AI is undergoing a paradigm shift with the rise of models (e.g., BERT, DALL-E, GPT-3) that are trained on broad data at scale and are adaptable to a wide range of downstream tasks. We call these models foundation models to underscore their critically central yet incomplete character. This report provides a thorough account of the opportunities and risks of foundation models, ranging from their capabilities (e.g., language, vision, robotics, reasoning, human interaction) and technical principles(e.g., model architectures, training procedures, data, systems, security, evaluation, theory) to their applications (e.g., law, healthcare, education) and societal impact (e.g., inequity, misuse, economic and environmental impact, legal and ethical considerations). Though foundation models are based on standard deep learning and transfer learning, their scale results in new emergent capabilities,and their effectiveness across so many tasks incentivizes homogenization. Homogenization provides powerful leverage but demands caution, as the defects of the foundation model are inherited by all the adapted models downstream. Despite the impending widespread deployment of foundation models, we currently lack a clear understanding of how they work, when they fail, and what they are even capable of due to their emergent properties. To tackle these questions, we believe much of the critical research on foundation models will require deep interdisciplinary collaboration commensurate with their fundamentally sociotechnical nature.

This report from the Montreal AI Ethics Institute covers the most salient progress in research and reporting over the second quarter of 2021 in the field of AI ethics with a special emphasis on "Environment and AI", "Creativity and AI", and "Geopolitics and AI." The report also features an exclusive piece titled "Critical Race Quantum Computer" that applies ideas from quantum physics to explain the complexities of human characteristics and how they can and should shape our interactions with each other. The report also features special contributions on the subject of pedagogy in AI ethics, sociology and AI ethics, and organizational challenges to implementing AI ethics in practice. Given MAIEI's mission to highlight scholars from around the world working on AI ethics issues, the report also features two spotlights sharing the work of scholars operating in Singapore and Mexico helping to shape policy measures as they relate to the responsible use of technology. The report also has an extensive section covering the gamut of issues when it comes to the societal impacts of AI covering areas of bias, privacy, transparency, accountability, fairness, interpretability, disinformation, policymaking, law, regulations, and moral philosophy.

Today's supply chains are laden with inefficiencies, as most companies rely on antiquated practices to oversee and manage how goods get from place to place. The supply chain is delicate -- even one disruption among suppliers, buyers, and logistics providers can have a trickle-down effect that causes waste, time loss, and increased carbon emissions. With the supply chain still managed manually, logistics managers are operating under intense pressure, with the sheer amount of data about material supply, demand, and transportation routes overwhelming. Even with machine learning providing managers with intelligent analysis, logistics managers can only react so quickly to the thousands of changes along a single supply chain. As managers are overburdened, their slow reactions to real-time problems and disruptions cause the supply chain inefficiencies that create higher costs, waste, and even greater environmental impact.

There is mounting public concern over the influence that AI based systems has in our society. Coalitions in all sectors are acting worldwide to resist hamful applications of AI. From indigenous people addressing the lack of reliable data, to smart city stakeholders, to students protesting the academic relationships with sex trafficker and MIT donor Jeffery Epstein, the questionable ethics and values of those heavily investing in and profiting from AI are under global scrutiny. There are biased, wrongful, and disturbing assumptions embedded in AI algorithms that could get locked in without intervention. Our best human judgment is needed to contain AI's harmful impact. Perhaps one of the greatest contributions of AI will be to make us ultimately understand how important human wisdom truly is in life on earth.

Climate change is a clear and present danger to the world economy. The tech industry bears its share of responsibility, not just for carbons emission but deforestation, plastic, chemical and other waste contamination, resource depletion and other damaging activities. But the tech industry also has the capacity to dramatically change the trajectory of all these problems; to at least slow down, if not reverse, the harm being done to our one and only planet. Artificial Intelligence (AI) in particular is already having a remarkable impact on issues that seemed intractable only a few years ago. Rather than being bad for the climate, AI is proving to help.

