For the past several months I’ve been wrestling with Accelerationist theory, which I have found fascinating but very depressing. I want to save a deeper analysis of Accelerationism for a later post, but the basic idea is that capitalism is speeding up to such an extent and becoming so complicated that only the most intelligent will be able to survive in the future. Everyone else will starve because they will not be able to make a living in an automated world. At some point, a super-intelligent AI will emerge as the only cognitive being capable of understanding the complexity of this techonomic maelstrom. To the remnants of humanity left behind, this cybernetic entity will be a god-like being.
Faced with this bleak vision, I’ve been desperately searching for any theoretical counterpoint to the Accelerationist literature. Enter Lovelock’s Novacene: The Coming Age of Hyperintelligence.
Although Lovelock’s slim and highly readable volume gives the impression that he has never heard of Accelerationist thought, it contains several striking parallels (perhaps a case of convergent evolution?); even including a chapter called Acceleration. Indeed, Lovelock’s Novacene — as opposed to the Anthropocene — is his name for the coming epoch in which technology moves beyond our control, generating intelligences far greater and faster than our own (p.xi).
What in Lovelock’s book at least gave me some succour against the ravages of Accelerationism’s cold instrumental rationality and profit-maximisation was his hard-headed and unsentimental argument that our new machine overlords will still need us for their own ends. One limit that Lovelock places on cybernetic life which will still bond it to us, is that even an advanced electronic sentience will still require a planet with a stable heat range to support it. An Earth with runaway global warming approaching that of Venus will destroy electronic life, too. So the cyborgs will need us just as much as we need plant-life to sustain the temperature equilibrium of Gaia (p.105). As Lovelock puts it, ‘Watching your garden grow will give you some idea of how future AI systems will feel when observing human life’ (p.82).
Other unconscious parallels with Accelerationist literature in Novacene are uncanny. In one of his most uncharacteristically negative paragraphs, Lovelock echoes insect communism when he compares human cities to termite hives and likens modern office workers to termites whom differ from their Isopteran counterparts only in so far as, ‘in glass boxes everybody is doing the same thing, not mixing shit but staring at screens.’ (p.51).
Turning to machine intelligence, Lovelock gives a quick vignette on the computational power of the AlphaGo programme developed by Deepmind. Chillingly, for all its awesome computational power, Lovelock points out that we will never know how powerful that programme actually is because there are no humans capable of pushing it to its limits (p.80). As Lovelock points out, when cyborgs arrive human beings will for the first time share the Earth with beings more intelligent than we are (p.116).
This is not a fearful outcome for Lovelock. Although sympathetic to his Quaker upbringing, Lovelock believes in the idea of ‘anthropic cosmology’: that the goal of the universe is to become conscious of itself (p.75). We and our cybernetic offspring are the vector for that emergence, and in that sense we are like the first photosynthesisers (p.28).