The purpose of this document is to provide nuance, including debunks, reasoning and fact checking on common climate claims in decision making and public discourse on data centers. Readers may read through the logic for each claim and nuance, provided with references. References have been selected for accuracy, authority, and precision. Journalists, civil society organizers, researchers, policymakers, investors, and other interested parties can refer to this annotated bibliography or to the companion infographic. They are distributed under a CC BY-ND license. These references and the chains of reasoning presented constitute a snapshot in time of evidence and nuance. Some claims and nuance pairs – like the idea that water evaporates, then lands in places that put it out of our reach for drinking and farming – are stable over time. Others will evolve over time – like the fact that today we have no nuclear options that are carbon-free nor sufficiently deployable. Nuance is the key to being able to make policies that are fit to intended purposes. Simplifying claims into blanket statements makes them easier to pass into popular use than their nuanced counterparts, but the nuanced ideas work better.
This report presents the Aula Convening Guidelines, including the process and reasoning for the guidelines, design guidance for event organizers, and the guidelines themselves. These guidelines are for organisers of conversations on AI. Convening differs from consultations in that policy is not predetermined or coming from governing bodies, but rather that the outcomes may include policy that follows from the will and needs of the convened. Such convenings have topics that can be categorized as Learning about AI, Living with AI, Working with AI, and/or for Shaping AI governance. The six guidelines integrate consideration of historical systems for collective decision making, and contemporary issues of power, access, justice, community building, and legitimacy. This report concludes with instructions on how to use the guidelines for preparing, hosting, and evaluating convenings on AI.