Citing opinions is a powerful yet understudied strategy in argumentation. For example, an environmental activist might say, "Leading scientists agree that global warming is a serious concern," framing a clause which affirms their own stance ("that global warming is serious") as an opinion endorsed ("[scientists] agree") by a reputable source ("leading"). In contrast, a global warming denier might frame the same clause as the opinion of an untrustworthy source with a predicate connoting doubt: "Mistaken scientists claim [...]." Our work studies opinion-framing in the global warming (GW) debate, an increasingly partisan issue that has received little attention in NLP. We introduce DeSMOG, a dataset of stance-labeled GW sentences, and train a BERT classifier to study novel aspects of argumentation in how different sides of a debate represent their own and each other's opinions. From 56K news articles, we find that similar linguistic devices for self-affirming and opponent-doubting discourse are used across GW-accepting and skeptic media, though GW-skeptical media shows more opponent-doubt. We also find that authors often characterize sources as hypocritical, by ascribing opinions expressing the author's own view to source entities known to publicly endorse the opposing view. We release our stance dataset, model, and lexicons of framing devices for future work on opinion-framing and the automatic detection of GW stance.

Leading conservationists have emphasized that conservation's priority is the protection of species and populations, not the welfare of individual nonhuman animals (hereafter “animals”) ([ 1 ][1]–[ 3 ][2]). Although individual conservationists often harbor concern for animal welfare, conservation organizations and scientists frequently downplay or ignore the ethical implications of actions they promote that harm individual animals, from culling and sport hunting to the discontinuation of wildlife rescue from oil spills ([ 3 ][2]–[ 5 ][3]). A growing body of scientific evidence should prompt conservation organizations to reconsider their inattention to animal welfare. A wide variety of vertebrate species (and perhaps some invertebrates) are capable of experiencing physical and emotional pain, engaging in substantive relationships, and executing cognitively complex tasks ([ 6 ][4]–[ 8 ][5]), bolstering claims that animal well-being is morally significant and policy-relevant. Addressing animal welfare in conservation would be politically challenging, and given the central role of predation and competition in ecosystems, conservation science cannot altogether avoid difficult decisions; harming animals can be a necessary step toward a worthwhile goal. Despite these trade-offs, conservation organizations face a singular opportunity to reshape conservation into a discipline that promotes both the quantity of species and the quality of animal life. Although humans are exceptional in many ways, the once-popular belief that it is unscientific to ascribe emotions or thoughts to animals is now regarded as inconsistent with evolutionary theory, experimental evidence, and any reasonable burden of proof ([ 9 ][6], [ 10 ][7]). Commonalities in basic neural functioning across vertebrate species, ranging from fish to mammals, suggest similarities in experiential capacities ([ 9 ][6], [ 11 ][8]). Evidence indicates that the thalamocingulate division of the limbic system and the anterior cingulate cortex evolved prior to the radiation of mammals, with all studied mammals sharing seven basic emotional systems including joy, fear, grief, parental nurturance, and playfulness. Deep neurological similarities underpin the extensive use of mammalian models in medical research, including for depression and anxiety. Further, research indicates that convergent evolution of the mammalian cortex and avian pallium has led to similar neural architecture between birds and mammals ([ 12 ][9]), with birds exhibiting similar forms of some affective states, consciousness, and attachment-oriented behaviors. Recent research has also demonstrated that various animal species are cognitively sophisticated, with findings including tool use in diverse taxa; spontaneous insight and innovative behavior; self-recognition and metacognition; collaboration to solve unfamiliar tasks; planning for future events; political strategy; empathetic concern; and the ability to recognize hundreds of human words (see supplementary materials). The accumulating scientific evidence that animals have vibrant inner lives was anticipated by modern philosophers, the field of animal welfare science, and numerous world cultures that have accorded moral relevance to the quality of animal life. Yet with limited exceptions, the most prominent international conservation organizations do not