EA Organization Updates: April 2021

By Aaron Gertler 🔸 @ 2021-04-28T10:52 (+32)

These monthly posts originated as the "Updates" section of the EA Newsletter.

You can see previous updates in our repository of past newsletters.

Organization Updates

80,000 Hours

This month, Rob Wiblin spoke with Ezra Klein on aligning journalism, politics, and what matters most, Kelly Wanser on whether to deliberately intervene in the climate, and Nina Schick on disinformation and the rise of synthetic media.

Ben Todd wrote about how to identify your personal strengths, and co-authored an EA Forum post (with Max Daniel) on how much performance differs between people.

Anima International

March saw the first meeting of the Parliamentary Group on the Future of Food and Innovative Agriculture in Poland, which was organised and inaugurated by Anima International activist Sabina Sosin. The aim of the group is to find solutions which enable Poland to maintain its position as a European leader in food production, while at the same time reducing the environmental and social costs which this production entails. The group hopes to achieve this goal by shifting Poland’s focus towards plant-based protein sources.

Otwarte Klatki published a photographic investigation carried out in Thailand by activist-photographer Andrew Skowron.

Animal Charity Evaluators

Animal Charity Evaluators shared a video presentation from their virtual all-staff retreat, in which board member Zach Freitas-Groff explores the history of longtermism in EA, implications for effective animal advocacy, and ideas for better understanding long-term impacts. ACE also produced a blog post on free courses for animal advocates, which highlights a number of resources for professionals seeking to up their skills and expand their networks. 

Thanks to a partnership with The Giving Block, ACE now provides a convenient option to donate a variety of cryptocurrencies!

Animal Ethics

Animal Ethics published “The use of insects for food” to spread awareness and increase concern about curbing this activity. They also published a variety of blog posts in Spanish, Portuguese, French, and German.

Animal Ethics has been developing a series of workshops based on each module of their Wild Animal Suffering video course. The workshops are scheduled through 5 June 2021.

The talk given by Oscar Horta of Animal Ethics and Michelle Graham of Wild Animal Initiative at the EA Global Student Summit last fall is now available in video and transcript form.

Aquatic Animal Alliance

AAA surpassed 30 member organizations worldwide in April. The Alliance submitted recommendations to the US’s National Oceanic and Atmospheric Administration covering methods for making aquaculture more resilient to climate change through animal welfare improvements.

Aquatic Life Institute

ALI released a report on “blue loss,” which found that 1.2 trillion aquatic animals are fed to farmed aquatic animals each year. ALI then published an amendment in the French Parliament calling for a reduction in aquatic animals used as farmed animal feed. ALI attended as a technical expert in several virtual consultation meetings to support the UN’s Food and Agriculture Organization in developing new global Guidelines for Sustainable Aquaculture.

Berkeley Existential Risk Initiative

BERI has published their 2020 annual report. This document includes both high-level and program-specific summaries of their activities in 2020, plans for each program in 2021, and some specific, quantitative predictions for 2021.

BERI is now hiring a full-time US-based Research Assistant to Professor Nick Bostrom.

Cellular Agriculture Society

CAS published a set of new webpages on their website, including:

CAS Fellows: A collaboration page highlighting the nonprofit’s international group of 25 Fellows, representing over 15 countries.

CAS Vision: A visioning page exploring the history, meaning, and development of the CAS website.

Harvard: A collaboration page outlining the various projects, partnerships, and Harvard-CAS team members associated with the nonprofit.

Google: A collaboration page describing CAS’s work with Google.

Education: An education page detailing how the CAS website can be used as an instrument for increasing global awareness of cellular agriculture.

Centre for Effective Altruism

CEA published an update on their activity in the first quarter of 2021.

Center for Human-Compatible AI

CHAI researchers and former interns Daniel Filan, Stephen Casper, Shlomi Hod, Cody Wild, Andrew Critch, and Stuart Russell published “Clusterability in Neural Networks” on arXiv.

Current and former CHAI members including Paul Knott, Micah Carroll, Anca Dragan, and Rohin Shah co-authored the paper “Evaluating the Robustness of Collaborative Agents” alongside Sam Devlin, Kamil Ciosek, and Katja Hofmann.

