1 Effective Altruism, Technoscience and the Making of Philanthropic Value Apolline Taillandiera*, Neil Stephensb and Samantha Vanderslottc aLeverhulme Centre for the Future of Intelligence, Department of Politics and International Studies, University of Cambridge, UK and Center for Science and Thought, Universität Bonn, Germany, amit3@cam.ac.uk; bSchool of Social Policy and Society, University of Birmingham, Birmingham, UK, n.stephens@bham.ac.uk; cOxford Vaccine Group, Department of Paediatrics, University of Oxford, Oxford, Oxfordshire, UK and NIHR Oxford Biomedical Research Centre, Oxford, Oxfordshire, UK, samantha.vanderslott@paediatrics.ox.ac.uk *Dr Apolline Taillandier, Leverhulme Centre for the Future of Intelligence, 16 Mill Lane Cambridge CB2 1SB, UK. amit3@cam.ac.uk Abstract: A recent philanthropic movement with advocates among high-profile tech entrepreneurs and philosophers, Effective Altruism (EA) has been widely disparaged for its flawed moral philosophy and conservative political effects. As philanthropic practice, however, it has been seldom studied. In this paper, we explore how EA advocates (EAs) seek to reorient philanthropic priorities through a set of measurement, categorization and ranking tools we label philanthropic metric work. By comparing how EAs portrait their interventions in three domains – neglected tropical diseases, cultured meat, and AI safety –, we show how they produce and challenge philanthropic and field-specific expertise, and how they draw on, and reshape various technoscientific imaginaries and worldmaking ambitions. Contributing to the social study of philanthrocapitalism, our analysis sheds light on the changing nature of philanthropic value creation. Keywords: effective altruism, philanthropy, sociotechnical imaginaries, neglected tropical diseases, cultured meat, AI safety 2 Introduction Little over fifteen years old, effective altruism (EA) aims to maximize ‘the most good’ (MacAskill, 2015) one can do through philanthropy. Its main principle, developed in popular books such as The Most Good You Can Do (Singer, 2015), The Precipice (Ord, 2020), and What We Owe the Future (MacAskill, 2022), is that individuals and non-profits should decide where and how to donate or volunteer based on scientific evidence, to ensure that their money or time contributes to maximising global welfare. In the words of philosopher and advocate Toby Ord, EA is ‘to the pursuit of the good [what] the scientific revolution was to the pursuit of truth’ (Future of Life Institute, 2016). This ambition is not new. Claims about the morality of capitalism have a long history, and the search for ‘efficiency’ in doing good can be seen as a key tenet of modern philanthropy ((Boston Review Forum, 2015; Brest, 2020). Philanthropists have long sought to impose their conception of the good by training political and economic elites (Guilhot, 2007) or actively shaping new fields of expertise (Bartley, 2007). EA organizations in part further this tradition by funding research related to the philanthropic causes they consider to be the most cost-efficient. With high-profile tech entrepreneurs advocating ‘earning to give’ or ‘longtermism’ (Kristof, 2015; Bajekal, 2022; MacAskill, 2022), EA recently received sustained public attention. It has been deemed the new ‘secular religion’ of Silicon Valley elites (Wenar, 2024; Gebru and Torres, 2024). EA, not unlike the Silicon Valley social milieu (Abeles 2003; Turner, 2006; Alexandre, 2022), is characterised by the coexistence of strong social and material hierarchies with a progressive and egalitarian ethos. But rather than a Californian phenomenon, it is a transnational complex with advocates spanning the United States and Europe. Known as ‘EAs’, these include wealthy entrepreneurs such as the recently disgraced Sam Bankman-Fried, high-profile philosophers based at elite universities in the United States, the United Kingdom or Australia, https://www.zotero.org/google-docs/?hhkVv6 https://www.zotero.org/google-docs/?8ss1Eq https://www.zotero.org/google-docs/?gZeZ5t 3 and advocates among graduate students and early career researchers (Rubenstein, 2016; Hyatt, 2022). Typically young, well-educated, and trained in the West, EAs are predominantly aged between 25 and 34, male, white, non-religious, politically on the left or centre left, living in the United States or Britain; 46% are vegan, more than 90% have a higher-education degree, and 20% attend one of the top 20 universities in the world (Dullaghan, 2019; see also Moss, 2020). Organizations committed to EA principles include private foundations – such as San Francisco- based Open Philanthropy and Good Ventures, both set up by Facebook co-founder and billionaire Dustin Moskovitz and former journalist Cari Tuna – and non-profits – such as the charity evaluator GiveWell created by former hedge-fund managers Elie Hassenfeld and Holden Karnofsky, and the Oxford-based Centre for Effective Altruism led by philosophers Toby Ord and William MacAskill. These organizations evaluate, recommend, and fund charities on the basis of number of lives improved per dollar donated: during 2018, GiveWell claimed to have directed more than $140 million to ‘highly effective charities’ through grants and by influencing individual donations (GiveWell, 2018). EA’s public facing initiatives include the website ‘80,000 Hours’ advising on ‘high-impact careers’ and the ‘Giving What We Can Pledge’ campaign to encourage ‘the richest 2% globally’ to donate 10% of their income to effective non- profits (Giving What We Can, 2024). EAs also meet at annual conferences, share houses, organise informal meetups, and interact through online forums and blogs. A growing scholarship explores EA’s ideological and narrative dimensions. Philosophers and political theorists have discussed at length how its ambition to increase global welfare recasts long-standing arguments about the source, nature, and scope of moral obligations (Parfit, 2015; Berkey, 2018; McMahan, 2016; Dietz, 2019; Adams, Crary and Gruen, 2023). They usually trace its broad inspiration in utilitarianism, a strand of moral philosophy concerned with https://www.zotero.org/google-docs/?Ju0meU 4 maximising individual or collective utility or welfare: as political theorist Jennifer Rubenstein put it, ‘the shadow of utilitarianism looms large’ over EA (Rubenstein, 2016, p. 513). Associated with analytic philosophers such as Peter Singer, William MacAskill, Toby Ord and Hilary Greaves, it recently became an object of philosophical inquiry (Greaves and Pummer, 2019; Kumar, 2020). Critics have found EA philanthropy to be mostly conservative: despite its ‘willingness to challenge the status quo’ (Future of Life Institute, 2016), it dilutes philosophical ideas into ‘an empowering investment opportunity’ (Srinivasan, 2015), focuses on individual donations rather than political advocacy and institutional action aiming at structural change (Herzog, 2016; Saunders-Hastings, 2019; Blunt 2022; Lechterman, 2022), and assumes a ‘rescue’ idea of philanthropy with deep depoliticising and paternalist effects (Deveaux, 2021, p. 48-79). Sociologists have described EA as the latest iteration of ‘philanthrocapitalism’ – a mode of philanthropic action privileging donations to for-profits and interventions into domains with high expected returns, manifesting the extension of market-based methods and private funding priorities to traditional spheres of state intervention, such as health and development (Eikenberry and Mirabella, 2017; McGoey and Thiel, 2018). While evidence-based philanthropy is not a new phenomenon, philanthrocapitalism also manifests the incursion of ‘high-net-worth individuals’ from the technology sector in a space traditionally dominated by large foundations (McGoey 2012a, 2015; Depecker, Déplaude, and Larchet, 2018; Haydon, Jung, and Russell, 2021). EA offers a case for studying how giving practices, institutions and rhetoric are transformed by these reconfigurations of the philanthropic field. In this paper, we argue that EA is best understood as a form of technoscientific worldmaking. We show that technoscience constitutes one of the central means through which EAs seek to change the world – from funding technological projects with high expected moral https://www.zotero.org/google-docs/?KUvEfl https://www.zotero.org/google-docs/?KUvEfl https://www.zotero.org/google-docs/?gZeZ5t https://www.zotero.org/google-docs/?aU6rm5 https://www.zotero.org/google-docs/?bylyMn https://www.zotero.org/google-docs/?cSozZQ 5 value to contesting the legitimacy of established knowledge production and aiming to reshape whole fields of expertise. This dimension of EA philanthropy been little analysed, even while EAs have been adamant to support and orient academic research to meet their goals. During 2018, Open Philanthropy recommended over $170 million donations in grants to organizations researching ‘potential risks of advanced AI’, ‘biosecurity and pandemic preparedness’, criminal justice reform, and animal welfare (Open Philanthropy, [2018] 2022), and in 2024 it partnered with the UK Department for Science, Innovation and Technology on a £5 million grants programme to support projects on ‘more effective ways of conducting and supporting research and development’ (UKRI 2024). Our analysis concentrates on three case-studies of EA intervention: the fight against neglected tropical diseases to reduce child mortality in the Global South, the development of cultured meat to reduce animal suffering, and the advancement of artificial intelligence (AI) safety to reduce risks of extinction to future humanity. In each case, we show how EAs translate philosophical views into actionable tools including rankings, forecasts, and assessments of efficiency, and how they seek to maximize moral and economic value through technoscientific innovation. By studying EA in the making, we offer an empirical complement to the ideology critique of philanthrocapitalist logics. This analysis brings together three strands of scholarship. We draw, firstly, on the study of the ‘politics of measurement’ (Lupton, 2013). As sociologists of quantification have extensively argued, both quantitative and systematic qualitative data collection aim to stabilize facts (i.e. some objective account of reality), and they are shaped by and contribute to materializing values and hierarchies (Best, 2012; Bowker