attempt to promote animal welfare in their mission or vision statements or to safeguard animal welfare in their readily available public policies. This contrasts with often robust ethics policies on a range of other social and environmental issues. From one perspective, the omission of animal welfare is befuddling. Conservationists must believe that animals deserve protection from human-induced harm; by combating habitat destruction and poaching, conservation often already promotes wild animal welfare. Officially recognizing the imperative of protecting animals as individuals could broaden conservation's constituency. Whereas the public often finds the value of biodiversity to be abstract and unrelatable, many people are concerned when human actions unnecessarily violate the freedom and well-being of wild animals. Conservation organizations have realized this, often using stories of human-induced suffering of individual animals to generate empathy and raise funds. Yet, owing to the pervasiveness of activities that compromise animal welfare, many conservation organizations could face a variety of political risks and programmatic complications if they were to officially endorse the legitimacy of animal welfare concerns. Conservation organizations often depend on a diverse coalition of political interests, including groups that habitually harm animals. For instance, the U.S. government is one of the largest bilateral sources of funding for international conservation largely because the U.S. Congress's International Conservation Caucus is among the largest bipartisan caucuses in the legislature, with many participants being vocal supporters of recreational hunting and fishing. For conservation organizations to acknowledge that killing animals for recreation might have moral implications ([ 4 ][10]) could complicate these politically important relationships in both halls of power and remote settings globally. There are well-evidenced concerns for how wild animals, especially wide-ranging species like elephants, some cetaceans, and carnivores, fare in captivity, but zoos can also inspire considerable support for wildlife conservation. Finally, conservation organizations and conservationists themselves (like other environmentalists) often regularly purchase factory farm products even though factory farms pose serious concerns about human-induced animal suffering. For conservation organizations, officially acknowledging the moral significance of animal welfare could complicate how many conservationists see themselves and generally cause discontent within their communities. Furthermore, conservation programming takes place in complex socioecological systems that pose practical trade-offs between animal welfare and biodiversity conservation or even human rights. At its extreme, efforts to curtail hunting and fishing in the world's poor rural areas could unjustly harm communities that rely on bushmeat or wild fish for their nutrition and livelihoods. Conservation groups can be seen as elitist, out-of-touch, or culturally oppressive where they oppose the killing of dangerous animals like elephants or traditional practices like subsistence whaling—such conflicts could become more common if conservation organizations consistently prioritize the interests of individual animals. In settings where wildlife tourism is not profitable, prohibiting sport hunting could deprive organizations of funding to protect wildlife from poaching, perversely leading to an increase in the killing of wildlife. Additionally, there are many examples of direct trade-offs between animal welfare and traditional conservation objectives like preventing extinction and maintaining ecosystem function. Invasive mammals—like goats on the Galapagos or feral cats on remote islands—suffer during eradication campaigns, but there may be no other way to secure the future for endangered native species. Programs to cull white-tailed deer similarly might be necessary to ensure the regeneration of forests in the eastern United States. Ecological research and reintroduction programs can also involve duress for the animals involved. Despite challenges posed by these trade-offs, conservation science should adjust its priorities in response to the overwhelming evidence that animals think and feel. Only explicit consideration of animal welfare in decision-making can ensure that conservation organizations do not unnecessarily compromise the well-being of individual animals. As a community, conservation organizations should set in motion three processes to (i) develop consensus principles, (ii) build the evidence base to identify best practices, and (iii) develop advisory institutions that can advance best practices. Each of these should engage diverse voices, including representatives from