Brian Tomasik

American researcher, ethicist, and essayist From Wikipedia, the free encyclopedia

Brian Tomasik (/tʌˈmɑːsɪk/ tuh-MAH-sick)[2] is an American researcher, ethicist, and essayist. He has written on wild animal suffering, suffering-focused ethics, the ethics of artificial intelligence, invertebrate sentience, artificial minds, and risks of astronomical suffering. He is the author of Essays on Reducing Suffering and of the 2009 essay "The Importance of Wild-Animal Suffering".

OthernameAlan Dawrst (former pseudonym)
CitizenshipAmerican[1]
Occupations
  • Researcher
  • ethicist
  • essayist
Quick facts Other name, Citizenship ...
Brian Tomasik
Tomasik in 2014
Other nameAlan Dawrst (former pseudonym)
CitizenshipAmerican[1]
EducationSwarthmore College (B.A. in computer science, 2009)
Occupations
  • Researcher
  • ethicist
  • essayist
Organizations
  • Center on Long-Term Risk (co-founder and advisor)
  • Center for Reducing Suffering (advisor)
Known forWork on wild animal suffering, suffering-focused ethics, and the ethics of artificial intelligence
Notable work
Website
Close

Education and career

Tomasik attended Guilderland Central High School in Guilderland, New York, graduating in 2005.[3] He has said that, as a teenager, he experienced chronic esophagitis, which contributed to his interest in suffering and ethics.[4]

He studied computer science at Swarthmore College, with minors in mathematics and statistics, and received a B.A. in 2009.[3][5] After college, he worked at Microsoft, applying statistics and machine learning to search ranking at Bing.[6]

Tomasik has said that by 2005 he had adopted a broadly utilitarian outlook, influenced in part by Peter Singer and Ralph Nader. He has also written that the works of Singer, Bernard Rollin, Yew-Kwang Ng, and David Pearce led him to consider wild animal suffering, including the suffering of insects and other invertebrates.[7]

He became involved in the strategy later known as earning to give, and in a 2012 interview described donating a substantial share of his income while working in the technology sector.[8] In 2013, he left Microsoft to focus on research on suffering reduction and co-founded the Center on Long-Term Risk, formerly the Foundational Research Institute.[9] From 2012 to 2015, he served on the board of Animal Charity Evaluators.[3]

Essays on Reducing Suffering

Tomasik has written that he began collecting philosophical writings in 2005 and 2006. After corresponding with David Pearce in 2006, he created a website to host them, first under the title Utilitarian Essays and later as Essays on Reducing Suffering.[10] He initially published under the pseudonym Alan Dawrst.[11][12]

The site contains essays on ethics, consciousness, artificial intelligence, and wild animal suffering, as well as interviews and donation recommendations.[13] Tomasik later moved the site to WordPress and then back to a hand-maintained format.[10]

Views

Moral anti-realism and moral progress

Tomasik identifies as a moral anti-realist. He argues that moral progress can still be made relative to personal values, which he considers meaningful because they reflect what individuals care about. He maintains that the emotional force often associated with moral truth can also apply to personal feelings about how one wants the world to be. Tomasik suggests that, in the long term, some convergence of values may occur through the emergence of a dominant decision-making system, or "singleton", as described by Nick Bostrom. He views such convergence as the result of power relations between competing factions rather than evidence of objective moral truth.[14]

Tomasik is an advocate of negative utilitarianism.[15] He supports a form of threshold negative utilitarianism based on consent. According to this view, some forms of suffering are so intense that they cannot be morally outweighed by happiness. To identify this threshold, Tomasik proposes considering whether the individual experiencing the suffering would consent to keep experiencing it in exchange for future benefit. If consent is withdrawn during the experience, the suffering is deemed to have exceeded the moral threshold and should not be justified.[16]

He contrasts this view with what he calls "consent-based positive utilitarianism", which would permit severe suffering if offset by sufficient happiness. Tomasik acknowledges the coherence of that view, but favours the negative utilitarian framework, which he regards as closer to his intuitions about the moral urgency of preventing suffering.[16]

Tomasik also questions the moral value of creating new happy beings, arguing that nonexistence is not inherently bad and that the drive to maximize happiness may reflect ideological bias. He characterizes his focus on reducing suffering as based on subjective intuition and acknowledges that moral values vary among individuals.[16]