and Star, 1999; Desrosières, 1998; Hacking, 1991; Porter, 1997). Rankings, for instance, enable action under uncertainty (Esposito and Stark, 2017). EA philanthropy materializes through what we call philanthropic metric work https://www.zotero.org/google-docs/?ep8v1o https://www.zotero.org/google-docs/?TMVBLx https://www.zotero.org/google-docs/?TMVBLx 6 (on metric work, see Adam, 2016), a set of practices involving the development, codification, and stabilisation of concrete instruments of commensuration, classification, and ranking to assess the effectiveness of charities and donations. Secondly, we expand on the analysis of ‘sociotechnical imaginaries’, the ‘collectively held and performed’ interplay between visions of technoscientific futures and visions of desirable social and moral orders (Jasanoff and Kim, 2015, p. 15). Philanthropists, like other economic actors, produce both an idea of the social and the tools to intervene into it. Our study shows how EAs in different settings perform distinct visions of desirable technology and society. Thirdly, we draw on the economic sociology literatures evidencing the role of technoscientific futures in driving innovation (Borup et al., 2006; Joly, 2010) and securing financial value (Beckert and Bronk, 2018; Muniesa, 2017). As we show across our cases, technoscientific promises, expectations of future technological capability, and claims to expertise play a central role in how EAs define and justify philanthropic investment. Philanthropic value creation, we argue further, is a contested process. A key EA problem has been to determine not only the effectiveness of individual donations, but also to assess which ‘causes’ are the most important, tractable, and neglected, and should therefore be prioritized. A core EA debate aims to delimit and rank areas of intervention and opportunities for investment – through what they call ‘cause prioritization’, EAs typically gauge whether extreme poverty, animal suffering, EA recruitment and organising, or extinction risks related to future technology should be given priority (e.g. Muehlhauser, 2015a). Many EA organizations shifted their focus from global poverty to AI risk in recent years, a turn that critics have analyzed as a departure from the movement’s original commitment to rationality and a sign of unwarranted futuristic speculation (Schuster and Woods, 2021, 45-78; Srinivasan, 2015). Rather than a sign of https://www.zotero.org/google-docs/?aGaTK3 https://www.zotero.org/google-docs/?mRnuju https://www.zotero.org/google-docs/?mRnuju 7 epistemological flaw or inconsistency, EA metric work brings light to the recurring contestations within EA practice, including negotiations between two opposing views of philanthropy: an ideal of decentralised philanthropy premised upon visions of disruptive technological breakthroughs and radical social change, and an ideal of technocratic philanthropy assuming a linear vision of technology-enabled moral progress. EA expertise, rather than neutralizing such contestation, is shaped by conflicting visions of both the means and ends of philanthropy: not a univocal project, EA involves a range of moral views, expert claims, and concrete ways of measuring effectiveness that justify different and partly conflicting investment strategies. The study of EA practice brings to light the conflicted politics of philanthrocapitalism. Methods Our analysis brings together three qualitative studies conducted independently by the three authors. The first study examined the policy history of neglected tropical diseases (NTDs), and the advocacy campaign to raise their profile, through documentary analysis of journal articles, books, media articles, and policy documents, and fifty-five interviews with scientists, policy officials, and NGO workers, from an initial purposeful sample expanded using snowballing technique (2013-2017) (Vanderslott, 2021). The second study focused on cultured meat communities and involved documentary analysis of texts produced by media actors and both cultured meat supporters and detractors, seventy-six interviews with scientists, funders, and proponents of cultured meat, and ethnographic observations (2010-2023) (Stephens 2013, Stephens, N. Sexton, A. & Driessen, C. 2019). The third study traced the political history of transhumanism from the 1970s onwards through published and archival materials and 46 interviews with actors in EA, studies of existential risk studies, and transhumanism in the San 8 Francisco Bay Area and in Oxford, Cambridge, and London (2016-2021) (Taillandier, 2021a). All three authors conducted additional text analysis and interviews for this paper, respectively: in addition to fifty-five interviews in NTDs, five EA focused informal interviews (two) and semi- structured interviews (three); twenty interviews including discussion of EA in cultured meat; and 15 interviews discussing EA or conducted with individuals defining themselves as EAs or affiliated with an EA organization in AI safety. Case-study one: Neglected tropical diseases and the remaking of ‘neglect’ NTDs have been a sustained topic of EA interest. Currently the World Health Organization lists 20 NTDs, including dengue fever, leprosy, and schistosomiasis. The label ‘neglected’ was applied to this disease group in the early 2000s in contrast with the ‘big three’ diseases of HIV/AIDS, tuberculosis, and malaria in order to attract funding from international donors, and it has subsequently received growing policy attention (Vanderslott, 2019). NTDs are clustered together under one title due to their shared status in global health politics rather than a shared biological mechanism. They are diseases of poverty, collectively affecting over one in four people in the world. EAs commonly describe NTDs as a low-hanging fruit in global health, an area they could meaningfully impact through supporting the deployment of existing solutions. As one EA said, ‘I think that, broadly speaking, when EA started, [it] was to a large extent about scaling up existing global poverty interventions, which it was widely believed that we had strong evidence for.’ EA interventions in global health often entail scaling-up low-technology and tried- and-tested interventions, such as bed-nets and deworming pills. EA organizations typically rank causes based on criteria of ‘Importance, Tractability, and Neglectedness’ (ITN). This ITN framework aims to determine the scale of a problem, the 9 possibility of meaningfully addressing it, and the significance of existing philanthropic efforts. In order to assess the cost-effectiveness of research into NTDs, Max Dalton, director of the Centre for Effective Altruism until 2023, conducted quantitative research to ‘assess the impact of an additional dollar given to a research programme, rather than the average impact of all of the dollars that fund a research programme’ (Dalton, 2015, p. 4). Dalton’s work, which emphasized the incremental contribution of the ‘next’ donation and the value of the ITN framework to assess it and was used by Giving What We Can to justify non-profit ranking, is an example of what we call philanthropic metric work. It aimed to identify how to maximize the impact of each individual donation, and assumed that well-informed, strategic donations could make a disproportionately large contribution to global welfare and poverty reduction. Dalton’s assessment illustrates how EAs redefine ‘neglect’ in the context of tropical disease. While NGOs have defined NTDs in contrast with HIV/AIDS, tuberculosis, and malaria, Dalton collated both categories in his calculations: as he explained, although the ‘big three’ are less underfunded than NTDs, they ‘fulfill all of the criteria for NTDs’ (ibid., p. 5), especially as they remain underfunded compared to other diseases. Dalton mobilized an alternative definition of ‘neglect’ to that commonly used by NTD scientists, activists, and policymakers: in opposition to the NTD community that sought to assert difference based upon profile, his definition bound together a broader group of tropical diseases. The same is evident at GiveWell. In 2017, the organization listed both the Malaria Consortium and the Against Malaria Foundation as key charities; the rationale for including malaria being its characterization as a high-mortality disease, the mortality of which can be lessened by low-cost interventions such as anti-malaria nets (GiveWell, 2017). This reinforces the view that ‘neglect’ should be balanced with cost- effectiveness calculations and contributes to making poverty-related diseases commensurable in https://www.zotero.org/google-docs/?3zX58t 10 a way that challenges dominant policy categorisations. By developing their own metrics, EAs contest the monopoly of NTD experts in defining what constitutes neglect. In doing so, they assert their own way of establishing global health priorities. Philanthropic metric work, though, can prove difficult. EAs, while typically prioritizing their own rankings above those commonly in use within the expert communities they interact with, also recognise that their own expertise lacks established credentials in medicine or policymaking. Dalton, for example, studied philosophy, politics, and economics before conducting his study of tropical diseases, and had no prior experience of working on NTDs. Some interviewees questioned the epistemic value of such outsider situation. As one asked, ‘how much can a smart but non-expert person read stuff and come to valid conclusions, and how much do you need actual expertise?’ Another interviewee provided further context: ‘the main advocates for cause areas, none of them have had an academic or professional background in these areas. GiveWell was started by hedge fund people. Giving What We Can was started by philosophers. [AI safety researcher] Eliezer Yudkowsky dropped out of high school. So, everybody is self-taught. And I think people have rightly criticised us for not having a professional background. I think people [in EA] have made up for that to a large degree [emphasis added].’ EA organizations also commonly point out the limits of their own efforts in assessing efficiency. In 2012, Giving What We Can revised its original 2009 estimate of the cost-effectiveness of deworming treatments, one of the leading interventions used by the major NTD charities (Cotton-Barratt, 2012). The revision came after significant errors in the estimates of the Disease Control Priorities Project that informed the original figure were identified by the British charity Cochrane Review, indicating that deworming treatments were less effective than first thought (ibid.). For a period, one of the key charities – Schistosomiasis Control Initiative – https://www.zotero.org/google-docs/?BxsWhj 11 was somewhat ambiguously labelled ‘Top Charities without capacity to use new donations effectively at this time’ (GiveWell, 2019). Following this recategorization, the organization stopped being one of GiveWell’s top-rated charities in 2022 (GiveWell, 2022). Interviewees typically questioned the precision and reliability of their own metrics. One commented on the shifting status of the Schistosomiasis Control Unit and malaria charities: ‘I give most of my donations to Against Malaria, which is still considered the EA charity. And Schistosomiasis Control Initiative. It’s considered, now, a safe bet…’ Discussing ‘longtermism’ – the view that long-term future welfare should weigh significantly more in resource allocation than it currently does – the interviewee explained that donating to charities fighting malaria ‘[is] not the intervention with the highest expected value, but, honestly, with a longtermist view, I don’t know if big, expected value really works, because they are kind of like magic numbers [emphasis added].’ Another noted the difficulty of grounding expected utility measures: ‘I think, basically over time, EAs have started to look at things with more error bars around them and expect that the numbers are less firm than they initially seemed.’ Commenting further on the ‘burden’ of measurement and concerned that expectations of conclusive results are not always met, one interviewee told us: ‘you get these RCTs [randomised controlled trials] done, it costs you so much money and effort. And then it’s like there’s barely any difference with your intervention. You were like, based on other evidence, do you really think it works? And it’s just maddening. And EA is committed to being about evidence. And I just think people have gotten tired of that level of rigour, frankly.’ Acknowledging the limits of metrics is part and parcel of EA’s metric work. In comparison to cultured meat and AI safety, interactions between EAs supporting intervention into NTDs and the scientists conducting research on tropical disease are limited. https://www.zotero.org/google-docs/?3zX58t 12 Alan Fenwick, a professor of tropical parasitology at Imperial College London who founded the Schistosomiasis Control Initiative in 2002, has pointed out the unexpected income stream for donations from the public and high-net-worth individuals since Giving What We Can and GiveWell both listed the Schistosomiasis Control Initiative as a top charity in 2011 (GiveWell, 2011; Fenwick, 2017). In the following years, while he attended EA events to further promote his work, EAs rarely engaged in developing innovations with him or other scientific researchers. Instead, they sought to funnel support to established but underfunded interventions. As one explained, ‘The EA approach isn’t really about development or new treatments or anything. So, it doesn’t really overlap with what most medical or academic people involved with those NTDs would be doing. It’s more like a supply line thing… just picking something where there’s low- hanging fruit. There’s not really a question, how do you treat this? It’s more like, how do you coordinate? … Definitely, EA is more involved with public health people, than in medicine.’ This suggests a logistically driven vision of public health that requires distributing known treatments more effectively or fairly, rather than innovating to produce new ones. In the case of NTDs, EAs identify problems and priorities differently both to the aid and development industry and to academic researchers. In contrast to our following cases studies, EA efforts with NTDs have sought to improve the distribution or allocation of existing resources rather than to lead innovation in science or policymaking. Mobilizing a distinctive definition of neglect premised upon the epistemic superiority of their ‘view from the outside’, EAs also enact a broader measurement ethos aimed at strategic prioritization and cost-efficiency within philanthropy. Assuming the diminishing utility of both financial and knowledge investment, they identify underfunded, high-mortality diseases as a high-value area for philanthropic support to targeted populations (MacAskill, 2015; Todd, 2013). At the same time, they often question the https://www.zotero.org/google-docs/?3zX58t https://www.zotero.org/google-docs/?3zX58t https://www.zotero.org/google-docs/?EqNT3r 13 soundness of their own metrics. Measuring neglect and effectiveness involves multiple re- categorisations, shedding light on the messy, contested character of philanthropic value creation. Case-study two: Cultured meat and the valorisation of uncertain futures Another key area of EA intervention is ‘effective animal advocacy’ (Broad, 2018). Several EA organizations advocate ‘cultured meat’ as a way to minimize animal suffering. Cultured meat refers to the growing of meat in a bioreactor directly from animal cells through biomedical tissue-engineering techniques applied to food (Post, 2012, Abrell 2023), a pursuit involving several university researchers and around a hundred related start-ups since the early 2000s. Several non-profit organizations and grant-giving bodies also advocate cultured meat, most notably New Harvest and the Good Food Institute, both based in the United States. This led to the production in 2013 of the world’s first laboratory-grown burger, as well as small-scale demonstration products such as sausages, meatballs, chicken nuggets, and small steaks (O'Riordan, K. Fotopoulou, A. & Stephens, N. 2016, Stephens, N. & Ruivenkamp, M. 2016). Cultured meat products were first commercialized in limited quantities in late 2020 in Singapore and have been available in the United States since 2023. While challenges remain to successfully commercialize them at volume – including making them cheap and sufficiently stabilized and standardized that they can be consistently produced at scale – the cultured meat project has been sustained by a set of promissory narratives about the moral cost of conventional meat production and animal farming, the value of a Silicon Valley-inspired ‘disruptive’ business model, and the power of technology to realize moral progress (Chiles, 2013; Jönsson, 2016; Sexton et al., 2019; Mouat and Prince, 2018; Stephens, N. Sexton, A. & Driessen, C. 2019). Supporters have argued that cultured meat will cause less environmental damage than conventional meat, improve food https://www.zotero.org/google-docs/?RkfY5m https://www.zotero.org/google-docs/?QTmLzi https://www.zotero.org/google-docs/?QTmLzi 14 safety by avoiding health risks related to animal-borne disease and antibiotic use, and allow the production of meat without killing animals. Consumed in vast quantities, it will cause a food revolution without the need to change the moral views of individuals. Consumers (the dominant framing of publics in this discourse) will spontaneously turn to cultured meat, so long as it is as cheap, accessible, and tasty as conventional products. For our interviewees, the influence of EA within the cultured meat community is clear: EAs have shaped debates about, funnelled additional support for, and directly contributed to producing expertise on cultured meat. Animal Charity Evaluators, an organization originally called ‘Effective Animal Activism’, was founded in 2012 by Eitan Fischer, then a philosophy graduate. Hosted initially by the Centre for Effective Altruism, it became an independent charity in 2013. Defining its mission as ‘to find and promote the most effective ways to help animals’ (Animal Charity Evaluators, 2024), it offers donation advice through online charity reviews, provides advocacy advice, and evaluates charity: Animal Charity Evaluators awarded New Harvest ‘Standout Charity’ status in 2015, and the Good Food Institute ‘Top Charity’ status in 2016 (Animal Charity Evaluators, 2017). Fischer, who went on to work for a plant-based and cultured meat company before co-founding its own, is one of several start-up founders associated with EA. Other EAs look to join companies or third sector groups. One interviewee in 2019 told us that a leading company in the field, Upside Foods, ‘has a ton of people in EA on staff.’ Another suggested that 20-40% of the cultured meat community were EAs. EAs also produce randomised control trials of potential consumer attitudes on cultured meat, with the aim to improve public acceptance (Anderson and Bryant, 2018). As one interviewee claimed, ‘a lot of Effective Altruists fund the NGOs – New Harvest, GFI – so they must make a huge impact on the community.’ By investing attention, time, and money in cultured meat, EAs assert its moral https://www.zotero.org/google-docs/?0cjxDz https://www.zotero.org/google-docs/?IsRRmX https://www.zotero.org/google-docs/?guwh5z 15 value, provide resources for the technology, and contribute to establishing its status as a credible business opportunity and technological project. In comparison to the case of NTDs, assessing the future efficiency of cultured meat requires accommodating high technological and economic risk. EAs, while asserting the need for data-driven analysis in identifying causes to support, also insist that cultured meat is an early- stage technoscientific project with limited material accomplishments. Both Animal Charity Evaluators and EA group Sentience Politics emphasise the risky nature of investment in cultured meat and the partial ignorance resulting from absence of robust empirical data on how it will be produced, what it will cost, or its potential environmental impact. In their report on New Harvest, Animal Charity Evaluators stated: ‘we think that developing cultured animal alternatives is a potentially high-impact way to influence the food system. It seems plausible that demand for conventional meat will decrease once cultured meat is put on the market.’ At the same time, they noted: ‘we believe that our back-of-the-envelope calculation of their cost-effectiveness is too speculative’ (Animal Charity Evaluators, 2017). When responding to a Good Food Institute projection that making cultured meat competitive could take about a decade, they claimed: ‘we suspect that a decade might be somewhat optimistic, and our current estimate is that 10-70 years might be more realistic.’ Similarly, Sentience Politics stated that ‘[a]ssessing the resource efficiency of industrial processes that don’t yet exist involves making many informed assumptions, many of which will later turn out incorrect’ before concluding that ‘cultured meat is expected to be significantly more resource-efficient than animal agriculture, especially when predictions of future meat consumption are taken into account’ (Sentience Politics, 2016b, p. 3). Ultimately, both claims of partial ignorance and the possibility of accelerating returns justify risky investments in cultured meat. 16 EA’s philanthropic metric work is instrumental in demonstrating the efficiency of cultured meat as a means to reduce global suffering. For instance, comparing the need for, and effectiveness of interventions aimed at reducing both human and non-human suffering, Sentience Politics defended cultured meat as part of an aim to expand humanity’s moral consideration to all sentient beings beyond humans (Sentience Politics, 2016). Defining cultured meat as a neglected pathway to addressing animal suffering, it noted: ‘cultured meat research has received very little attention so far, making it relatively easy to conduct basic research that may later prove catalytic to further development’ and ‘[t]his, in combination with its potentially extraordinary return on animal and human welfare in the long term, convinces us that accelerating cultured meat is a worthwhile investment at this time’ (ibid. p. 7). Animal Charity Evaluators considered that ‘the availability of cost-competitive cultured animal products has the potential to cause a massive decrease in the demand for non-cultured animal products’, insisting that estimates of future cultured meat availability ‘meaningfully affect the prioritization of animal advocacy interventions’ (Animal Charity Evaluators, 2018). All the while acknowledging the speculative nature of the technology, EAs do not question the legitimacy of standard risk assessment tools or the heuristic value of the ITN criteria. Investment in cultured meat is also underwritten by judgements about what the good society should look like and who is best placed to bring it into being. It encodes a non-anthropocentric order in which the welfare of all sentient beings is given equal consideration – a view known in moral philosophy as personism (see e.g. Singer, 2011). Made commensurable, interventions aimed at animal and at human welfare can be assessed and ranked on one common scale. EA has received ambivalent responses from engineers and entrepreneurs involved in cultured meat. One scientist explained: ‘I’m very uncomfortable with the amount of dogmatic https://www.zotero.org/google-docs/?7ukaYX 17 thinking in EA… there’s a lack of humility… but I think that EA has brought in a lot of dollars, which has been good.’ A start-up founder also raised concerns that ‘the way [EAs] do moral calculus is not just immoral and shady, but also uneducated and not well thought out’. Questioning the credentials and expertise of EAs, he noted: ‘this ends up with [EAs] who are like, “I’m going to form a company.” And it’s like, “what are your qualifications?” ... they have functioned in an environment which encourages participation, but it doesn’t require content.’ Another interviewee, explaining how they lost confidence in EA practice, made a similar critique: ‘the problem with Effective Altruism in general is people sitting in a room and deciding unilaterally what’s the best use of funding money, when they’re not experts in the space… It’s extremely frustrating because I often see these individuals who… love thinking they know a whole lot about things that they’re really not that informed on, and making some little, tiny equation that supposedly justifies how good their spending is.’ These critics see EAs as ‘dogmatic,’ ‘uneducated,’ and ‘not experts’: influential but self-selected figures, whose lack of technical training can lead to poor investment decisions and potentially threaten the credibility of the cultured meat project. The cultured meat case shows how EAs assess the future profitability of cultured meat through its moral value, derived from its presumed efficiency to reduce animal suffering. Underlying EA metric work in this case is a commitment to the idea of commensurability between human and non-human suffering. EAs also mobilize technoscientific promises that are common in venture capital valorization or funding startups to advance research and development on cultured meat products. As with the NTDs case, entrepreneurs and scientists involved in cultured meat development contest EA interventions for their lack of relevant expertise. EAs themselves, meanwhile, have produced ‘back-of-the-envelope calculations’ of the future moral 18 and financial benefits that cultured meat could generate, seeking to establish at once EA expertise and cultured meat as valuable philanthropic opportunity. Case-study 3: AI safety and the securing of long-term value AI safety, the aim to mitigate potential risks from advanced AI, is another key EA cause. A matter of ‘extremely high importance, high neglectedness, and reasonable tractability’ according to EA organizations GiveWell and Open Philanthropy (Karnofsky 2016b), AI safety involves theoretical mathematical models of hypothetical AI general or superhuman capabilities (Legg, 2008), machine learning techniques to ‘align’ the behaviour of AI systems with users’ preferences (Amodei et al., 2016), and global policy agendas for governing ‘frontier AI’ (UK Department for Science, Innovation and Technology, 2023). Prominent AI scientists, such as Stuart Russell, a computer science professor at the University of California, Berkeley, have described it as the incorporation of basic engineering considerations to their discipline (Russell, 2019). More controversial claims have pointed to the ‘existential risks’ to humanity’s survival over the long-term that future AI could pose (Bostrom, 2014; Hawking et al., 2014; Altman, Brockman, Sutskever, 2023). AI safety research has received substantial funding and public support in recent years, including through Elon Musk donating $10 million to fund ‘long-term research’ on AI risks (Russell, Dewey and Tegmark, 2015), technology companies DeepMind and OpenAI pledging to ensure that ‘artificial general intelligence (AGI)… benefits ‘all of humanity’ (OpenAI, 2018), and the UK government launching in 2023 the Frontier AI Taskforce, a ‘start-up inside government’ to build ‘state capacity’ in AI safety (Frontier AI Taskforce, 2023). https://www.zotero.org/google-docs/?4wjke3 https://www.zotero.org/google-docs/?4wjke3 https://www.zotero.org/google-docs/?TG0FH2 19 In contrast with NTDs and cultured meat, the main challenge of AI safety, as defined by EA organizations, has been to produce relevant academic expertise (Todd, 2015). The ambition to develop academic subjects relevant to AI safety is evident in Open Philanthropy’s grant- making strategy: between 2016 and 2023, the organization directed more than $270 million to AI research with a ‘safety’ focus, including $30 million to the then non-profit OpenAI, $211 million to the Center for Human-Compatible AI (CHAI) at UC Berkeley, and $14 million to the Machine Intelligence Research Institute (MIRI), a transhumanist organization launched in 2000 to prepare for the advent of a technological singularity (Yudkowsky, 2001; see Taillandier, 2021b). Interviewees generally insisted that AI safety was a contested research area. A graduate student in computer science described it as a topic that one would choose for moral, rather than scientific reasons: ‘for a while I didn’t agree with [EA] arguments about AI safety… because they seemed to conflict with what I knew about AI and machine learning.’ As he explained, ‘what convinced me to work on AI safety was thinking more about population ethics,’ which led him to reassess the moral value of future people’s welfare and reconsider his own career strategy on this basis. A computer science professor similarly described in moral terms his decision to shy away from a ‘fringe’ research topic: ‘maybe if I were a better person, I would be spending more time researching AI safety … [but] if I did, I would stop getting funding, and I wouldn’t be able to support that many students.’ For those settling on AI safety as priority cause, the expected moral value of AI safety outweighed its material and symbolic costs. EAs typically assess the importance and legitimacy of AI safety as philanthropic cause and research through invoking their intuition about its high expected moral value. In 2011, Holden Karnofsky, GiveWell co-founder and CEO of the Open Philanthropy Project, explained why he would not recommend offering a grant to support MIRI even though he agreed ‘with https://www.zotero.org/google-docs/?Ba94FJ 20 many of [its] most controversial views’ – notably ‘that [unfriendly AI] is an existential risk and that [Friendly AI], if it could be created before [unfriendly AI], would be astronomically beneficial’ (Karnofsky, 2015). As in the NTD case, Karnofsky based his decision on an ‘outsider’ judgment about the institute’s capacity to contribute to the field: ‘my intuitions don’t matter much because I know so little about artificial intelligence research and other relevant issues; but when I look at the actions of the people who seem to me that they ought to know better… few explicitly endorse your mission and capability to follow through on it… [M]y view of your organization hinges heavily on the question of whether you have people who have rare insights, capabilities, and general rationality, and if in fact you have those things, you ought to be able to translate them into impressive achievements/affiliations.’ In 2015, Open Philanthropy revised its decision about MIRI. It recommended AI safety as priority area, following a ’shallow investigation’ of AI risks consisting of informal discussions with researchers and investors who had publicly raised the alarm about global catastrophic risks (Open Philanthropy, 2015). Karnofsky explained how an Open Letter by the Future of Life Institute and the publication of Nick Bostrom’s Superintelligence (2014b) had led him to the ‘new interpretation’ that ‘there simply is no mainstream academic or other field (as of today) that can be considered to be “the locus of relevant expertise” regarding potential risks from advanced AI… There is no one who I think clearly qualifies as an expert on such matters, and no one I’m aware of who has clearly put more thought into the relevant issues than Nick Bostrom or relevant staff at MIRI’ (Karnofsky, 2016c, p. 5, emphasis in original). Having previously ‘misjudged’ the importance of AI risks, he hoped that ‘next time a cause this potentially promising emerges’, he would ‘recognize it faster’ (Karnofsky, 2016c, p. 6). Others within the organization noted that neither expert surveys nor quantitative extrapolations from current progress constituted reliable sources 21 to anticipate future breakthroughs in AI (Muehlhauser, 2015b. As in the cultured meat case, EAs primarily drew on entrepreneurial intuition to assess the urgency and importance of AI safety for a ‘hits-based’ approach to philanthropy (Karnofsky, 2016a). AI safety remains largely debated within EA. Some EA philosophers advocated prioritizing AI safety on the basis of longtermism, the idea that ‘the long-term future of humanity matters more than anything else’ (Wiblin, 2017). Longtermism, they argued, required adjusting EA metrics, rather than departing from evidence-based philanthropy or a more general commitment to ‘the scientific method’ (MacAskill, 2019, p. 15; MacAskill, 2022; Ord, 2020) i. Longtermism was also, as William MacAskill insisted, ‘is compatible with any empirical view about the best way of improving the long-run future’ (MacAskill, 2019). Others, for instance Singer, expressed some reservation on EA prioritizing existential risk mitigation over the fight against poverty (Singer, 2015, 165-174). As several interviewees noted, whether AI safety should be prioritized also revealed tensions between ‘the elite of the movement [Ord, MacAskill, Karnofsky]’ and activists. A researcher at Open Philanthropy noted: ‘there is a little bit of a disconnect with the leadership of EA, having shifted in that direction and being more interested in [existential] risks… I think there is some tension around that, and maybe that feels out of touch with the rest of the EA community.’ One interviewee, then a fellow of the Oxford-based Future of Humanity Institute led by transhumanist philosopher Nick Bostrom, explained how longtermism had played out in his own view: ‘[EA] was initially very much like, global health is this area where we’ve got… roughly crisp measures of what we can work on in terms of costs to save a life, so we can rank everything, we can have these top options like malaria or deworming’, but longtermism ‘[weighed] against the malaria and deworming agenda’. At the same time, he acknowledged the speculative nature of the argument, hoping that ‘it’s not the fad of a weird https://www.zotero.org/google-docs/?4EzdtN https://www.zotero.org/google-docs/?uHAPRk https://www.zotero.org/google-docs/?uHAPRk https://www.zotero.org/google-docs/?EFvmH9 https://www.zotero.org/google-docs/?v85v2B 22 longtermist cult, and we are just following the evidence wherever it leads us.’ As in our previous cases, EAs acknowledge the epistemic difficulty of producing evidence. This latter case further illustrates how EA metric work translates moral valuation of future welfare into funding decisions. As in the case of cultured meat, EAs typically acknowledged the speculative character of long-term predictions and the limits of existing expertise, including their own. Instead of strong ‘tractability’, they found the decisive criterion for investing in AI safety to be its potential moral ‘high reward’. EA support to AI safety includes research at leading academic institutions such as Berkeley, as well as support for transhumanist academic and independent researchers selected for their long-time involvement in the study of existential risk – making EA, as a former researcher at the Oxford Future of Humanity Institute noted, ‘the continuation of transhumanism in a much wider sense’.ii In line with computer scientists stating that social responsibility should be part of the very definition of AI, EA organizations have sought to lead innovation in AI research and regulation, as well as to reshape the field as a whole. Discussion: Philanthropic metric work as technoscientific worldmaking One of the most prominent forms of philanthrocapitalism today, EA philanthropy promises to improve the world by optimising individual donations, career strategies, venture capital investment, and research grant allocation. EA translates philosophical principles – for instance, the notion that the wellbeing of each person should be given equal consideration regardless of space, time, or species – into a variety of ranking or ordering tools. EA’s philanthropic metric work involves a range of quantitative methods – such as RCTs, cost-benefit analysis, technology forecasting, market predictions, and the application of what EAs called the ITN framework – 23 andqualitative ones – such as interviews with AI experts and historical comparison to assess ‘highly uncertain’ but ‘plausible’ future AI risks (Open Philanthropy, 2015; Muehlhauser, 2015b). Metric work stabilizes definitions of well-being and need – in the fight against extreme poverty, EA interventions typically prioritize access to deworming treatments over public health infrastructure development – and establishes commensurability across sites of intervention, allowing actions aimed at human and actions aimed at animal welfare to be ranked on the same scale – a process that Marion Fourcade has termed ‘ordinalization’ (Fourcade, 2016). EA metrics also reduce the uncertainty associated with contested actions, for instance when calculations about the extremely large potential value of AI safety for future generations overweigh the difficulty of assessing the long-term tractability of both AI risks and mitigation strategies. Acknowledging epistemic limitations and ignorance – both theirs and those of traditional experts – EAs produce evidence of expected value not in spite of moral and economic uncertainty, but on the basis of it. EA philanthropy, in this sense, illustrates the well-studied role of risk mitigation strategies and promissory narratives in technoscientific innovation (Borup et al. 2006; Beckert and Bronk, 2018). Through producing criteria and evidence of importance, neglect, and effectiveness and mitigating the uncertainty of risky investments, metric work constructs the world that EAs seek to change, as well as the means to do so. EA philanthropy entails a familiar ‘economics of techno-scientific promises’ (Joly, 2010), whereby the moral imperative to improve human and non-human welfare justifies innovation. But where expectations of moral progress and profitability are often found to take part in the construction of technological inevitability, our cases evidence that EA metric work often fails to materialize objectivity and indisputability or to stabilize the horizon of a linear future. EAs hold different views on whose wellbeing matters the most and on the most appropriate ways to 24 improve present and future global welfare. These contrasting views are especially evident in debates around cause prioritization, with some defending the ‘safe bet’ of supporting NTDs and others advocating AI safety as a ‘risk worth taking’ (Karnofsky, 2016a). This diversity of views, far from unacknowledged, is a core tenet of EA’s rhetoric and practice – as reflected in Open Philanthropy’s commitment to ‘worldview diversification’ (Karnofsky, 2016d). Changes in EA priorities, rather than a straightforward shift away from the ‘low-hanging fruit’ of anti-malaria bednets to the ‘high-risk, high-reward’ projects of cultured meat and AI safety or a logical consequence of its philosophical commitments – for instance the broadening of moral concern from humans regardless of their location in space, to non-human animals, and finally to future people, human or non-humaniii – is best understood as the result of struggles between conflicting moral and economic expectations. Our cases also draw a more nuanced picture of philanthrocapitalism to the one that critics have commonly portrayed. EA, while contributing to what McGoey and Thiel have called the ‘sanctification of the super-rich’ (2018), also effectively turns average-income donors into philanthropic experts in their own right. As the cases of cultured meat and AI safety make clear, EA draws upon and enacts entrepreneurial imaginaries that endow investors with special foresight and intuition – reiterating the mainstream liberal assumption that entrepreneurs and investors can foresee future market states and stir progress through risk-taking (Boenig-Liptsin and Hurlbut, 2016). More generally, EA metrics, based on the apparently rigid and decontextualized criteria of importance, tractability and neglectedness, in fact largely accommodate the subjective preferences, intuitions, and value judgements of ‘self-taught’ individuals. In part through a critique of traditional knowledge authorities and through performing ‘strategic ignorance’ (McGoey, 2012b), EAs assert their authority to steer innovation 25 and moral progress: self-appointed experts, they insist, are more likely to make morally valuable choices than academics. Further elements of EA practice – such as its grassroots component, self-description as a social movement, and rhetoric of (partial) privilege acknowledgment – reflect its ambition to democratise philanthropy and achieve social change through investments and donations unimpeded by traditional political oversight or corporate governance. In line with ‘alternative philanthropy’ (Silver, 2007) and ‘philanthrolocalism’ (Beer, 2015), EAs reclaim the right for every potential donor in the two percent global top-earners to define and address world problems in their own terms. More than an instrument or a tool for enforcing a unified, consistent view of the good, technoscientific production is an integral part of EA’s attempt to make philanthropic expertise accessible to every donor. EAs challenge the monopoly of traditional experts and professional communities over the definition of what constitutes optimal investment and qualified philanthropic knowledge, either through tailoring existing objectifying tools