The Science article “Who needs a teacher? Artificial intelligence designs lesson plans for itself” features two recent papers from members of CHAI.

Thomas Krendl Gilbert co-authored “AI Development for the Public Interest: From Abstraction Traps to Sociotechnical Risks,” published in IEEE ISTAS 20. 

CHAI faculty and affiliates presented the following papers at AAAI 2021:

• “Probabilistic Dependency Graphs” by Oliver Richardson and Joseph Y. Halpern of Cornell University
• “Asking the Right Questions: Learning Interpretable Action Models Through Query Answering” by Pulkit Verma, Shashank Rao Marpally, and Siddharth Srivastava of Arizona State University
• “Unifying Principles and Metrics for Safe and Assistive AI” by Siddharth Srivastava

Stuart Russell participated in Transformation in Action: Positive Futures, a plenary in the World Economic Forum 2021 Global Technology Governance Summit.

Pieter Abbeel recently launched a new podcast, The Robot Brains, where he hosts leading experts in AI robotics. 

Daniel Filan’s podcast AXRP released three new episodes featuring conversations with Beth Barnes, Vanessa Kosoy, and Evan Hubinger.

The podcast TalkRL features Michael Dennis in a recent episode.

Center on Long-Term Risk

• Jesse Clifton and Stefan Torges were confirmed as Co-Executive Directors by the board of CLR. Jesse will set the overall strategy and lead their research efforts. Stefan will be in charge of their operations, grantmaking, and community-building efforts.
• Jia Yuan Loke published his report on case studies of self-governance to reduce technology risk on the EA Forum. In it, he distills takeaways for current self-governance efforts in the realm of artificial intelligence. He started this project as a summer research fellow at CLR in 2020. 
• In a short post on LessWrong, Alex Lyzhov argues that the "AI and Compute" trend isn't predictive of what is happening. It was first postulated by researchers at OpenAI in 2018 but seems to have stalled significantly since then. This could have implications for how quickly we should expect the development of transformative systems.
• The Centre for the Governance of AI at the University of Oxford published “AI Policy Levers: A Review of the U.S. Government’s Tools to Shape AI Research, Development, and Deployment.” Stefan Torges from CLR is one of the co-authors. He contributed to this project while he was a research fellow at the Centre last year.

Centre for the Study of Existential Risk (CSER)

CSER researchers published two papers on AI ethics in the British Medical Journal:

• “Using AI ethically to tackle covid-19” looks at how relaxing ethical requirements in a crisis could have unintended harmful consequences that may outlast the pandemic. 
• “Does “AI” stand for augmenting inequality in the era of covid-19 healthcare?” highlights potential consequences arising from the AI now making clinical choices at scale if it does so based on biased data.

Lalitha Sundaram published a paper on biosafety in DIY‐bio laboratories in Embro Reports, showing how discussions about regulating DIY biology tend to ignore the extent of self‐regulation and oversight of DIY laboratories.

S. J. Beard and Julius Weitzdӧrfer published Double Debt Disaster, a book on law, policy, and social justice in the wake of Japan's 2011 tsunami. The book examines the increasingly serious and widespread phenomenon of 'double-loan problems' following mega-disasters.

Matthijs Maas has published a paper presented at the International Studies Association's 2021 convention. “AI, Governance Displacement, and the (De)Fragmentation of International Law” describes AI’s impacts on the global governance architecture. Matthijs also appeared on The Leiden Security and Global Affairs podcast to discuss his work.

Lauren Holt and Bram Thomas Arnold produced an audio artwork as part of their contribution to the recently published Dasgupta Review. “Sunlight, but dreaming” was created through a remote workshop, where philosophers, thinkers, and theologians were invited to comment on the non-instrumental and non-economic value of biodiversity.

Faunalytics

This month, Faunalytics is raising funds to support their core operating expenses. Donations to this campaign will help Faunaytics focus on their capacity-building programs, and they would be grateful for the support of the EA community.

Faunalytics produced a new webinar: Designing Effective Surveys. This resource covers survey preparation, defining and operationalizing a research question, sampling considerations, question wording, attrition, data quality checks, ethical considerations, and more. 