different cultures, countries with diverse political realities, and researchers and practitioners from both animal welfare science and conservation. The process of developing consensus principles to bring animal welfare concerns into conservation science has already begun, with ideas coming from national regulatory bodies, nongovernmental organizations concerned with wild animal welfare, the World Association of Zoos and Aquariums ([ 13 ][11]), animal welfare experts ([ 14 ][12], [ 15 ][13]), and the burgeoning compassionate conservation movement ([ 3 ][2], [ 4 ][10]). Conservation and animal welfare organizations should collaborate to systematically refine and select practically applicable ethical principles. Given the diverse cultural practices and economic systems that involve harm to animals, prohibitions on animal captivity, killing animals, and eating meat are unlikely to gain consensus support—but that need not prevent constructive discussions on minimizing human-induced suffering of animals, general agreement to minimize suffering during killing, and principles guiding the circumstances when killing animals might be acceptable. Animal welfare principles can alert conservationists to when the harm an activity causes to individual animals outweighs the benefits to biodiversity. Second, international conservation and animal welfare organizations should fund the development of an evidence base for how best to engage with wildlife in a way that minimizes avoidable suffering. Again, scientists have begun this process ([ 4 ][10], [ 13 ][11], [ 14 ][12])—but the evidence compiled must come from more diverse settings and situations and reflect practical limitations and trade-offs faced by conservation organizations in places where even human rights are not adequately realized. In addition to improving conservation practice, such evidence would help animal welfare organizations recognize where the protection of biodiversity, ecological function, and local communities might necessitate harming individual animals. This evidence review process would also highlight areas of research that could help resolve ethical dilemmas posed by conservation programming. International conservation bodies should also work with animal welfare scientists to establish advisory committees that review (voluntarily submitted) conservation project proposals to assess whether they satisfy principles of animal welfare. The process could be modeled as a voluntary version of the Institutional Animal Care and Use Committee that reviews animal research in the United States, working to promote best practices, build precedent, and collect real-life cases that can improve the evidence base. The committees' recommendations should provide a basis for informed debate about the trade-offs between wildlife conservation and animal welfare, helping better define whether the suffering of individual animals might be commensurate with conservation benefits ([ 14 ][12]). Over time, the cumulative experience of these committees should allow conservation organizations to recommend evidence-based animal welfare safeguards that can fit into the broader category of social and environmental safeguards, much like policies striving to minimize carbon emissions or protect human rights in conservation and development. Inevitably, these processes will take time. In the meanwhile, conservation organizations can take two steps toward building a better world for all animals: publicly commit to considering animal welfare in their decisions, and adopt policies against the purchase of factory farm meat where less harmful alternatives are available. Given the implications of factory farming not just for animal welfare but also for climate change and biodiversity, such action would further demonstrate the sincerity of conservation organizations' pursuit of a more just and sustainable planet. [science.sciencemag.org/content/369/6504/629/suppl/DC1][14] 1. [↵][15]1. M. E. Soule , Bioscience 35, 727 (1985). [OpenUrl][16][CrossRef][17][Web of Science][18] 2. 1. P. Kareiva, 2. M. Marvier , Bioscience 62, 962 (2012). [OpenUrl][19][CrossRef][20][Web of Science][21] 3. [↵][22]1. D. Ramp, 2. M. Bekoff , Bioscience 65, 323 (2015). [OpenUrl][23][CrossRef][24] 4. [↵][25]1. A. D. Wallach et al ., Conserv. Biol. 32, 1255 (2018). [OpenUrl][26] 5. [↵][27]1. P. Kareiva, 2. M. Marvier, 3. B. Silliman 1. J. A. Estes, 2. M. T. Tinker , in Effective Conservation Science: Data Not Dogma, P. Kareiva, M. Marvier, B. Silliman, Eds. (Oxford Univ. Press, 2017), pp. 128–134. 6. [↵][28]1. V. A. Braithwaite, 2. P. Boulcott , Dis. Aquat. Organ. 75, 131 (2007). [OpenUrl][29][PubMed][30] 7. 1. I. B.