WE HAVE DISCLOSED THE DANGER. NOW INVEST.

Oct. 2026

Civilization announces a power that could threaten humanity, then invites investors to finance its expansion. The warning enters the prospectus. Money approaches. Something in this encounter exceeds ordinary recklessness: danger can fascinate, destructive capacity can confer prestige, and the spectacle of power can disarm the judgment required to confront it.

Anthropic’s October 2026 IPO filing warns of catastrophic or existential AI risks while seeking capital for a business promising scientific breakthroughs and growth. Disclosed dangers include malicious uses and possible loss of control (Associated Press 2026). Destruction is not its announced purpose. The contradiction lies in acknowledging such possibilities while soliciting the means to expand.

The atomic mushroom cloud supplies a terrible image of this fascination: human ingenuity contemplating its own capacity for annihilation. Look what we can do. Horror and admiration can occupy the same gaze. With AI, the spectacle enters ordinary life and answers back. It writes, predicts, imitates, astonishes. Each demonstration invites another surrender of attention. Power becomes palpable; its consequences remain dispersed.

For me, the long-term effects I investigate are more radical than those of the atomic bomb. Computational systems can act upon the conditions through which populations perceive, desire, think and organize resistance. The loss of discernment belongs inside this thesis. The machinery can help weaken the capacities required to judge its expansion, while fascination makes that expansion desirable. The spectacle supplies both the force that unsettles judgment and the achievements offered as reasons to applaud.

An exhausted attention encounters another astonishing performance. Fear demands another update. The promise of being left behind makes withdrawal feel dangerous. Who has time to connect the demonstration to the infrastructure, the infrastructure to dependency, dependency to the disappearance of alternatives? Understanding requires continuity. The next marvel has already arrived.

Debord’s account of spectacle as a social relation mediated by images helps formulate the problem (1967). Here, the image of unprecedented intelligence can acquire more authority than our examination of what its deployment does. My hypothesis is that danger itself can become promotional: if this power could overturn civilization, surely we must possess it, invest in it, remain close to it.

Meanwhile, bodies bear the demands on attention and recovery; societies confront dependencies and consequences exceeding their powers of correction. Dark Machinic Epistemology deepens the fracture: operational knowledge accumulates behind proprietary walls while public understanding struggles to connect the effects.

The most disturbing possibility is that civilization will finance the erosion of its discernment and experience the transaction as participation in greatness.

Liviu Poenaru

References

BBC News. 2026. Report on Anthropic’s planned stock market listing and warnings about AI risks. https://www.bbc.com/news/articles/cq398p1wd1evoDebord, G. 1967. La Société du spectacle. Buchet-Chastel.