Evidence-based ethical decision-making

Tomasik has argued for evidence-based reasoning, cost-effectiveness, and long-term strategy in ethical decision-making. He contends that emotional intuition and rigid ideology can obstruct more effective ways to reduce suffering, and has recommended focusing on neglected and tractable issues.[14][17]

Wild animal suffering

Tomasik's work centres on the reduction of suffering, including suffering among non-human animals.[17] He has described the suffering of animals in nature as "the most important current issue due to its sheer scale", and argues that wild animal suffering may exceed all other forms of suffering on Earth by several orders of magnitude.[8]

According to Tomasik, many wild animals, particularly small invertebrates, live short lives marked by hunger, disease, parasitism, predation, and early death. Given their numbers, he contends that their cumulative suffering may be among the largest sources of suffering on Earth.[17][18]

Interventions to reduce wild animal suffering

Tomasik has supported research, welfare-oriented environmental management, and long-term planning for ecosystems as possible ways to reduce wild animal suffering. He has also argued that reducing the number of wild animals through habitat destruction could be ethically justified if it reduced overall suffering.[17][19]

Among the measures he has discussed is replacing grass lawns with gravel to reduce invertebrate populations. He has also considered the possible use of gene drives to spread traits such as reduced pain sensitivity within wild populations. Tomasik notes that large-scale ecological interventions would probably disrupt existing equilibria and lead to biodiversity loss. He maintains that species extinction is not inherently morally relevant, and that biodiversity and ecosystem stability are valuable only to the extent that they benefit sentient beings.[14] He has also expressed support for the controlled reduction of wild invertebrate populations, including through sterilisation, as a possible way to prevent future suffering.[20]

Future risks of replicating wild animal suffering

Tomasik has argued that some advanced technologies could multiply wild animal suffering. He cites examples such as terraforming Mars or initiating directed panspermia to spread life to other planets, which could recreate Earth-like environments with high levels of suffering. He has also written that future computer simulations, especially those incorporating artificial intelligence, might reach a level of complexity at which simulated wild animals become sentient and capable of suffering. He argues that such possibilities should be evaluated before being pursued.[18] He has also expressed concern that the animal rights movement could increase support for wilderness preservation and non-interference, which he argues may perpetuate wild animal suffering rather than reduce it.[21][22]

Animal welfare assessment

Tomasik co-authored the 2017 paper "Framework for Integrating Animal Welfare into Life Cycle Sustainability Assessment", published in The International Journal of Life Cycle Assessment. The study proposed three indicators for evaluating animal welfare based on life quality, lifespan, and the number of animals required per unit of food. One indicator also incorporated cognitive traits such as intelligence. The study found that insect-based foods scored worst across all indicators because of high mortality and low yield, despite lower assumed sentience. The authors argued that simplified welfare metrics can be incorporated into sustainability assessments.[23]

Moral scope and animal ethics

Tomasik rejects speciesism, arguing that moral concern should be based on ethically relevant traits such as sentience and cognitive capacity rather than species membership. He also states that differences in treatment may be justified when they reflect relevant distinctions, such as communication ability or social complexity.[17]

He supports a pragmatic ethical framework resembling Robert Nozick's formulation of "utilitarianism for animals, Kantianism for people".[17] While he views consequentialist reasoning as appropriate when addressing animal suffering, he considers deontological norms, such as honesty, nonviolence, and respect for rights, tools for maintaining trust and cooperation within human societies.[17]

Ethical concerns about eating insects

Tomasik has argued against entomophagy, the practice of eating insects, on ethical grounds.[24] He acknowledges claims that insect farming may offer environmental benefits and cultural acceptability in some regions, but expresses concern about possible suffering on a large scale.[25] In particular, he notes that insects have high rates of reproduction and mortality, meaning that farming them for food could involve the deaths of very large numbers of individuals.[24]

Because insect consciousness is uncertain, Tomasik recommends caution. He argues that if insects are sentient, farming and slaughtering them could involve substantial suffering. He supports welfare standards where insect farming occurs, including more humane slaughter methods, but concludes that cultivating insects for food should be avoided.[25]