-A. Bartal et al ., Science 334, 1427 (2011). [OpenUrl][31][Abstract/FREE Full Text][32] 8. [↵][33]1. N. S. Clayton, 2. A. Dickinson , Nature 395, 272 (1998). [OpenUrl][34][CrossRef][35][PubMed][36][Web of Science][37] 9. [↵][38]1. J. Panksepp , PLOS ONE 6, e21236 (2011). [OpenUrl][39][CrossRef][40][PubMed][41] 10. [↵][42]1. G. A. Mashour, 2. M. T. Alkire , Proc. Natl. Acad. Sci. U.S.A. 110 (suppl.2), 10357 (2013). [OpenUrl][43][Abstract/FREE Full Text][44] 11. [↵][45]1. T. E. Feinberg, 2. J. Mallatt , Front. Psychol. 4, 667 (2013). [OpenUrl][46][PubMed][47] 12. [↵][48]1. A. B. Butler, 2. R. M. J. Cotterill , Biol. Bull. 211, 106 (2006). [OpenUrl][49][CrossRef][50][PubMed][51][Web of Science][52] 13. [↵][53]1. D. J. Mellor et al ., Caring for Wildlife: The World Zoo and Aquarium Animal Welfare Strategy (WAZA Executive Office, 2015). 14. [↵][54]1. S. Dubois et al ., Conserv. Biol. 31, 753 (2017). [OpenUrl][55] 15. [↵][56]1. J. O. Hampton et al ., Conserv. Biol. 33, 751 (2019). [OpenUrl][57] Acknowledgments: We thank H. Telkänranta, N. Shah, S. Sekar, N. Mohapatra, D. Mistree, M.Malik, A.Lerner, K. Kolappa, S. Kishore, P. Hannam, G. Fricchione, M. Doshi, P. Chanchani, and four anonymous reviewers. This piece reflects the views of the authors and not the official positions of their organizations. [1]: #ref-1 [2]: #ref-3 [3]: #ref-5 [4]: #ref-6 [5]: #ref-8 [6]: #ref-9 [7]: #ref-10 [8]: #ref-11 [9]: #ref-12 [10]: #ref-4 [11]: #ref-13 [12]: #ref-14 [13]: #ref-15 [14]: http://science.sciencemag.org/content/369/6504/629/suppl/DC1 [15]: #xref-ref-1-1 "View reference 1 in text" [16]: {openurl}?query=rft.jtitle%253DBioscience%26rft_id%253Dinfo%253Adoi%252F10.2307%252F1310054%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [17]: /lookup/external-ref?access_num=10.2307/1310054&link_type=DOI [18]: /lookup/external-ref?access_num=A1985AUV2000010&link_type=ISI [19]: {openurl}?query=rft.jtitle%253DBioscience%26rft_id%253Dinfo%253Adoi%252F10.1525%252Fbio.2012.62.11.5%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [20]: /lookup/external-ref?access_num=10.1525/bio.2012.62.11.5&link_type=DOI [21]: /lookup/external-ref?access_num=000311661000006&link_type=ISI [22]: #xref-ref-3-1 "View reference 3 in text" [23]: {openurl}?query=rft.jtitle%253DBioscience%26rft_id%253Dinfo%253Adoi%252F10.1093%252Fbiosci%252Fbiu223%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [24]: /lookup/external-ref?access_num=10.1093/biosci/biu223&link_type=DOI [25]: #xref-ref-4-1 "View reference 4 in text" [26]: {openurl}?query=rft.jtitle%253DConserv.%2BBiol.%26rft.volume%253D32%26rft.spage%253D1255%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [27]: #xref-ref-5-1 "View reference 5 in text" [28]: #xref-ref-6-1 "View reference 6 in text" [29]: {openurl}?query=rft.jtitle%253DDiseases%2Bof%2Baquatic%2Borganisms%26rft.stitle%253DDis%2BAquat%2BOrgan%26rft.aulast%253DBraithwaite%26rft.auinit1%253DV.%2BA.%26rft.volume%253D75%26rft.issue%253D2%26rft.spage%253D131%26rft.epage%253D138%26rft.atitle%253DPain%2Bperception%252C%2Baversion%2Band%2Bfear%2Bin%2Bfish.%26rft_id%253Dinfo%253Apmid%252F17578252%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [30]: /lookup/external-ref?access_num=17578252&link_type=MED&atom=%2Fsci%2F369%2F6504%2F629.atom [31]: {openurl}?query=rft.jtitle%253DScience%26rft.stitle%253DScience%26rft.aulast%253DBartal%26rft.auinit1%253DI.%2BB.-A.%26rft.volume%253D334%26rft.issue%253D6061%26rft.spage%253D1427%26rft.epage%253D1430%26rft.atitle%253DEmpathy%2Band%2BPro-Social%2BBehavior%2Bin%2BRats%26rft_id%253Dinfo%253Adoi%252F10.1126%252Fscience.1210789%26rft_id%253Dinfo%253Apmid%252F22158823%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [32]: /lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEzOiIzMzQvNjA2MS8xNDI3IjtzOjQ6ImF0b20iO3M6MjI6Ii9zY2kvMzY5LzY1MDQvNjI5LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ== [33]: #xref-ref-8-1 "View reference 8 in text" [34]: {openurl}?query=rft.jtitle%253DNature%26rft.stitle%253DNature%26rft.aulast%253DClayton%26rft.auinit1%253DN.%2BS.%26rft.volume%253D395%26rft.issue%253D6699%26rft.spage%253D272%26rft.epage%253D274%26rft.atitle%253DEpisodic-like%2Bmemory%2Bduring%2Bcache%2Brecovery%2Bby%2Bscrub%2Bjays.%26rft_id%253Dinfo%253Adoi%252F10.1038%252F26216%26rft_id%253Dinfo%253Apmid%252F9751053%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [35]: /lookup/external-ref?access_num=10.1038/26216&link_type=DOI [36]: /lookup/external-ref?access_num=9751053&link_type=MED&atom=%2Fsci%2F369%2F6504%2F629.atom [37]: /lookup/external-ref?access_num=000075974600049&link_type=ISI [38]: #xref-ref-9-1 "View reference 9 in text" [39]: {openurl}?query=rft.stitle%253DPLoS%2BONE%26rft.aulast%253DPanksepp%26rft.auinit1%253DJ.%26rft.volume%253D6%26rft.issue%253D9%26rft.spage%253De21236%26rft.epage%253De21236%26rft.atitle%253DCross-species%2Baffective%2Bneuroscience%2Bdecoding%2Bof%2Bthe%2Bprimal%2Baffective%2Bexperiences%2Bof%2Bhumans%2Band%2Brelated%2Banimals.