Deborah W.A. Foulkes
Participation2
- Completed the In-Depth EA Virtual Program
- Attended an EA Global conference
Posts 10
Comments65
Topic contributions1
The very first premise of this post, "Things are getting better." is flawed. Our life-support system, the biosphere, for example, has been gradually deteriorating over the past decades (centuries?) and is in danger of collapse. Currently, humanity has transgressed 6 of the 9 planetary boundaries. See:
https://www.weforum.org/agenda/2024/02/planetary-boundary-health-checks/
Technological 'progress', on balance, is accelerating this potential collapse (e.g. the enormous energy resources consumed by AI). Ord should revise his position accordingly.
Please note that I have been a great fan of Ord's work in the past (particularly The Precipice, which I donated to my school library also), along with his colleagues' work on long-termism. Nevertheless, his latest work (both the blog post and the chapter in the forthcoming OUP book, Essays in Longtermism) feels like a philosophical version of Mark Zuckerberg's Facebook motto, Move fast and (risk) break(ing) things. It downplays uncertainty to an irresponsible degree, and its quantitative mathematical approach fails to sufficiently take into account advances in complex systems science, where the mathematics of dynamical systems and chaos hold sway.
Regarding existential risk of AI and global AI governance: the UN has convened a High-Level Advisory Body on AI which produced an interim report, Governing AI for Humanity, and issued a call for submissions/feedback on it (which I also responded to). It is due to publish the final, revised version based on this feedback in 'the summer of 2024', in time to be presented at the UN Summit of the Future, 20-23 September. (The final report has not yet been published at the time of writing this comment, 6 August.)
https://www.un.org/sites/un2.un.org/files/231025_press-release-aiab.pdf
https://www.un.org/en/ai-advisory-body
https://www.un.org/en/summit-of-the-future
See also report on interim workshop:
https://royalsociety.org/-/media/policy/publications/2024/un-role-in-international-ai-governance.pdf
Have expanded this comment and turned it into a forum post:
I am not seeing the issues posed by uncertainty implemented fully in your tools. I'd like to see an in-depth treatment (and incorporation into your tools) of the position stated by Andreas Mogensen in his paper 'Maximal Cluelessness', Global Priorities Institute Working Paper No. 2/2019:
"We lack a compelling decision theory that is consistent with a long-termist perspective and does not downplay the depth of our uncertainty while supporting orthodox effective altruist conclusions about cause prioritisation."
In my view, if one accepts 100% the implications of maximal cluelessness (which is ever more strongly supported by dynamical systems and chaos theory, the more longtermist the perspective), then the logical conclusion from that position is to fund projects randomly, with random amounts.
The RP team may wish to consider prioritising the study of complexity and dynamical systems etc. as part of their continuing professional development (CPD). I recommend the courses offered by the Santa Fe Institute. You can register for most courses at any time, but the agent-based modelling course requires registration and starts at the end of August: https://www.complexityexplorer.org/courses/183-introduction-to-agent-based-modeling
Stephen Hawking famously once said that the 21st century would be the century of complexity. I wholeheartedly agree. IMHO, in these non-linear times, it should be a part of every scientist's (and philosopher's) basic education.
From a complex systems perspective, the cross-cause cost effectiveness model is inadequate, since it fails to fully take into consideration or model the complex interactions and interdependencies between cause areas. Did you know, for example, that combatting inequality (a global development goal) is also a proven way of reducing carbon emissions, i.e. reducing the existential risk of climate change, which in turn would reduce biodiversity loss (an animal welfare goal)?
I invite the RP team to consider two of many similar examples:
[1] The 2019 paper published in Nature Sustainability by Nerini et al., Connecting climate action with other Sustainable Development Goals:
"Abstract
The international community has committed to combating climate change and achieve 17 Sustainable Development Goals (SDGs). Here we explore (dis)connections in evidence and governance between these commitments. Our structured evidence review suggests that climate change can undermine 16 SDGs, while combatting climate change can reinforce all 17 SDGs but under- mine efforts to achieve 12. Understanding these relationships requires wider and deeper interdisciplinary collaboration. Climate change and sustainable development governance should be better connected to maximize the effectiveness of action in both domains. The emergence around the world of new coordinating institutions and sustainable development planning represent promising progress."
