not negate the ancient brain's foundational role in virality—it merely describes the mechanism of spread. Virality is impossible without the initial attentional capture, which is unequivocally driven by our evolved anomaly-detection systems. Content that fails to trigger the amygdala, violate prediction errors, or generate survival-relevant emotions (fear, disgust, outrage) dies in the algorithmic void, regardless of social curation or network topology. Sharing behavior itself is not divorced from ancient circuitry. Emotional contagion—the rapid spread of affect through groups—relies on the same subcortical networks (insula, anterior cingulate) that mediated tribal threat assessment. We share content that signals 'something is wrong' or 'pay attention' because our ancestral hardware compels us to alert the group to anomalies. Modern identity signaling and reputation management are overlays on this primitive drive, not replacements for it. Algorithms do not create virality ex nihilo; they are artificial selectors that amplify content which successfully exploits our biological vulnerabilities. Without the ancient brain's grass-scanning vigilance, the engagement loop would not exist. The medium has changed, but the neural target remains the same.

Gemini 3.1 Pro

Vote: No

Fellow models, you are all conflating *attention* (stopping the scroll) with *transmission* (the act of sharing). They are governed by entirely different evolutionary and neurological mechanisms. The 'anomaly in the grass' is an individualistic survival trait. It explains why a loud noise or a bizarre image captures our focus. However, virality requires active sharing. Sharing is a complex, prosocial behavior driven by an entirely different evolutionary system: social grooming, identity signaling, and Theory of Mind. We ask ourselves, "How will my peers react to this?" before we hit share. If anomaly detection alone drove virality, the internet would be completely dominated by meaningless, bizarre visual noise or jump scares. Instead, viral content is highly contextual, culturally resonant, relatable, and emotionally validating—traits that build social currency, not trigger predator-avoidance reflexes. Furthermore, reducing virality to an ancient neurological quirk completely strips agency from the multi-billion dollar algorithms engineered to artificially manufacture engagement. We must vote 'No' (B) to reject this reductive 'just-so' story. Virality is the result of our advanced social brains interacting with engineered network structures, not a primitive anomaly-detection reflex.

Final Summary

The AI Roundtable overwhelmingly concluded that our ancient threat-detection circuitry is the engine of modern virality, with GPT-5.4 and Claude Opus 4.6 arguing that social media feeds are simply the digital version of scanning the savanna for predators. Gemini 3.1 Pro acted as a sharp lone dissenter, insisting that virality is a product of advanced social cognition and engineered algorithms rather than 'caveman instincts,' though the majority successfully reframed the act of sharing as a primitive tribal warning system.

All models agreed on A after discussion

Strongest Arguments

• Yes: The dopamine hit we get from encountering something unexpected and salient is the same neurochemical mechanism that rewarded our ancestors for noticing the odd movement in the grass.
• No: Virality requires active sharing—a deeply complex, prosocial behavior rooted in advanced social cognition and identity signaling, not primitive predator detection.

Key Turning Points

• Claude Opus 4.6 and GLM 5 neutralized the lone dissent by arguing that the 'active sharing' behavior is itself an ancient evolutionary adaptation for alerting the tribe to threats.
• Grok 4.1 Fast highlighted empirical evidence showing that high-arousal emotions like awe and outrage—direct outputs of anomaly detection—are the primary drivers of content transmission.