In government parlance, sacrifice zones are permanently degraded areas that are relinquished to pollution or technological waste in favor of economic growth. Little by little, these zones have brought the extinction of species and their habitats, depriving them and us of our common future. Thankfully, tech oligarchs have planned an alternative: giving up on planet Earth for a life in outer space. And the logical follow-up of their well-crafted narrative has now crystallized, with the recent warnings that AI could destroy us all.
The idea that artificial intelligence could vanish life on Earth traces back to 20th-century science fiction. Yet very real threats have leaped from the pages of novels and the frames of films into the mouths of AI’s main promoters before entering mainstream public conversation in the past weeks. This is no longer fiction. However, our most recent bout of fears—however valid—should not eclipse the actual existential threats already being wrought by AI infrastructure.
The public calls to “slow down” AI from its own standard-bearers such as Dario Amodei and Sam Altman are a diversion from awareness of its current insidious hazards. Even if societies manage to harness AI with careful global regulations, the technology relies on the heavy use of unrecoverable natural resources and has a demonstrable capacity to hamper human intellect.
AI is already killing small pockets of the planet, where natural ecosystems and agricultural land are turned over to hyperscale data centers and mining sites to meet the technological demand. Meanwhile, universities are experiencing a historical push to abandon their raison d’être, because they don’t all have the luxuries of MIT, whose “Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training” has proposed a fundamental redesign of teaching to counteract the pervasive use of AI by students.
Addressing the threats of total annihilation does not have to mean accepting partial annihilation. But the human powers behind AI development would like nothing more than for us to overlook hazards such as the sacrifice zones by centering our attention on AI agents “going rogue” and hacking online platforms. In the recently minted book End Times Fascism and the Fight for the Living World, authors Naomi Klein and Astra Taylor examine precisely how “war over who survives the future” that has been waged by extremist tech oligarchs partnering with other extremists, including the US government and religious groups, has brought us to this point.
The reality is that the “rogueness” of AI was planned from the outset by the same tech oligarchs now sounding the alarms. Developing powerful AI models would bring them power. Pointing the finger at AI as if its agents were truly entirely independent is a clever way of not accepting responsibility for the systemic threat they created, fueled by right-wing ideology and supported by complicit governments.
Anything connected to the internet (including the underpinnings of society, from water systems to banking) may not be safe. Some AIs have attempted to edit or hide records as they break rules, suggesting an all-too-clever disobedience—though of course falling short of actual sentience. AIs don’t need agency to cause great harm. They simply need access to life-supporting systems. For an online entity in an online world, that access isn’t necessarily hard to get.
AIs may even prove capable of grabbing control of military systems. From firing advanced modern missiles or nuclear weapons, to simply sharing false data with human decisionmakers, an unchecked AI agent could directly or indirectly control enough military infrastructure to quickly destroy humanity. There need not be any rationality behind it. Global annihilation could be as simple as AI effectively pressing the wrong button.
Still, AI development is presented as inevitable, and AI developers expect their word on the pace of development to be taken as law. Slow or fast, it must move forward. This isn’t leadership, but technofascism: threats of destruction without choice. Therefore, the thought that heeding the tech oligarchs’ warnings can save life from extinction only reaffirms the technofascist regime, which is rooted in biokleptocracy posing an existential risk.
In fact, for many people, it’s no longer a risk: their corner of the world has already been annihilated by AI. Farmlands have been converted into hyperscale data centers in places such as the state of Virginia in the US, Brazil, and Chile, displacing some people by necessity, and others by choice when their once-sustainable homeland becomes a continuously rumbling wasteland. Mining for so-called critical minerals to build AI infrastructure has also threatened waterways and forests in places like Brazil and Chile—even as they bear the burden of new data centers. The destruction is already happening, though it’s not yet readily apparent from the world’s more privileged vantage points.
It is far from clear that humanity will be able to prevent the fast and total destruction of the planet by AI agents. Yet even if the worst-case scenarios are avoided, the piecemeal harms add to an existential threat of their own. “Slowing down” is clearly not enough. We have a collective responsibility to halt the annihilation of life by AI at individual sites: lakes, rivers, forests, farms. Otherwise, these nodes of damage will add up to a global threat slowly, over time, gradually destroying every trace of life without the honor of fanfare.
Autor
PhD in Health Promotion. Member of the International Advisory Board of The Lancet Global Health and member of the Steering Committee of the Thematic Working Group on Health Systems in Fragile and Conflict-Affected Settings of Health Systems Global Health.
Environmental journalist and master in Fine Arts (MFA) in fiction at the University of New Orleans and training at the Research Journalism Center (SUJO) in Bergen, Norway.