Ethical concerns about eating bivalves

Tomasik has expressed ethical reservations about eating mussels and other bivalves, citing evidence such as changes in morphine levels and environmental responsiveness that may indicate rudimentary sentience. While he considers their capacity for suffering uncertain and probably lower than that of insects, he avoids consuming them because of the number of individuals typically killed per meal and the common practice of boiling them alive. He has said that he finds dairy ethically preferable for this reason, while acknowledging that others may reach different conclusions.[26]

Ethics of artificial intelligence

Tomasik has written about the moral status of artificial minds, especially those built using reinforcement learning and related techniques. He argues that even simple artificial agents may warrant moral consideration because of structural similarities with animal learning systems.[27]

Artificial suffering and s-risks

Tomasik has warned that future technologies could create large numbers of suffering artificial minds, particularly if AI goals are misaligned or if computer simulations are used extensively to model sentient processes. He has described such scenarios as a possible source of astronomical suffering ("s-risks") and has called for AI governance and ethical safeguards to reduce the risk of such outcomes.[28]

Moral consideration of video game characters

In a 2014 interview with Vox, Tomasik argued that some non-player characters (NPCs) in video games may deserve limited moral consideration, depending on their behavioural complexity. He suggested that simple NPCs such as Goombas in Super Mario Bros. probably have negligible ethical relevance, while more advanced characters that display goal-directed behaviour, such as avoiding harm or adapting to player actions, may have a small degree of ethical relevance. He drew parallels between such NPCs and simple reinforcement learning agents, noting that if these systems pursue rewards or avoid punishments, they may embody forms of processing that could be morally relevant. While stating that individual NPCs have minimal ethical weight, Tomasik argued that the aggregate harm caused by large-scale simulated violence might become nontrivial. He also expressed concern that as NPCs become more lifelike and intelligent, their moral status could increase.[29]

In a 2021 interview with Wired, he reiterated that large-scale simulated harm could pose ethical risks, and suggested that some NPCs displaying goal-directed behaviour might merit moral attention.[30]

Consciousness

Tomasik has been described as endorsing a form of consciousness eliminativism, the view that consciousness does not exist as an ontologically distinct entity.[31] He describes consciousness as a high-level concept that humans ascribe to physical systems, rather than an objectively existing property. He rejects the existence of ontological qualia and does not accept the hard problem of consciousness, identifying instead as a type-A physicalist in the terminology of David Chalmers. He compares consciousness to concepts such as justice: socially constructed, morally relevant, and vague at the margins. In his view, people may agree on clear cases while differing about borderline instances.[14]

Tomasik combines reductionism with elements of panpsychism, treating consciousness as an emergent property of information-processing systems. He argues that whether a system is considered conscious, and the moral concern it warrants, is partly a normative question. He also supports using neuroscience and computer science to refine judgements about which systems are more likely to exhibit sentience.[14]

Reception and criticism

Wild animal suffering

Tomasik has contributed to interest in wild animal suffering among academic researchers and within the effective altruism community. He corresponded with figures such as Oscar Horta and helped establish early online spaces for discussion, including a Facebook group that later became "Reducing wild-animal suffering".[7]

His 2009 essay "The Importance of Wild-Animal Suffering", originally published on his website and later reprinted by the Center on Long-Term Risk and in the journal Relations. Beyond Anthropocentrism in 2015,[18][32] has been cited in discussions of animal ethics and welfare biology by scholars including Horta, Jeff Sebo, Alasdair Cochrane, Catia Faria, Kyle Johannsen, and Jacy Reese Anthis.[33] A 2016 research plan by the Foundational Research Institute described the essay as an early contribution to work presenting wild animal suffering as a moral concern and welfare biology as a possible framework for intervention.[34]

Tomasik's 2009 essay "How Many Wild Animals Are There?" has also been cited in academic literature on the scale of wild animal populations and their ethical relevance, particularly within debates on wild animal suffering, welfare biology, and longtermist ethics.[35]

Peter Singer has credited Tomasik, along with Oscar Horta and Catia Faria, with helping to establish wild animal suffering as a topic within animal ethics. Singer said their work influenced his decision to include the subject in Animal Liberation Now, after he had considered it too speculative to address in earlier editions of the book.[36] Singer also cited Tomasik and Horta in his 2015 book on effective altruism, The Most Good You Can Do.[37]