to their needs, or through developing their own expertise. In the NTD case, they identify global health problems and priorities in their own way, contesting the categories in use in global development, public health, and biology research. In cultured meat and AI safety, by contrast, they seek more direct involvement in the making of technoscience, either through founding cultured meat startups or through seeking to make a ‘foundational contribution’ to AI and machine learning (Whittlestone, 2015). The continuing relevance of traditional knowledge hierarchies, credentials, and ambitions not only constraints EA claims to expertise, but it also puts crucial limits to their stated ambition to radically transform and democratise philanthropy through the aggregation of individual values. Ultimately, EA’s imaginary of social change through the rule of experts challenges the 26 vanguard, entrepreneurial model of philanthropy led by visionary and benevolent technoscientists. Conclusion: Philanthropic metric work as technoscientific worldmaking Philanthropy performs a complex ideological function. It materializes the moral superiority of economic elites, hides the role of donations in tax avoidance, increases global inequalities, and reinforces the view that capitalism can benefit the greater number. Endorsed by prominent technology billionaires and philosophers in recent years, effective philanthropy has become increasingly visible. While most scholarship on EA discusses its utilitarian commitments or reiterates long-standing critiques of philanthropy more generally, our paper brings emphasis to the material practices through which EAs translate moral claims into action. Through comparing three case-studies – NTDs, cultured meat, and AI safety – we fleshed out some of the associated ambiguities, contestations, identity-work, and localized interactions that take part in EA philanthropy. By studying EA in the making, we provide empirical support to the claim that it contributes to fostering existing configurations of social and market power as privileged ways to social change. We showed that philanthrocapitalism is a complicated ideological project, one that entails egalitarian aspirations and decentralized knowledge practices alongside justifications of the socioeconomic status quo. A heterogeneous set of practices, EA’s philanthropic metric work entails measuring how a singular donation increases the welfare of a particular category of benefactors (the global poor, non-humans, or future people), recasting what counts as neglected in global health policy, or incorporating entrepreneurial intuition into risk-benefit analysis. It is also an operation in sociotechnical worldmaking, whereby the tools to assess social change are 27 themselves a particular modality of philanthropy. By developing its own risk assessments, categorizations and rankings, EA de facto contests dominant forms of philanthropic and academic expertise. Empowered with rationalization techniques such as the ITN framework, each donor can decide not only whether and where to give, but also which areas of philanthropic intervention are legitimate and which are not. Well aligned with other forms of philanthrocapitalism, EA enacts dominant entrepreneurial imaginaries, whereby gifted individuals are seen as best positioned to lead social change, knowledge production, and technological innovation. But EA also furthers a more traditional view of philanthropy as production of expertise. Ultimately seeking to reshape technoscientific domains by remaking academic fields from within, as in the case of AI safety and computer science, EA philanthropy entails both challenging traditional philanthropic expertise and reiterating its high-modernist ambition to change the world from above. By shifting the focus from metrics to metric work, we showed that philanthropic instruments express conflicting moral views. Metrics, in our cases, act not only as neutralising tools that enable philanthropic action under uncertainty but also as highly politicized knowledge interventions based on different understandings on what constitutes the world’s most pressing problems and the most efficient ways to address them. Seeking to maximize the value of philanthropic actions through supporting, reshaping, or fostering technoscientific projects with high expected market and moral value, EAs remake promissory narratives and channel resources towards controversial projects – be they upscaling bednet distribution as a way to address NTDs, or embracing the financial promises of the ‘AI fetish’ (see Kampmann, 2024). Effectively blurring the boundary between philanthropic and technoscientific worldmaking, the study of EA 28 sheds light on the changing repertoires of philanthropic value creation, and the ambitions of the technology elite to remake the world in their own image. Acknowledgements We would like to thank the interviewees in our case studies for their contribution, as well as Benjamin Bourcier, Nicolas Duvoux, the participants to the research axis ‘Mobilisations philanthropiques’ at the 2022 Meeting of the Association Francaise de Science Politique, the editors to this journal and our three anonymous reviewers for comments on the paper. Funding details Apolline Taillandier’s work was supported by a doctoral fellowship from the Max Planck Sciences Po Center, a postdoctoral fellowship from the Thomas und Ulla Kolbeck-Stiftung, and funding from the Mercator Stiftung. Neil Stephens' work was supported by the Wellcome Trust under Grant WT208198/Z/17/Z and the EPSRC under Grant EP/X038114/1. Samantha Vanderslott’s work was supported by the ESRC under Grant ES/J500185/1. Disclosure statement Apolline Taillandier is postdoctoral research fellow at the Leverhulme Centre for the Future of Intelligence, Newnham College, and the Department of Politics and International Studies at the University of Cambridge, jointly with the Center for Science and Thought at the University of Bonn. Her work focuses on the political theory and history of transhumanism and future imaginaries of artificial intelligence. Neil Stephens is Associate Professor of Technology and Society at the University of Birmingham. He is a sociologist and Science and Technology Studies scholar focused upon innovation in biotechnology, with themes including cultured meat, mitochondrial donation, and regenerative medicine. Samantha Vanderslott leads the Vaccines and Society Unit (VAS), hosted by the Oxford Vaccine Group at the University of Oxford. She is a health sociologist and associate professor working on topics at the 29 intersection of health and society. Her research currently focuses on public attitudes and decisions on vaccination, particularly in relation to pro-vaccination behaviours and vaccine acceptance. Ethical approval statement The paper reports research from three projects, which each received their own ethical approval. Apolline Taillandier's work was approved by Sciences Po Doctoral School and the Max Planck Sciences Po Center on Coping with Instability in Market Societies. Neil Stephens' work was approved by Brunel University London College of Business, Arts and Social Sciences Research Ethics Committee (10699-MHR- May/2018- 12767-1), and Samantha Vanderslott's work was approved by University College London (UCL) Research Committee and the University of Oxford Central University Research Committee (R50658/RE001). References Abrell, E. (2023). The empty promises of cultured meat. In C. J. Adams, A. Crary & L. Gruen (eds.), The good it promises, the harm it does: Critical essays on effective altruism (pp. 149–159). Adams, V. (2016). Metrics: What counts in global health. Duke University Press. Adams, C. J., Crary A. & Gruen L. (eds.). (2023). The good it promises, the harm it does: Critical essays on effective altruism. Oxford University Press. Alexandre, O. (2022). Une normalité ostentatoire: Ethos égalitaire et hiérarchies sociales dans la Silicon Valley. Politix 4(140), 47-76. Altman, S., Brockman, G. & Sutskever, I. (2024). Governance of superintelligence [Blogpost]. Retrieved from https://web.archive.org/web/20240503045637/https://openai.com/index/governance-of- superintelligence/. Amodei, D., Olah, C., Steinhardt, J., et al. (2016). Concrete problems in AI safety. Retrieved from https://arxiv.org/abs/1606.06565v1. https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 30 Anderson, J. & Bryant, C. (2018). Naturalness concerns and clean meat acceptance: A faunalytics study. Wellbeing international studies repository report. Retrieved from https://web.archive.org/web/20210422111435/https://www.wellbeingintlstudiesrepositor y.org/cgi/viewcontent.cgi?article=1001&context=hw_diet. Animal Charity Evaluators (2017). New Harvest. Retrieved from https://web.archive.org/web/20180420203014/https://animalcharityevaluators.org/charity -review/new-harvest/. Animal Charity Evaluators (2018). Cause area priorities for animal advocacy. In Animal Charity Evaluators. Retrieved from https://web.archive.org/web/20181116212611/https://animalcharityevaluators.org/blog/ac e-highlight-when-will-there-be-cost-competitive-cultured-animal-products/. Animal Charity Evaluators (2024). Who we are. Retrieved from https://web.archive.org/web/20240519121146/https://animalcharityevaluators.org/about/ who-we-are/. Bajekal, N. (2022, August 10). Want to do more good? This movement might have the answer. Time. Retrieved from https://time.com/6204627/effective-altruism-longtermism-william- macaskill-interview/. Bartley, T. (2007). How foundations shape social movements: The construction of an organizational field and the rise of forest certification. Social Problems 54(3), 229–55. Beckert, J. & Bronk, R. (eds.). (2018). Uncertain futures: Imaginaries, narratives, and calculation in the economy. Oxford University Press. Beckstead, N. (2013). On the overwhelming importance of shaping the far future (Unpublished doctoral dissertation). Retrieved from https://rucore.libraries.rutgers.edu/rutgers- lib/40469/PDF/1/play/. Beer, J. (2015). The philanthropic revolution: An alternative history of American charity. University of Pennsylvania Press. Best, R. K. (2012). Disease politics and medical research funding: Three ways advocacy shapes policy. American Sociological Review 77(5), 780–803. Boenig-Liptsin, M. & Hurlbut, J. B. (2016). Technologies of transcendence at Singularity University. In J.B. Hurlbut & H. Tirosh-Samuelson (eds.), Perfecting human futures: Transhuman visions and technological imaginations (pp. 239–68). Springer. https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 31 Borup, M., Brown, N., Konrad, K. & Van Lente, H. (2006). The sociology of expectations in science and technology. Technology Analysis & Strategic Management 18(3–4), 285–98. Boston Review Forum (2015). The logic of effective altruism. Boston Review. Retrieved from https://bostonreview.net/forum/logic-effective-altruism/angus-deaton-response-effective- altruism. Bostrom, N. (2002). Existential risks: Analyzing human extinction scenarios and related hazards. Journal of Evolution and Technology 9(1): 1–30. Bostrom, N. (2003). Astronomical waste: The opportunity cost of delayed technological development. Utilitas 15(3), 308–314. Bostrom, N. (2013). Existential risk prevention as global priority. Global Policy 4(1), 15–31. Bostrom, N. (2014a). Crucial considerations and wise philanthropy. EA Forum. Retrieved from https://www.effectivealtruism.org/articles/crucial-considerations-and-wise-philanthropy- nick-bostrom. Bostrom, N. (2014b). Superintelligence: Paths, dangers, strategies. Oxford University Press. Bowker, G. C. & Star, S. L. (1999). Sorting things out: Classification and its consequences. MIT Press. Blunt, G. D. (2022). The Gates Foundation, global health and domination: A republican critique of transnational philanthropy. International Affairs 98(6), 2039–2056. Brest, P. (2020). The outcomes movement in philanthropy and the nonprofit sector. In W. W. Powell & P. Bromley (eds.), The nonprofit sector: A research handbook (3rd ed.; pp. 381–412). Stanford University Press. Chiles, R. M. (2013). Intertwined ambiguities: Meat, in vitro meat, and the ideological construction of the marketplace. Journal of Consumer Behaviour 12(6), 472–482. Cotton-Barratt, C. (2012). Neglected tropical diseases: Are they cost-effective to treat? Retrieved from https://www.givingwhatwecan.org/post/2012/10/neglected-tropical- disEAses-are-they-cost-effective-to-treat/. Crary, A. (2023). Against ‘effective altruism’. In C.J. Adams, A. Crary & L. Gruen (eds.), The good it promises, the harm it does (pp. 225–247). Dalton, M. (2015). Giving What We Can: Medical research part 1. Retrieved from https://www.givingwhatwecan.org/report/medical-research-1/. https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 32 Davidson, J. P. L. (2022). Extinctiopolitics: Existential risk studies, the extinctiopolitical unconscious, and the billionaires’ exodus from Earth. New Formations 107–108(2022), 48–66. Depecker, T., Déplaude, M.O. & Larchet, N. (2018). La philanthropie comme investissement. Contribution à l’étude des stratégies de reproduction et de légitimation des élites économiques’. Politix 121(1), 9–27. Desrosières, A. (2002). The politics of large numbers: A history of statistical reasoning. Harvard University Press. Deveaux, M. (2021). Poverty, solidarity, and poor-led social movements. Oxford University Press. Dullaghan, N. (2019). EA survey 2019: Community demographics & characteristics. Retrieved from https://web.archive.org/web/20211122233347/https://rethinkpriorities.org/publications/E As2019-community-demographics-characteristics. Eikenberry, A. M. & Mirabella, R. M. (2017). Extreme philanthropy: Philanthrocapitalism, effective altruism, and the discourse of neoliberalism. PS: Political Science & Politics 51(1), 43–47. Esposito, E. & Stark, D. (2019). What’s observed in a rating? Rankings as orientation in the face of uncertainty. Theory, Culture & Society 36(4), 3–26. Fenwick, A. (2017). Schistosomiasis research and control since the retirement of Sir Patrick Manson in 1914. Transactions of The Royal Society of Tropical Medicine and Hygiene 111(5), 191–198. Fourcade, M. (2016). Ordinalization: Lewis A. Coser memorial award for theoretical agenda setting 2014. Sociological Theory 34(3), 175–95. Frontier AI Taskforce (2023). Frontier AI taskforce: First progress report. Retrieved from https://web.archive.org/web/20230907134432/https://www.gov.uk/government/publicatio ns/frontier-ai-taskforce-first-progress-report/frontier-ai-taskforce-first-progress-report. Future of Life Institute (2016). Effective altruism 2016. Retrieved from https://web.archive.org/web/20230329041044/https://futureoflife.org/event/effective- altruism-2016/. https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 33 Gebru, T. & Torres, É. P. (2024). The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence. First Monday. Retrieved from https://firstmonday.org/ojs/index.php/fm/article/view/13636. GiveWell (2011). Schistosomiasis Control Initiative: November 2011 review. Retrieved from https://web.archive.org/web/20200526202928/https://www.givewell.org/charities/schisto somiasis-control-initiative/November-2011-review. GiveWell (2018). GiveWell metrics report: 2018 annual review. Retrieved from https://web.archive.org/web/20201028032223/https://files.givewell.org/files/metrics/Give Well_Metrics_Report_2018.pdf. GiveWell (2019). Allocation of discretionary funds from Q4 2019. Retrieved from https://blog.givewell.org/2020/03/17/allocation-of-discretionary-funds-from-q4-2019/. GiveWell (2022). SCI Foundation: November 2021 version. Retrieved from https://web.archive.org/web/20221121153635/https://www.givewell.org/charities/sci- foundation/November-2021-version#Doesitwork. Giving What We Can (2024). The Giving What We Can pledge. Retrieved from https://web.archive.org/web/20240208030504/https://www.givingwhatwecan.org/pledge. Guilhot, N. (2007). Reforming the world: George Soros, global capitalism and the philanthropic management of the social sciences. Critical Sociology 33(3), 447–77. Greaves, H. & Pummer, T. (eds.). (2019). Effective altruism: Philosophical issues. Oxford University Press. Hacking, I. (1991). The Making and molding of child abuse. Critical Inquiry 17(2), 253–288. Hawking, S., Tegmark, M., Russell, S., & Wilczek, F. (2014, April 19). Transcending complacency on superintelligent machines. Huffington Post. Retrieved from https://www.huffingtonpost.com/stephen-hawking/artificial- intelligence_b_5174265.html. Haydon, S., Jung, T., & Russell, S. (2021). ‘You've been framed’: A critical review of academic discourse on philanthrocapitalism. International Journal of Management Reviews, 23(3), 353–375. Herzog, L. (2016). Can ‘effective altruism’ really change the world? Open Democracy (blog). Retrieved from https://www.opendemocracy.net/en/transformation/can-effective- altruism-really-change-world/. https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 34 Hutchinson, M. (2015). Giving What We Can. In R. Carey (ed.), The effective altruism handbook (pp. 114–116). Retrieved from https://web.archive.org/web/20150701172316/http://www.careyryan.com/files/EA_Hand book.pdf. Hyatt, J. (2022, November 17). Sam Bankman-Fried’s donations to effective altruism nonprofits tied to an Oxford professor are at risk of being clawed back. Forbes. Retrieved from https://www.forbes.com/sites/johnhyatt/2022/11/17/disgraced-crypto-trader-sam- bankman-fried-was-a-big-backer-of-effective-altruism-now-that-movement-has-a-big- black-eye/?sh=571bd4554ce7. Jasanoff, S. & Kim, S. H. (eds.). (2015). Dreamscapes of modernity: Sociotechnical imaginaries and the fabrication of power. University of Chicago Press. Joly, P. B. (2010). ‘On the economics of techno-scientific promises’ In M. Akrich, Y. Barthe, F. Muniesa & P. Mustar (eds.), Débordements: Mélanges offerts à Michel Callon, Presse des Mines (pp. 203–222). Jönsson, E. (2016). Benevolent technotopias and hitherto unimaginable meats: Tracing the promises of in vitro meat. Social Studies of Science 46(5), 725–748. Kampmann, D. (2024). Venture capital, the fetish of artificial intelligence, and the contradictions of making intangible assets. Economy and Society 53 (1), 39–66. Karnofsky, H. (2015, February 1). Jasen Murray and others, Singularity Institute for Artificial Intelligence [Notes from an interview conducted on GiveWell’s behalf, dated 2011, March 2]. Retrieved from https://web.archive.org/web/20220815124048/https://www.openphilanthropy.org/researc h/jasen-murray-and-others-singularity-institute-for-artificial-intelligence/. Karnofsky, H. (2016a). Hits-based giving. Retrieved from https://web.archive.org/web/20220621171027/https://www.openphilanthropy.org/researc h/hits-based-giving/. Karnofsky, H. (2016b). Potential risks from advanced artificial intelligence: The philanthropic opportunity. Retrieved from https://web.archive.org/web/20220610124101/https://www.openphilanthropy.org/researc h/potential-risks-from-advanced-artificial-intelligence-the-philanthropic-opportunity/. https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 35 Karnofsky, H. (2016c). Some key ways in which I’ve changed my mind over the last several years [Personal notes]. Retrieved from https://web.archive.org/web/20190506231003/https://docs.google.com/document/d/1hKZ NRSLm7zubKZmfA7vsXvkIofprQLGUoW43CYXPRrk/edit. Karnofsky, H. (2016d). Worldview diversification, Retrieved from https://web.archive.org/web/20220701180603/https://www.openphilanthropy.org/researc h/worldview-diversification/. Kumar, V. (2020). The science of effective altruism. In D. W. Portmore (ed.), The Oxford handbook of consequentialism. Oxford University Press (pp. 530–547). Lechterman, T. (2022). The tyranny of generosity: Why philanthropy corrupts our politics and how we can fix it. Oxford University Press. Legg, S. (2008). Machine super intelligence (Unpublished doctoral dissertation). Retrieved from https://web.archive.org/web/20110719142043/http://www.vetta.org/documents/Machine_ Super_Intelligence.pdf. Lupton, D. (2013). Quantifying the body: Monitoring and measuring health in the age of health technologies. Critical Public Health 23(4), 393–403. MacAskill, W. (2015). Doing good better: How effective altruism can help you make a difference. Penguin. MacAskill, W. (2019). Longtermism. Retrieved from https://web.archive.org/web/20200401063437/https://forum.effectivealtruism.org/posts/q ZyshHCNkjs3TvSem/longtermism. MacAskill, W. (2022). What we owe the future: A million-year view. Oneworld. McGoey, L. (2012a). Philanthrocapitalism and its critics. Poetics 40(2), 185–199. McGoey, L. (2012b). The logic of strategic ignorance. The British Journal of Sociology 63(3), 553–576. McGoey, L. (2015). No such thing as a free gift: The