In the city of Belém, Brazil, the 30th United Nations Climate Change Conference (COP30) began on the 10th of November and will be continuing until the 21st, gathering international politicians. An entry point to Brazil’s lower Amazon, the city and its surrounding is home to many Indigenous Peoples, thousands of which have converged to the conference to protest outside the venue and attempt to steer decision-making towards better protections for the Amazon and its inhabitants.¹ In preparation for the event, Oxfam released a report entitled “Climate Plunder: How a powerful few are locking the world into disaster.” ² It presents new data that shows that the lifestyle of some super-rich are ‘blowing through the world’s remaining carbon budget,’ and lays out the political and economic influence that billionaires exert to keep a majority of people hooked on fossil fuels for their own interest (i.e, to maximize their own profit).

Recent discourse around AI has seen a lot of talks of “AI bubble:” the idea of a bubblegrowing market amid the current AI boom that could burst, even after having affected the broader economy. This would suggest that accompanying phenomena, such as the growing concentration of power within the AI industry, may be ephemeral. However, whether or not this ‘bubble’ is solid or even real, it suggests an underlying condition: a strong and self-amplifying omnidirectional sphere that, given its current and projected economical power, may just end up steering some of humanity’s major decisions.

This applies to climate-related decisions, as well as determining the fate of technology itself, one of the major challenges of our generation. Indeed, the decisions which are slipping through our fingers to fall in the hands of the tech giants, also include whether we attempt to figure out AGI (Artificial General Intelligence) and ASI (Artificial SuperIntelligence); and if so, what form that attempt will take. Notice that, following the logic presented by leading figures in AI safety,³ we write ‘attempt’ in the singular. Following their logic, if we consider some probable arguments about AGI and ASI, the first form of such intelligent systems could radically determine humanity’s faith. This major importance would rule out the probability of getting a second chance to figure out alignment and related questions.

Like many liberatory and activist movements highlight,⁴ everyone’s liberation is bound, as was first brought up by Indigenous activist Lilla Watson at the 1985 UN Decade for Women Conference in Nairobi when she said:

If you have come here to help me, you are wasting your time. But if you have come because your liberation is bound up with mine, then let us work together.⁵

In this case particularly, the relation between ecological struggles and technological progress is also reinforced by direct causal links. As François Jarrige explains in AI and Law: A Critical Overview, we have passed the point where ecological activism will bring on an ecological transition on its own. Instead, we must “free ourselves from the hold of technological progress:” dealing with the impacts (present and forthcoming) of AI has become both a pre- and co-requisite to tackle the climate crisis.⁶ Indeed, ecological struggles will inevitably follow from the rapid expansion of a resource-intensive AI industry. Already today, most readers must’ve encountered some of the alarming statistics around the consumption levels of existing AI systems, such as the fact that GPT-4’s total environmental footprint is equivalent to the yearly energy use of a small town. To tackle the climate crisis, the trajectory and pace of AI must inevitably also be recalculated.

Our liberations are bound, all the way. Actions must be taken to address the multifaceted problem of oppression that has pervaded every sphere of human life, especially when technology fueled by “the powerful few,” in turn, enables that group to establish and maintain their power across society.