%26rft_id%253Dinfo%253Adoi%252F10.1371%252Fjournal.pone.0021236%26rft_id%253Dinfo%253Apmid%252F21915252%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [40]: /lookup/external-ref?access_num=10.1371/journal.pone.0021236&link_type=DOI [41]: /lookup/external-ref?access_num=21915252&link_type=MED&atom=%2Fsci%2F369%2F6504%2F629.atom [42]: #xref-ref-10-1 "View reference 10 in text" [43]: {openurl}?query=rft.jtitle%253DProc.%2BNatl.%2BAcad.%2BSci.%2BU.S.A.%26rft_id%253Dinfo%253Adoi%252F10.1073%252Fpnas.1301188110%26rft_id%253Dinfo%253Apmid%252F23754370%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [44]: /lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoyMjoiMTEwL1N1cHBsZW1lbnRfMi8xMDM1NyI7czo0OiJhdG9tIjtzOjIyOiIvc2NpLzM2OS82NTA0LzYyOS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30= [45]: #xref-ref-11-1 "View reference 11 in text" [46]: {openurl}?query=rft.jtitle%253DFront.%2BPsychol.%26rft.volume%253D4%26rft.spage%253D667%26rft_id%253Dinfo%253Apmid%252F24109460%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [47]: /lookup/external-ref?access_num=24109460&link_type=MED&atom=%2Fsci%2F369%2F6504%2F629.atom [48]: #xref-ref-12-1 "View reference 12 in text" [49]: {openurl}?query=rft.jtitle%253DThe%2BBiological%2BBulletin%26rft.stitle%253DBiol.%2BBull.%26rft.aulast%253DButler%26rft.auinit1%253DA.%2BB.%26rft.volume%253D211%26rft.issue%253D2%26rft.spage%253D106%26rft.epage%253D127%26rft.atitle%253DMammalian%2Band%2Bavian%2Bneuroanatomy%2Band%2Bthe%2Bquestion%2Bof%2Bconsciousness%2Bin%2Bbirds.%26rft_id%253Dinfo%253Adoi%252F10.2307%252F4134586%26rft_id%253Dinfo%253Apmid%252F17062871%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [50]: /lookup/external-ref?access_num=10.2307/4134586&link_type=DOI [51]: /lookup/external-ref?access_num=17062871&link_type=MED&atom=%2Fsci%2F369%2F6504%2F629.atom [52]: /lookup/external-ref?access_num=000241793700003&link_type=ISI [53]: #xref-ref-13-1 "View reference 13 in text" [54]: #xref-ref-14-1 "View reference 14 in text" [55]: {openurl}?query=rft.jtitle%253DConserv.%2BBiol.%26rft.volume%253D31%26rft.spage%253D753%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [56]: #xref-ref-15-1 "View reference 15 in text" [57]: {openurl}?query=rft.jtitle%253DConserv.%2BBiol.%26rft.volume%253D33%26rft.spage%253D751%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx

In spatial statistics, a common objective is to predict the values of a spatial process at unobserved locations by exploiting spatial dependence. In geostatistics, Kriging provides the best linear unbiased predictor using covariance functions and is often associated with Gaussian processes. However, when considering non-linear prediction for non-Gaussian and categorical data, the Kriging prediction is not necessarily optimal, and the associated variance is often overly optimistic. We propose to use deep neural networks (DNNs) for spatial prediction. Although DNNs are widely used for general classification and prediction, they have not been studied thoroughly for data with spatial dependence. In this work, we propose a novel neural network structure for spatial prediction by adding an embedding layer of spatial coordinates with basis functions. We show in theory that the proposed DeepKriging method has multiple advantages over Kriging and classical DNNs only with spatial coordinates as features. We also provide density prediction for uncertainty quantification without any distributional assumption and apply the method to PM$_{2.5}$ concentrations across the continental United States.

Climate change is seen as the biggest challenge to the future of human life on Earth, and understanding the scientific language used to describe it can sometimes feel just as difficult. But help is at hand. Use our translator tool to find out what some of the words and phrases relating to climate change mean. Keeping the rise in global average temperature below 1.5 degrees Celsius will avoid the worst impacts of climate change, scientists say.

We invited three industry expert speakers using AI to battle climate change. During the hour long webinar, Anita Faul, Data Scientist at the British Antarctic Survey, Lauren Kuntz, CEO and Co-Founder of Gaiascope and Topher White, CEO and Founder of Rainforest Connections walked us through their business use applications of AI to fight the change in climate. Anita started her talk with an explanation of the Thwaites Glacier, otherwise know as the'Doomsday Glacier'. This glacier is responsible for 4% of all sea level increase - if it were to melt completely, sea levels would rise by half a meter in total (hence the name). Therefore, Anita's objective at the Antarctic Survey is to identify icebergs efficiently and reliably in Synthetics Aperture Radar (SAR) satellite images to estimate ice loss.