DRAFT, IN PROGRESS
Three different types of language are distinguished here: The languages of biological life, of humans (natural human language), and of artificial life. This perspective on language and mental models represents a step towards simplifying later analysis. It helps us focus on the respective climate-relevant decision-making processes of different lifeforms, and avoid anthromophizing non-human activity.
This linguistic angle allows adressing multiple global problems, defined as language-driven processes. For efficient analysis, I combine Wittgenstein (philosophy of language) and Max Tegmark (physics: machine learning (ML) / AI).
A focus on language may help handling present and future disruptions, including at COVID–19-scale, elegantly based on their mathematical properties. For example, pandemics and ML in a narrow sense fall under ‘Language 1.0,’ societal reactions to them under ‘Language 2.0.’ This difference looks trivial on paper. But in practice, people use false mental models all the time, as Q1 2020 has amply shown.
That we need alternatives is easy to see. We know that building the systems to remove billions of tons of carbon debt accumulated in the atmosphere is a generational project unlike all others (Rockström, Johan et al. 2017; Hayhoe and Kopp 2016).
Analyzing multiple global problems across different time horizons is allows policy analysis at climate-relevant timescales, potentially stretching from split-seconds (climate impacts from ML-driven decisions) to decades if not millennia into the future (feedbacks in the carbon cycle). Even COVID–19 is changing societies, the international system, and future climate policy trajectories already in its first phase.
- Rockström, Johan et al. 2017. “A Roadmap for Rapid Decarbonization.” Science 355(6331): 1269–71. doi: 10.1126/science.aah3443
- Hayhoe, Katharine, and Robert E Kopp. 2016. “What Surprises Lurk within the Climate System?” Environmental Research Letters 11(12): 120202. doi: 10.1088/1748-9326/11/12/120202