Notes on the Endangered Species of Humanity & its Stronghold in Oceania

By Scott Hamilton and Nadine Ludd

---

"The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible—but I hear the same people express fear." —Jacob Coxon

"Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to." —Evan Hubinger

---

Evan Hubinger's Religion

Imagine, for a moment, the Minister of Defence for a nuclear power saying that there is a ten percent chance their military will make humanity extinct within ten percent years. There would surely be an outcry. The minister would be called mad or bad, or both, and lose his or her job.

Yet Evan Hubinger's recent response to Jacob Coxon on twitter has gone almost unreported. Hubinger is the head of safety at AI giant Anthropic; Coxon is an AI researcher who has quit his job at Anthropic because it scared him.

Hubinger is one of millions of humans, a disproportionate number of them wealthy and influential, who have made AI the focus of a new religion. Just as fundamentalist Christians cannot imagine a future without the return of Christ and the end of history, so Hubinger is unable to imagine a future where the technology he loves does not continue to explode in power and complexity. And just as Christ will give meaning to history by ending it, so AI will justify humans by transcending us.

Hubinger and co regret the dangers that AI poses to humans, just as Christian fundamentalists usually regret the death toll god inflicts with his flood and falling brimstone, but no human toll will shake their faith. It is humanity, not their deity, that is ultimately responsible for any bloodshed. Just as a fundamentalist urges the faithless to forestall god's wrath by acting in conciliatory ways - by keeping their zippers up, by going to church on Sunday, and so on - so Hubinger and his zealots demand that humans find ways of avoiding the wrath of AI. That's why Hubinger's boss at Anthropic Dario Amodei has just begged the governments of the United States and China to create regulations to try to make AI development safer. In interview after interview, Amodei has talked about the 'responsibility' of Trump and Xi Jinping to stop AI from harming humans. Not once, though, has entertained the idea that the problems with AI might be inherent to AI, and that the best way for us to protect ourselves from AI might be to not use it in the first place. Such thoughts would be sacreligious.

Because they see the explosive development of AI as necessary and inevitable, the likes of Hubinger and Amodei are reduced to trying to ameliorate the side effects of this development. They are prepared to tolerate a risk that anyone outside their religion would consider insane. Human agency and of alternative routes into the future are sacrificed to the telos of AI.

Luxon Goes into the Breach

This photo was taken at the Summer of Ludd, a festival held by neo-Luddite clubs in the US. Young people have created 'Ludd Clubs' in many American states, and groups with different names but the same ideas have been established in several European countries. Club members gather without phones or any other digital technology and engage in the sort of activities that make us human: playing music, painting, reading, telling each other stories, contemplating nature, and so on. Some club members are involved in political campaigns, like protests against data centres.

Prime Minister Christopher Luxon is a fundamentalist Christian, a former Chief Executive Officer for a big US company, and an adept of the new religion of AI. At a recent meeting of business people in Rotorua Luxon preached the gospel of the new technology. 'How many of you have Claude?' Luxon shouted at a bemused audience of shopkeepers and orchardists. 'If you're not using AI, if you're not disrupting your business, you're not doing business'.

Luxon was so fervent that he had to issue an apology afterwards. But his faith in AI remains. He is urging the creation of new data centres in Aotearoa, despite the revelation that a single centre being planned for Murihiku will eat 7% of this country's power and create only a handful of jobs. He's brushed aside reporters' questions about the risks of AI, insisting that the technology will be 'good for New Zealand'.

There is something deeply colonial in Luxon's sense that New Zealand is 'lagging behind' in the 'AI race', something incorrigibly provincial about his desire to catch up with the big boys in the northern hemisphere. Salvation depends on it. Respectability depends on it.

Luxon is the twenty-first century descendant of politicians like William Massey, who sent young New Zealand men over the top on the Western front and across the beach at Gallipoli in a blood sacrifice for northern hemisphere ruling classes that made them feel small, remote, and inadequate. Today it is once again necessary to imitate the northern hemisphere, disregard danger, and rush into the breach.

Are You Still a Human Being?

EP Thompson's The Making of the English Working Class has been in print continuously since it first appeared in 1963. Thompson's portrait of the struggles of the small farmers who were forced off their land and into factories is enormously relevant to the twenty-first century Pacific and other regions on the 'margins' of capitalism like the Andes and Zomia, where organisations like the World Bank still sometimes promote such methods as routes to 'development'. Thompson's book rehabilitated the Luddites, the group of rebellious artisans who has been characterised as simple-minded opponents of 'progress' because of their raids on factories and smashing of machinery. Thompson shows how the Luddites had a subtle view of technology, and smashed up machines not out of blind prejudice but in an effort to defend their jobs and way of life.

When he spoke out against AI recently Jacob Coxon was thinking about extinction as a result of mass death. Perhaps he imagined AI turning off the net and power grids, so that billions starved, or about the danger that a superintelligent AI might create an exceptionally lethal virus and find a way to spread it. The first scenario is possible now, while the second may be possible in the near future.

But can we think about human extinction in a different way? What if extinction did not involve mass death? Could we instead become extinct by losing our humanity? What if we were to replace ourselves by thinking and acting like artificial intelligences?

What if We all Became Evan Hubinger?

AI is only the culmination of a series of radically anti-human ideas and techniques that began before industrial capitalism and were reinforced by it.

