Atlântimor

A name, a label, and the promise of salty Atlantic winds coming ashore. Clearly, wine decisions are rarely rational, but in this case, I backed mine up with a few gold medals and recommendations.

Passing through Lund earlier today, one could feel that the semester is just around the corner, with students everywhere. I cannot say that I am entirely envious of them, starting their studies at a time when artificial intelligence is reshaping so many certainties. On the train, I revisited my 2013 paper Ambivalence, Irony and Democracy in the Anthropocene, which has remained one of my most cited publications. In it, I explored what I saw as a basic dilemma of modernity: either we find technological ways of transcending ecological limits, or we accept those limits and somehow create the political institutions necessary to enforce them on a global scale.

Neither path looked particularly plausible. Radical technological modernization required a Promethean faith in human ingenuity that many environmentalists rejected, while comprehensive degrowth would demand levels of political coordination and personal restraint that few people appeared prepare to accept. For all the rhetorical appeal of “sufficiency”, its advocates rarely seemed to see any contradiction in expecting rising retirement incomes, ever more advanced medicine, and ever more elaborate forms of inconspicuous consumption in their own lives. My conclusion was therefore deliberately ambivalent: neither radical modernization nor comprehensive degrowth seemed politically within reach.

Reading the paper now, it is striking how much I treated advanced machine intelligence as belonging to the realm of science fiction. Today, we know better. In one sense, AI gives the anti-modern intuition perhaps its strongest possible case: maybe there really are doors that humanity should never open. Yet actually keeping those doors locked would itself require unprecedented global coordination over technological development, probably indefinitely. The more powerful the technology becomes, the more compelling the argument for restraint may become – and the less plausible it seems that such restraint could ever be universally enforced.