During the early months of the Covid-19 pandemic, global populations demonstrated a largely spontaneous solidarity driven by fear and a sense of duty, signaling a collective response to an imminent crisis. This phenomenon, while not reaching the levels of existential risk discussed in specialized circles focused on preventing species-wide catastrophes, highlights a broader pattern in how societies anticipate and respond to threats labeled as apocalyptic. The term apocalypse itself, historically meaning revelation rather than ultimate destruction, frames how futures perceived as catastrophic are processed and managed.

These patterns are evident not only in public health crises but also in climate change responses. Societies often expect disaster trajectories to rise exponentially and demand solutions of equal speed and scale. When interventions fail to produce rapid, sweeping improvements, reactions tend to tilt toward disappointment, sometimes described as “doom.” The global experience with Covid-19 exemplified this dynamic: outcomes were simultaneously better and worse than many initial projections suggested. Similarly, ongoing climate challenges, despite progress in decarbonization efforts, maintain a looming threat of severe consequences, although the likelihood of truly catastrophic global warming scenarios is arguably diminishing.

Artificial intelligence presents a more urgent and complex case. Experts suggest the critical window to avert significant disruption has likely closed, yet public discourse frequently defaults to narratives of inevitable catastrophe. In May, Amodei, the CEO of Anthropic, provided a stark forecast, predicting that ideology itself may not endure technological advancement. He envisioned a future in which superintelligent systems could resolve political disputes over resources, policy, and social organization so thoroughly that democratic debate would appear primitive in hindsight.

This perspective reflects a faith in the exponential logic of technological progress but invites critical scrutiny. Historically, ceding complex social decisions to an all-knowing entity—whether human or artificial—raises profound questions about the appropriate balance between risk tolerance, progress, and public anxiety. If artificial intelligence is imagined as an infallible political oracle, the dilemma becomes how societies should ethically respond to the potential for disaster while managing the competing impulses of innovation and caution.

Within the AI industry itself, concerns about catastrophic risks are not merely theoretical. Some researchers estimate nearly a one-in-five chance of near-term human extinction, and company leaders frequently acknowledge the possibility of severe negative outcomes. Nonetheless, investment and development continue unabated, with the sector accelerating despite these warnings.

Public opinion appears at odds with the industry’s trajectory. Polling frequently shows Americans favoring prohibitions on superintelligent AI by more than a two-to-one margin, with those expressing apprehension far outnumbering enthusiasts. Additionally, a substantial majority believe AI is advancing too rapidly—by more than 30 to one—suggesting widespread unease that far outstrips acceptance. These figures raise questions about the extent to which democratic oversight can still influence technological development or whether the pace of AI progress has already placed control beyond effective public governance.

As artificial intelligence advances, these tensions between innovation, risk, and societal control underscore an unsettled future, one that will require navigating unfamiliar terrain between technological possibility and the preservation of human judgment.