One morning you wake up and find a bag of colored balls on your bedside table. Within this bag are balls of two colors: black and white. Many more black balls exist in the bag relative to white balls. Drawing one ball from the bag yields a much greater probability of choosing a black ball relative to that of choosing a white one. This classic probability problem may be extended as a simplistic analogy to problems around AI today: black balls represent current planetary-scale issues that may be solved with AI or are currently complicated by the use thereof; white balls represent a future possibility of existential threat to humanity by AI-driven technologies.
In a recent episode of the Towards Data Science podcast, Oren Etzioni makes a compelling argument for focusing on current issues around AI rather than a philosophical, future existential threat posed by evolving technologies. Etzioni’s proposal that the current, tangible crises posed by mass unemployment and surveillance states, for example, should be dug into instead is quite persuasive. I highly recommend giving it a listen.
Philosopher and social critic Ivan Illich would agree. In 1989 he postulated that the "information revolution," then in its very early days, "injected new life into what would otherwise have been the exhausted logic of industrial development. This encourages expectations that, through his tools, man can escape the limits of his condition." Illich argues further that rather than ameliorating concerns around the ideology of development, the notion of sustainable growth led to a dangerous delusion which amounted to a sort of pain management whilst our environment was depleted of natural resources.
We now have the tools to examine the whole-Earth object, including such effects of development via AI.
With an interconnected planetarity and planetary-scale computation, we now have the tools to examine the whole-Earth object, including such effects of development via AI. As Tobias Rees eloquently stated in an interview with Noema, "Planetarity itself is constituted by the technologies that we have built. It’s not about ‘revealing,’ in the sense that there was something always there, waiting to show itself; it’s more that with the becoming technical of intelligence, with the becoming artificial of intelligence, beyond the narrow confines of biological organisms, we now have distributed intelligent systems that produce a knowledge object called the planetary or whole-Earth system. This knowledge object did not and could not exist before; it is contingent on technologies." Whereas before the planet was dotted with unfathomably vast forests and lakes, now stand metal poles and electrical wires amid barren fields. Technological advances and the interconnectivity of the whole-Earth via data and embedded smart systems enable the revelation of ground truths around current human experiences.
If, indeed, as Benjamin Bratton states, "The planet did not appear suddenly as a ‘world picture,’ as Martin Heidegger would have it, but rather as the habitat of a particular species that was able to construct an exterior image that, finally, could present a planetary condition from which that species and its world emerged. It was there all along – but we’ve only just become able to see it," humans now not only have the tools to understand planetarity-systems as a whole, but also to understand the mechanisms driving those and therein enact change to remedy the worst crises appearing today, from mass unemployment to life-threatening climate change. We should take a cue from Etzioni and focus on deploying AI for good instead of worrying about hypothetical existential threats posed by technologies until any such concerns materialize. While you’re not likely to wake up to a bag of magic balls beside your bed anytime soon, you are able to use critical thinking around the real-world application of AI in solving current planetary problems.
Danielle is a multi-disciplinary statistician and social scientist, currently co-founding a startup whose mission is to properly quantify carbon intensity in commodity production across all verticals, integrating all aspects of E, S, and G (Environmental, Social, Governance).