Gates Foundation and the price of philanthropy. Verso. McGoey, L. & Thiel, D. (2018). Charismatic violence and the sanctification of the super-rich. Economy and Society 47(1), 111–134. https://web.archive.org/web/20190506231003/https://docs.google.com/document/d/1hKZNRSLm7zubKZmfA7vsXvkIofprQLGUoW43CYXPRrk/edit https://web.archive.org/web/20190506231003/https://docs.google.com/document/d/1hKZNRSLm7zubKZmfA7vsXvkIofprQLGUoW43CYXPRrk/edit https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 36 Moss, D. (2020). EA survey 2020: Demographics. Retrieved from https://web.archive.org/web/20210512090834/https://forum.effectivealtruism.org/posts/T hdR8FzcfA8wckTJi/ea-survey-2020-demographics. Mouat, M.J. & Prince, R. (2018). Cultured meat and cowless milk: On making markets for animal-free food. Journal of Cultural Economy 11(4), 315–329. Muehlhauser, L. (2015a). Four focus areas of effective altruism. In R. Carey (ed.), The effective altruism handbook (pp. 99-106). Retrieved from https://web.archive.org/web/20150701172316/http://www.careyryan.com/files/EA_Hand book.pdf. Muehlhauser, L. (2015b). What do we know about AI timelines? Retrieved from https://web.archive.org/web/20160510174151/https://www.openphilanthropy.org/focus/g lobal-catastrophic-risks/potential-risks-advanced-artificial-intelligence/ai-timelines. Muniesa, F. (2017). On the political vernaculars of value creation. Science as Culture 26(4), 445–54. OpenAI (2018, March 9). OpenAI Charter. Retrieved from https://web.archive.org/web/20190311213352/https://openai.com/charter/. Open Philanthropy (2018). Our progress in 2018 and plans for 2019. Retrieved from https://web.archive.org/web/20220701180235/https://www.openphilanthropy.org/researc h/our-progress-in-2018-and-plans-for-2019/ Ord, T. (2020). The precipice: Existential risk and the future of humanity. Hachette. O'Riordan, K. Fotopoulou, A. & Stephens, N. (2016) The first bite: Imaginaries, promotional publics and the laboratory grown burger Public Understanding of Science 26(2), 148- 163. Parfit, D. (2015). Reasons, persons, and effective altruism [Talk at Harvard EA]. Retrieved from https://www.youtube.com/watch?v=q6glXJ7dVU0. Porter, T. M. (1997). Trust in numbers: The pursuit of objectivity in science and public life. Princeton University Press. Post, M. J. (2012). Cultured meat from stem cells: Challenges and prospects. Meat Science 92(3), 297–301. Rubenstein, J. (2016). The lessons of effective altruism. Ethics & International Affairs 30(4), 511–526. https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 37 Russell, S., Dewey, D. & Tegmark, M. (2015). Research priorities for robust and beneficial artificial intelligence. AI Magazine 36(4), 105–114. Russell, S. J. (2019). Human compatible: Artificial intelligence and the problem of control. Viking. Saunders-Hastings, E. (2019). Benevolent giving and the problem of paternalism. In H. Greaves & T. Pummer (eds.), Effective Altruism: Philosophical Issues (pp. 115–136). Schuster, J. & Woods, D. (2021). Calamity theory: Three critiques of existential risk. University of Minnesota Press. Sentience Politics (2016a). Fundamental Rights for Primates (Policy Paper). Retrieved from https://web.archive.org/web/20201215000000*/https://ea- foundation.org/files/Fundamental-Rights-for-Primates.pdf. Sentience Politics (2016b). Cultured meat: A pragmatic solution to the problems posed by industrial animal farming (Policy paper). Retrieved from https://web.archive.org/web/20210813005813/https://ea-foundation.org/files/cultured- meat.pdf. Sexton, A. E., Garnett, T., & Lorimer, J. (2019). Framing the future of food: The contested promises of alternative proteins. Environment and Planning E: Nature and Space 2(1), 47–72. Silver, I. (2007). Disentangling class from philanthropy: The double-edged sword of alternative giving’. Critical Sociology 33(3), 537–49. Singer, P. (2011). The expanding circle: Ethics, evolution, and moral Progress. Princeton University Press. Singer, P. (2015). The most good you can do: How effective altruism is changing ideas about living ethically. Yale University Press. Srinivasan, A. (2015, September). Stop the robot apocalypse. London Review of Books. Retrieved from https://www.lrb.co.uk/v37/n18/amia-srinivasan/stop-the-robot- apocalypse. Stephens, N. (2013) Growing meat in laboratories: The promise, ontology and ethical boundary- work of using muscle cells to make food Configurations: A Journal of Literature, Science and Technology 21(2), p159-83. https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR file:///C:/Users/stephenn/Downloads https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.lrb.co.uk/v37/n18/amia-srinivasan/stop-the-robot-apocalypse https://www.lrb.co.uk/v37/n18/amia-srinivasan/stop-the-robot-apocalypse 38 Stephens, N. & Ruivenkamp, M. (2016) Promise and Ontological Ambiguity in the In vitro Meat Imagescape: From Laboratory Myotubes to the Cultured Burger Science as Culture 25(3), p327-355. Stephens, N., Sexton, A. E. & Driessen, C. (2019). Making sense of making meat: Key moments in the first 20 years of tissue engineering muscle to make food. Frontiers in Sustainable Food Systems 3. Retrieved from https://www.frontiersin.org/article/10.3389/fsufs.2019.00045. Taillandier, A. (2021a). In the name of posthumanity: Visions and justifications of liberal order in contemporary anglophone transhumanism (Unpublished doctoral dissertation). Sciences Po, Paris. Taillandier, A. (2021b). ‘Staring into the singularity’ and other posthuman tales: Transhumanist stories of future change. History and Theory 60 (2), 215–33. Todd, B. (2013). How to assess the impact of a career. Retrieved from https://web.archive.org/web/20141117052130/https://80000hours.org/2013/07/how-to- assess-the-impact-of-a-career/. Todd, B. (2015). Why you should focus more on talent gaps, not funding gaps. Retrieved from https://web.archive.org/web/20160122001916/https://80000hours.org/2015/11/why-you- should-focus-more-on-talent-gaps-not-funding-gaps/. Torres, É. (2023). Human extinction: The science and ethics of human annihilation. Routledge. Turner, F. (2006). From counterculture to cyberculture: Stewart Brand, the Whole Earth Network, and the rise of digital utopianism. University of Chicago Press. UK Department for Science, Innovation and Technology (2023). Capabilities and risks from Frontier AI: A discussion paper on the need for further research into AI risk. [Discussion paper in preparation of the UK AI safety summit, 1-2 November 2023]. Retrieved from https://web.archive.org/web/20231026162808/https://assets.publishing.service.gov.uk/me dia/65395abae6c968000daa9b25/frontier-ai-capabilities-risks-report.pdf. Vanderslott, S. (2019). Moving from outsider to insider status through metrics: the inclusion of “neglected tropical diseases” into the sustainable development goals. Journal of Human Development and Capabilities, 20(4), 418-435. Vanderslott, S. (2021). Attention and Responsibility in Global Health: The Currency of Neglect. Routledge. Retrieved from https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 39 https://www.taylorfrancis.com/books/mono/10.4324/9780429355448/attention- responsibility-global-health-samantha-vanderslott Wenar. L. (2024, March 27) The deaths of effective altruism. Wired. Retrieved from https://www.wired.com/story/deaths-of-effective-altruism/. Wiblin, R. (2017) Toby Ord on why the long-term future of humanity matters more than anything else, and what we should do about it. Retrieved from https://web.archive.org/web/20180808133146/https://80000hours.org/podcast/episodes/w hy-the-long-run-future-matters-more-than-anything-else-and-what-we-should-do-about- it/. Whittlestone, J. (2015). Research into risks from artificial intelligence. Retrieved from https://web.archive.org/web/20160627024844/https://80000hours.org/career- reviews/artificial-intelligence-risk-research/. Yudkowsky, E. (2001). Creating friendly AI 1.0: The analysis and design of benevolent goal architectures (Report of the Singularity Institute for Artificial Intelligence). Retrieved from https://web.archive.org/web/20130810014031/https://intelligence.org/files/CFAI.pdf. Authors biography Apolline Taillandier is postdoctoral research fellow at the Leverhulme Centre for the Future of Intelligence, Newnham College, and the Department of Politics and International Studies at the University of Cambridge, jointly with the Center for Science and Thought at the University of Bonn. Her work focuses on the political theory and history of transhumanism, AI, and liberalism focusing on the United States and Britain since the 1960s. Neil Stephens is Associate Professor of Technology and Society at the University of Birmingham. He is a sociologist and Science and Technology Studies scholar focused upon innovation in biotechnology, with themes including cultured meat, mitochondrial donation, and regenerative medicine. Samantha Vanderslott leads the Vaccines and Society Unit (VAS), hosted by the Oxford Vaccine Group at the University of Oxford. She is a health sociologist and associate professor working on topics at the intersection of health and society. Her research currently focuses on public attitudes and decisions on vaccination, particularly in relation to pro-vaccination behaviours and vaccine acceptance. https://www.wired.com/story/deaths-of-effective-altruism/ https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR https://www.zotero.org/google-docs/?HZ2JFR 40 Notes i MacAskill (2020) draws in part on Nick Bostrom’s claim that ‘existential risks’ – unlikely risks to humanity’s survival or potential for future flourishing, such as extreme technological risks related to AI, should be granted specific epistemic status and prioritized in global policy (Bostrom, 2002, 2013, 2014) – and on a well-established philosophical discussion in ‘population ethics’ investigating moral obligations towards future generations, most notably by Derek Parfit’s influential Reasons and Persons (1984). See also Beckstead (2013). ii Transhumanism aims to fundamentally change the human condition through technoscience. On the relationship between EA and transhumanism, see Srinivasan (2015), Torres (2023). iii This idea is captured by the notion of an expanding ‘moral circle’ developed by moral philosopher and EA advocate Peter Singer (2011). https://www.zotero.org/google-docs/?Izzzf8 https://www.zotero.org/google-docs/?Izzzf8