The yellow sunlight and muted grey skies of these mornings look like something is burning far away. Both colours flatten everything they touch, painting walls and floors in the same milky gold against the heavy charcoal above.
The common factors of entanglement and complexity that mark discussion of AI and ‘nature’ seem to me to be getting louder. AI obviously has practical applications in Earth sciences such as in studying migration but there are more abstract and inarticulable parallels that folks are chipping away at. Very often AI is used as a technical framework for comprehending or relating to ecology. For example, I’ve admired Foam’s Random Forests as a project attempting to unpick the application of machine learning to environmental science, asking if a machine could or should become ‘environmentally literate’ and whether we could comprehend this literacy. Daisy Ginsberg’s The Substitute is a sharper critique of this way of thinking; desperate obedience to the hype of AI to solve climate collapse while species, like the Javan White Rhino in the project, go extinct. Both of these projects and others like them point to a strained and unrealised relationship between humans and our two existential threats – a planet that is turning on us and a tool that may decide to. And of course this connection isn’t new, the connection between ecology and computation and the conjoined history of those sciences is well explored through amongst other things, Adam Curtis’ All Watched Over By Machines of Loving Grace.
I think a part of the increasing proximity between critical approaches to these fields is in trying to understand and articulate systems in new ways. Machine learning as a form of AI is perhaps most well known for being inexplicable; neural networks draw connections between input and output in ways that are often alien to human phenomenology or experience. In doing so they can sometimes be more efficient calculators but they display alien cognitive properties. Simultaneously there is a turn against reducing natural phenomena to human terms; recognising and embracing the complexity of ecology and de-centering ourselves in it. AI, as a technical project, opens the potential to develop tools that can help us do that. The relationship goes the other way to – it may be better to consider AI as new, other form of intelligence, interpreted perhaps through a renewed popular fascination with cephalopods and mushrooms than as simulating human cognition and behaviour.
Augury also took a stab at unpicking this relationship – in particular the way humans tend to interpret complex patterns as some marker of cognitive intelligence or meaning. The ancient Greeks and Romans believed the complexity of the movement of birds implied some sort of divine will and by extension, meaning could be drawn from them, predictions made. Augury satired this with the movement of planes – as if secret future prognostications were being written in the sky. AI is often spoken about in the same hushed tones; that meaning can be drawn from the complexity of the underlying system. Sascha Pohflepp, in drawing on another ancient Greek divination tool – marble, referred to this as pattern agnosia.
I only bring this up because of two articles I came across last week. Firstly, We Make Money Not Art carried a review of AI In The Wild by Peter Dauvergne which concerns itself with the more pragmatic entanglements of AI and nature such as the enormous cost of running AI systems and the way that it may be deployed to drive the worst excesses of exploitation in agriculture and mining. Eirini Maliaraki wrote a piece taking a critical but more optimistic view – that AI has the potential to improve energy demand scheduling and mitigate the worst impacts of climate change through better preparation while also drawing attention to the energy impact of these systems.
It’s an interesting time for this stuff, I wish I had time to delve into it more deeply. I’m drawn to the potential of machine learning to leverage humans into a new way of thinking about their place in the universe – like an alleyway encounter with Cthulu.
Small Stuff
Honestly I don’t know what I did last week but there was a lot of it as a consequence I didn’t browse the Internet much and don’t have anything on my list. I got to meet some new folks again so if you want to hang out and chat let me know, I’m always around.
Natalie and I are going to bring Scrycasts back we decided so that should be fun.
I suppose that’s it for now. Love you, bye.