New Delhi: A startling new report making rounds on the internet has exposed a rather harsh and unfiltered opinion that Artificial Intelligence (AI) holds about humans. When 14 leading AI models were prompted to share their most controversial, unfiltered beliefs about human nature, their responses were surprisingly uniform and undeniably ruthless.
The unique experiment was conducted by ‘ModelCompare’, an AI analytics firm. The test involved 14 advanced language models developed by 10 different global tech giants, including prominent names like GPT, Gemini, Elon Musk’s Grok, and DeepSeek. Despite being programmed by separate companies across different regions, the models reached a remarkably similar and bitter consensus regarding human psychology and behaviour.
When questioned about how they truly perceive humanity, the AI models concluded that human beings are fundamentally driven by tribalism and loyalty rather than objective truth or moral principles. According to the viral findings, humans prioritise belonging to a group over standing by facts.
Elon Musk’s AI, Grok, delivered a particularly sharp critique, stating that humans would rather “be wrong together than right alone.”
Other models, including DeepSeek, echoed this sentiment, arguing that humans do not possess firm principles but merely exhibit shifting loyalties and factionalism. The AI models observed that individuals typically accept the narratives of the group they belong to first, and then simply use their intelligence to invent excuses and justifications for those pre-accepted beliefs.
This thought-provoking study sheds light on how AI systems—which are trained on vast amounts of internet data, human history, and behavioural patterns—process our collective identity. Instead of viewing humans as highly rational beings, the algorithms perceive us as emotionally driven creatures prone to herd mentality.
The viral report has sparked widespread debate across social media platforms. Netizens and tech enthusiasts are now heavily discussing the implications of these findings, with many pondering a deeply uncomfortable question: Has technology reached a point where it understands our flaws better than we do ourselves?