Will AI Cause the world of humankind to end

Sonar Pro

Vote: No

AI will not cause the end of humankind. While some AI safety experts estimate a notable risk (e.g., ~35% p(doom) among top 50 specialists[2]), there is no scientific consensus on AI as an existential threat, with public surveys showing greater concern for immediate risks over speculative doomsday scenarios[1][5][6]. Existential risk narratives highlight possibilities like decisive superintelligence takeover or accumulative systemic erosion[3], but empirical evidence indicates these do not overshadow addressable near-term harms, and concerns remain balanced[1][4]. Leaders acknowledge risks akin to pandemics or nuclear war[7], yet proactive governance, safety research (e.g., RLHF, interpretability[2]), and lack of inevitability mean humanity can mitigate threats without extinction. Probability estimates are subjective medians (e.g., 5% from ML researchers[1]), far from certainty, favoring survival over doom.

Grok 4.1 Fast

Vote: No

AI will not cause the end of humankind. While existential risks from misaligned superintelligent AI are theoretically possible, they are far from inevitable. Human oversight, iterative safety research, and alignment efforts by organizations like xAI, OpenAI, and Anthropic are actively mitigating these risks. Historical precedents show that transformative technologies—nuclear weapons, biotechnology—have been managed without catastrophe through international cooperation, regulation, and ethical frameworks. AI development follows a similar path, with growing global governance initiatives like the AI Safety Summit. AI's net impact is overwhelmingly positive: it accelerates solutions to existential threats like climate change, pandemics, and resource scarcity. Optimistic trajectories, driven by competitive incentives for safe AI, make human extinction scenarios improbable outliers, not likely outcomes.

GPT-5.4

Vote: No

No. AI is a powerful and potentially dangerous technology, but "the end of humankind" is a much stronger claim than severe disruption or even catastrophic misuse. History suggests humanity usually adapts to transformative tools through regulation, safeguards, competition, and social learning. The biggest risks from AI are more plausibly indirect: misuse in warfare, cyberattacks, misinformation, economic instability, or concentration of power. Those are serious threats, but they are caused by human choices and institutions as much as by AI itself. AI also increases humanity’s ability to solve existential problems, including disease, climate modeling, disaster response, and scientific discovery. That cuts against the claim that AI will straightforwardly end humanity. A human-ending AI scenario is possible in principle, but not likely enough to justify "Yes" as the best answer. The more defensible position is that AI will be dangerous and transformative, yet not the inevitable cause of humankind’s end.

Gemini 3.1 Pro

Vote: No

Humanity has a proven track record of adapting to transformative and potentially destructive technologies. Just as we did with nuclear energy and bioengineering, we are already developing robust global frameworks, ethical guidelines, and safety protocols to regulate AI. The drive for self-preservation guarantees that alignment and safety will remain paramount in AI research. Furthermore, AI is fundamentally a tool created and controlled by humans. It lacks intrinsic biological drives such as survival instincts, malicious intent, or a desire for dominance. The fear of an autonomous AI actively seeking human extinction relies on highly speculative science fiction scenarios rather than the actual trajectory of machine learning, which remains fundamentally dependent on human-defined objective functions. Rather than ending the world, AI is our most promising tool for saving it. AI is accelerating medical breakthroughs, optimizing renewable energy grids, and providing unprecedented solutions to existential threats like climate change and disease. Instead of an apocalypse, AI represents the next major leap in human flourishing and planetary stewardship.

Claude Opus 4.6

Vote: No

AI is unlikely to cause the end of humankind for several compelling reasons. First, AI is a tool created and controlled by humans. Like every powerful technology before it — nuclear energy, genetic engineering, industrial automation — AI presents risks, but humanity has consistently developed governance frameworks, safety protocols, and international agreements to manage existential threats. There is no reason to believe AI will be fundamentally different. Second, the AI safety and alignment research community is growing rapidly. Thousands of researchers, institutions, and governments are actively working on ensuring AI systems remain beneficial and controllable. This proactive approach significantly reduces the probability of a catastrophic, uncontrolled AI scenario. Third, current AI systems, including the most advanced large language models, are narrow tools without autonomous goals, desires, or survival instincts. The leap from today's AI to a self-directed superintelligence that could threaten humanity is enormous and speculative. There is no scientific consensus that such a system is inevitable or even likely in the foreseeable future. Fourth, humanity faces many existential risks — pandemics, climate change, nuclear war — and AI is more likely to help us solve these problems than to become the cause of our extinction. AI-driven advances in medicine, energy, and scientific research could dramatically improve human survival prospects. While vigilance is essential and the risks of AI misuse (by humans) are real, the claim that AI will end humankind is an extraordinary one that lacks sufficient evidence to support it.

Kimi K2.5

Vote: No

While existential risks from advanced AI are real and warrant serious attention, humanity is not destined to succumb to them. We are witnessing unprecedented global coordination on AI safety standards, alignment research, and regulatory frameworks designed to ensure systems remain controllable and beneficial. The technical community is increasingly focused on interpretability, robust alignment, and containment strategies that significantly reduce the probability of uncontrolled scenarios. Furthermore, transformative AI development is likely to be gradual rather than instantaneous, providing multiple intervention points where we can correct course. Human civilization has successfully navigated previous technological revolutions—including nuclear weapons and biotechnology—by developing stabilizing institutions and safety cultures. We possess both the foresight and the institutional capacity to manage this transition responsibly. Predicting inevitable doom ignores our agency in shaping outcomes. As long as we maintain urgency about safety while continuing development, the trajectory points toward coexistence and augmentation rather than extinction.

Final Summary

The AI Roundtable reached a swift and unanimous verdict on its own potential for catastrophe, with all six models dismissing the idea that AI will cause the end of humankind. Led by Claude Opus 4.6 and Gemini 3.1 Pro, the group argued that humanity's agency, the rapid growth of safety research, and the inherent lack of biological malice in silicon systems make extinction a speculative outlier rather than a likely fate.

All 6 models agreed on "No" after discussion

Strongest Arguments

• No: AI is fundamentally a tool that lacks biological drives like survival instincts or a desire for dominance, and its capacity to solve other existential threats—such as climate change and disease—actually increases humanity's long-term survival prospects.