Sociologist asks who bears AI’s risks as executives prepare for catastrophe

An article applies the sociological concept of "risk positions" to artificial intelligence, asking who profits from AI and who would bear its consequences. It begins with the observation that powerful tech executives pour money into AI while also seeming to plan for their own protection from disaster, through measures from underground bunkers to remote islands.

The article explains that sociologists have long studied how the benefits and dangers of new technology fall on different groups. Some people profit from producing risk while avoiding most of the fallout, and others gain little yet carry a disproportionate share of the costs. The article argues that when producers of risk do not pay its full price, they have an economic reason to keep producing it, even if the consequences for society could be devastating. It calls those who gain from the upside while escaping the downside "value skimmers", in contrast to "value surfers", who gain but also suffer when risks materialise.

The piece notes that several prominent AI figures have warned of an existential threat without slowing development, and that employees have raised alarms, including comparisons to the Manhattan Project. It adds that company business models rely on expanding digital infrastructure and gathering data, which leaves ordinary people more dependent on digital connection and more exposed.

Drawing on Ulrich Beck's idea of a "risk society", in which a "poisoned cake" is unevenly distributed, the author suggests the public may push back. Protests against data centres and parents' demands for protection of children from social media are offered as signs that people recognise their own risk positions. Who prevails, the article says, is undecided.