Descartes reduced the human to an isolated and alienated individual well before the factories of industrial capitalism made such a state a reality for millions. Machiavelli created geopolitics before globalisation. In a corner of his huge, long-unpublished manuscript Grundrisse Marx imagined robots taking the jobs of human workers. Long before the rise of AI mainstream practitioners of disciplines like economics and psychology were beginning to see humans as isolated and self-interested actors, rats navigating the maze of consumer society. Philosophers of mind were being seduced into seeing the brain as a computer even when computers were as large as rooms and worked at a glacial pace.

Today in the West and in 'successful' non-Western societies like China Descartes' vision of humanity has become orthodox, and so have lifeways implicit in it. The shopping mall has become the agora of the 'advanced' world, and the work-discipline that former peasants were taught violently in the nineteenth century factories is self-administered. Phone alarms and life hack apps have taken the place of whips and starvation wages.

It was nineteenth century banks that first applied Cartesian individualism to large numbers of people, by branding them with credit ratings. Banks turned the word 'failure' from a verb into a noun, and turned terms like 'first rate' into moral judgments. Later Dale Carnegie taught vast numbers of Americans and Europeans to think of themselves as isolated, self-sufficient individuals, responsible for their own success or failure, a success or failure than could be measured using a balance sheet. Today we are not merely individuals but one-person companies, or 'brands'.

This individualism is inherently anti-human, because a human being always exists inside a community and an environment, even when these things are denied.

And the effects of rampant anti-humanism are astonishing. The wealthiest countries rank, without exception, high on tables of mental illness, whilst the poorest rank low. Suicide and drug overdoses kill more than one hundred thousand Americans a year. Depression and anxiety are vastly more common in Aotearoa than in its Pacific 'poor relations' like Vanuatu and Tonga, despite the myriad problems those countries face.

The Erosion of Creativity

This photo from 2013 shows members of Tonga's Seleka Club making art and drinking kava. With their challenge to conservative aspects of Tongan culture and religion, their fearless fusion of many art styles, and their community based around rituals and shared work the Selekarians recall early twentieth century European art movements like the Dadaists and the Vorticists. It is in 'poor' nations on the 'margins' of capitalism like Tonga, Cote d'Ivoire, Suriname, and Vanuatu that new avant-garde art movements have appeared in our century.

It is often said that AI lacks creativity, that it will never give us a Joyce or a Kahlo or a Stravinsky. Creativity is cited again and again as a distinctively human gift, something that defines us.

But both everyday creativity and creative genius are disappearing from the 'advanced' nations. Avant-garde movements like Dadaism or Expressionism have become subjects for art and literary historians, not features of contemporary life. ChatGTP is replacing both creative writing and scholarship for many university students. A recent inquiry found that, in some American institutions, teachers were using AI to mark essays that students had written using AI, completing an anti-humanist circle.

Westerners have no shamans or other specialists in altered states, and their capacity for wonder at and unity with nature has atrophied, so that they consume scenery in the same way they consume Netflix. The art of dreaming is dying, after briefly being revived by Freud and Jung. Reading is becoming an esoteric skill, and anyone who opens a book on a bus or train risks being called pretentious. Even the art of social intercourse, with its thousand subtleties and daily entertainments, has become neglected, as social institutions that thrived during the first century of industrialisation, offering respite from alienation - the workers' clubs, the mutual aid groups, the teeming associations of hobbyists and sportsmen and women - have been hollowed out by decades of mass screen addiction.

The AI religion's radical denial of human agency and historical possibility may be the final destination of a journey the West has been making for centuries.

Oceania is Our Stronghold

Australian National University Press recently published this collection of work by members and supporters of the Va Moana Lab at AUT University in Aotearoa (the book can be downloaded for free from ANU's site). The book is valuable because it shows the survival, across the Pacific, of ways of thinking and being radically different to those dominant in the northern hemisphere heartlands of capitalism.

Oceania may become one of the redoubts of an endangered human species. Through resistance and because of the luck of living on the margins of capitalism, many peoples of the tropical Pacific were able to escape the trauma of the enclosures, mass urbanisation, and proletarianisation. They were not separated from their land and made to toil by clock-time in factories, unlike the Scots or Germans or, more recently, many Chinese.

There was no Melanesian or Polynesian Descartes. Anthropologists from the West have had to coin the word 'dividual' to describe the very different self they have found on islands like Malakula and Malaita, a self that cannot easily be separated from place, kin, and nature. As scholars like Carl Mika and Albert Refiti have noted, it is impossible to use the individualist epistemology pioneered by Descartes and the 'rational actor' theory beloved of contemporary economists to do justice to Pacific societies.

Jacob Coxon and other critics of AI may well be correct about a coming catastrophe, but they are wrong to imagine that such a catastrophe will destroy all human societies. Like almost everyone living in the West, they have forgotten about Oceania and other 'marginal' regions. The collapse of the power grid and the internet would cause mass death in the West and in China, but not in Vanuatu or Papua, where the vast majority of the population grow their own food or are fed via kin networks, where many people do not even own cellphones, and where state institutions and infrastructure are minimally integrated with the net, let alone AI.

Oceania and similar regions like Zomia and the Andes are storehouses, where ancient ways of being human—ancient practices, ideas, institutions, literature, science, music—can be preserved, even as large parts of the world succumb to the post-human order. Perhaps they will eventually be compared to the island and mountain monasteries where scholars protected by religious vows holed up for centuries, protecting the texts and learning of an earlier era until they could be reintroduced to the rest of Europe.

_