[2] Carbon emissions, income inequality and economic development
Abebe Hailemariam, Ratbek Dzhumashev, Muhammad Shahbaz
Empirical Economics 59 (3), 1139-1159, 2020
This paper investigates whether changes in income inequality affect carbon dioxide () emissions in OECD countries. We examine the relationship between economic growth and emissions by considering the role of income inequality in carbon emissions function. To do so, we use a new source of data on top income inequality measured by the share of pretax income earned by the richest 10% of the population in OECD countries. We also use Gini coefficients, as the two measures capture different features of income distribution. Using recently innovated panel data estimation techniques, we find that an increase in top income inequality is positively associated with emissions. Further, our findings reveal a nonlinear relationship between economic growth and emissions, consistent with environmental Kuznets curve. We find that an increase in the Gini index of inequality is associated with a decrease in carbon emissions, consistent with the marginal propensity to emit approach. Our results are robust to various alternative specifications. Importantly, from a policy perspective, our findings suggest that policies designed to reduce top income inequality can reduce carbon emissions and improve environmental quality.
https://link.springer.com/article/10.1007/s00181-019-01664-x
In my view, Rethink Priorities should take on board the conclusion of these and similar papers by promoting 'wider and deeper interdisciplinary collaboration', and incorporating the results of that collaboration in your models.
Although you have addressed the question of uncertainty in recent work, I am not seeing it implemented fully in your tools. I'd like to see an in-depth treatment (and incorporation into your tools) of the position stated by Andreas Mogensen in his paper 'Maximal Cluelessness', Global Priorities Institute Working Paper No. 2/2019:
"We lack a compelling decision theory that is consistent with a long-termist perspective and does not downplay the depth of our uncertainty while supporting orthodox effective altruist conclusions about cause prioritisation."
In my view, if one accepts 100% the implications of maximal cluelessness (which is ever more strongly supported by dynamical systems and chaos theory, the more longtermist the perspective), then the logical conclusion from that position is to fund projects randomly, with random amounts.
The RP team may wish to consider prioritising the study of complexity and dynamical systems etc. as part of their continuing professional development (CPD). I recommend the courses offered by the Santa Fe Institute. You can register for most courses at any time, but the agent-based modelling course requires registration and starts at the end of August: https://www.complexityexplorer.org/courses/183-introduction-to-agent-based-modeling
Stephen Hawking famously once said that the 21st century would be the century of complexity. I wholeheartedly agree. IMHO, in these non-linear times, it should be a part of every scientist's (and philosopher's) basic education.
See also article on mounting evidence for negative effect of microplastics - the material from which most bednets are made - now found everywhere in the human body, including in brain tissue:
https://www.theguardian.com/environment/article/2024/aug/21/microplastics-brain-pollution-health
"A growing body of scientific evidence shows that microplastics are accumulating in critical human organs, including the brain, leading researchers to call for more urgent actions to rein in plastic pollution.
Studies have detected tiny shards and specks of plastics in human lungs, placentas, reproductive organs, livers, kidneys, knee and elbow joints, blood vessels and bone marrow.
Given the research findings, “it is now imperative to declare a global emergency” to deal with plastic pollution, said Sedat Gündoğdu, who studies microplastics at Cukurova University in Turkey.
Humans are exposed to microplastics – defined as fragments smaller than 5mm in diameter – and the chemicals used to make plastics from widespread plastic pollution in air, water and even food.
There’s much more plastic in our brains than I ever would have imagined or been comfortable with Matthew Campen, University of New Mexico The health hazards of microplastics within the human body are not yet well-known. Recent studies are just beginning to suggest they could increase the risk of various conditions such as oxidative stress, which can lead to cell damage and inflammation, as well as cardiovascular disease.
Animal studies have also linked microplastics to fertility issues, various cancers, a disrupted endocrine and immune system, and impaired learning and memory.
There are currently no governmental standards for plastic particles in food or water in the United States. The Environmental Protection Agency is working on crafting guidelines for measuring them, and has been giving out grants since 2018 to develop new ways to quickly detect and quantify them."