Insect welfare and human exceptionalism

In a 2014 article for National Review, Wesley J. Smith criticized Tomasik's concern for insect welfare as an example of rejecting human exceptionalism. Smith questioned the implications of prioritizing the potential suffering of insects, particularly in comparison with human interests, and expressed skepticism about proposals to reduce insect populations on moral grounds. He also cited Tomasik's suggestions for insect farming standards and concern over insect deaths in nature as examples of what he viewed as an excessive focus on sentience as the basis of moral concern.[38]

Habitat destruction

Philosopher Kyle Johannsen has criticized Tomasik's argument that intentional habitat destruction could be justified as a way to reduce wild animal suffering, particularly among r-strategist species. While acknowledging that such measures might reduce total suffering, Johannsen argues that Tomasik's position depends on a utilitarian framework that gives less weight to moral constraints against directly causing harm. From a moderate deontological perspective, Johannsen contends that negative duties, such as the duty not to kill or displace animals, are more stringent than positive duties to prevent suffering. He concludes that intentional habitat destruction is morally impermissible even if it would improve net outcomes, and instead supports cautious, harm-avoiding interventions in nature.[39]

In The Oxford Handbook of Consequentialism, philosophers Tyler M. John and Jeff Sebo describe Tomasik's view as exemplifying the "Logic of the Logger".[40] While recognizing the coherence of this position within consequentialism, they caution that it could unintentionally justify ecologically harmful actions and reduce empathy toward wild animals. They recommend a more cautious ethical stance, based on further empirical study of ecological systems, before endorsing large interventions based on such reasoning.[40]

Publications

  • ; Sutherland, Dougal (2008). "An Efficient Python Module for Lexical Distributional Similarity" (PDF). Swarthmore College. Department of Computer Science.
  • Golub, Stephen S.; (May 15, 2008). Measures of International Transport Cost for OECD Countries (Report). OECD Economics Department Working Papers. doi:10.1787/241707325051.
  • (2009). "A Minimum Description Length Approach to Multitask Feature Selection". arXiv:0906.0052 [cs.LG].
  • Dhillon, Paramveer S.; ; Foster, Dean; Ungar, Lyle (2009). "Multi-task Feature Selection Using the Multiple Inclusion Criterion (MIC)". In Buntine, Wray; Grobelnik, Marko; Mladenić, Dunja; Shawe-Taylor, John (eds.). Machine Learning and Knowledge Discovery in Databases. Lecture Notes in Computer Science. Vol. 5781. Berlin, Heidelberg: Springer. pp. 276–289. doi:10.1007/978-3-642-04180-8_35. ISBN 978-3-642-04180-8.
  • ; Kim, Joon Hee; Ladlow, Margaret; Augat, Malcolm; Tingle, Derek; Wicentowski, Richard; Turnbull, Douglas (2009). Using Regression to Combine Data Sources for Semantic Music Discovery (PDF). 10th International Society for Music Information Retrieval Conference (ISMIR 2009).
  • ; Thiha, Phyo; Turnbull, Douglas (July 19, 2009). "Tagging products using image classification" (PDF). Proceedings of the 32nd international ACM SIGIR conference on Research and development in information retrieval. SIGIR '09. New York, NY, USA: Association for Computing Machinery. pp. 792–793. doi:10.1145/1571941.1572131. ISBN 978-1-60558-483-6.
  • (2014). "Do Artificial Reinforcement-Learning Agents Matter Morally?". arXiv:1410.8233 [cs.AI].
  • (2015). "The Importance of Wild-Animal Suffering". Relations. Beyond Anthropocentrism. 3 (2) 1: 133–152. doi:10.7358/rela-2015-002-toma.
  • (2016). "Insects Should Not Be a Part of People's Diets". In Espejo, Roman (ed.). What Should We Eat?. New York, NY: Greenhaven Publishing. p. 82. ISBN 978-0-7377-7390-3.
  • Scherer, Laura; ; Rueda, Oscar; Pfister, Stephan (2018). "Framework for integrating animal welfare into life cycle sustainability assessment". The International Journal of Life Cycle Assessment. 23 (7): 1476–1490. Bibcode:2018IJLCA..23.1476S. doi:10.1007/s11367-017-1420-x. ISSN 0948-3349. PMC 6435210. PMID 30996531.
  • (August 7, 2019) [2009]. "How Many Wild Animals Are There?". Essays on Reducing Suffering.

See also

References

Further reading

Related Articles

Wikiwand AI