Their research library includes new study summaries on topics including plant-based food labels, the online wildlife trade, and cognitive function in arthropods. Additionally, their recent guest-blog examines the ethology of lower trophic sea life, and why consuming them is an ethical problem.

Fish Welfare Initiative

In India, FWI finalized their farm selection for an upcoming pilot study. They also hired two new staff, a Fish Welfare Expert and a Program Coordinator, to lead parts of the work there.

In the Philippines, they are working part-time with Chiawen and Pia (members of EA Philippines) to investigate and launch opportunities for fish advocacy work there.

Future of Humanity Institute

The Centre for the Governance of AI published “AI Policy Levers: A Review of the U.S. Government’s Tools to Shape AI Research, Development, and Deployment” (Fischer at al.), Carolyn Ashurst spoke at the Ethics in AI Seminar: Responsible Research and Publication in AI, and Toby Ord published “The Edges of Our Universe”, which explores the fundamental causal limits on how much of the universe we can observe or affect.

GiveWell

GiveWell published pages on two recent Incubation Grants related to their expanding work on policy:

• In January 2021, GiveWell recommended a grant of approximately $7 million from Open Philanthropy to the Centre for Pesticide Suicide Prevention (CPSP). With this funding, CPSP plans to expand to provide support to China and work with relevant UN agencies and regulatory bodies to assist countries in which CPSP does not currently have a program. CPSP will also continue its work in Nepal and India. This grant follows a grant of $198,320 that GiveWell recommended to the Effective Altruism Global Health and Development Fund to maintain CPSP's operations.
• In January 2021, Vital Strategies received a GiveWell Incubation Grant of $100,000 to convene a consortium to develop a proposal to work on advocacy and technical assistance for alcohol policy in low- and middle-income countries.

GiveWell also updated their Research on Programs page.

Giving What We Can

Giving What We Can members have been publishing a series of commonly asked questions, myths, and misconceptions about charity. So far this has included: “Can an individual donor really make a difference?”, “Does corruption in recipient governments interfere with foreign aid?”, “Can foreign aid and international charity make a difference?”, and “Why should we donate money to charity?” If you would like to be involved with this effort, please volunteer (you don’t need to be a member of GWWC).

GWWC has also been collaborating on an event series alongside One For The World and various other EA organisations. 

The Humane League

THL UK worked alongside coalition partners to negotiate a Better Chicken Commitment from Burger King in the UK, which will reduce the suffering of chickens raised for meat sold at more than 500 UK locations. THL UK is continuing to pressure supermarkets across the country, while THL US is campaigning against major restaurant chains and retailers like Costco — take action here.

Around the world, THL continues to work to free hens from battery cages. The Open Wing Alliance is pressuring Hard Rock International to go cage-free globally, while Nestle and Valora Group fulfilled their European cage-free commitments. In the US, more than 95 million hens now live in cage-free housing, which Vox called “easily one of the biggest successes of the animal welfare movement.”

To further these efforts, THL distributed $1,425,000 in grants to 39 groups on six continents through the Open Wing Alliance.

THL also published their 2020 Annual Report.

Open Philanthropy

Open Philanthropy announced grants including $27.1M to the Malaria Consortium to support seasonal malaria chemoprevention programs, $11.3M to UC Berkeley to support the Center for Human-Compatible Artificial Intelligence, $7.5M to the Centre for Effective Altruism for general support, and $1.7M to Dignity and Power Now to support efforts to reduce incarceration. Open Philanthropy also published a blog post examining one approach to forecasting when artificial general intelligence will be developed. 

Ought

Ought has continued to develop Elicit’s classification capabilities using GPT-3. In two weeks, they’ve worked closely with 20 users to help answer questions such as:

• Is this publication a systematic review? Should this publication be included in a systematic review?
• How should we categorize these free-form messages?
• Is this publication AI-related? 
• What is a higher-level schema for the UN’s keywords that will help predict country-level voting behavior? 

Ought has learned that GPT-3 based classification works best when the category labels are descriptive (“not AI-related” works better than “false”). They are now testing whether explanations and decomposition of the classification reasoning improve performance. 

If you’re spending too much time or money labeling large datasets, they’d love to help; you can find their contact information on this page. 

Rethink Priorities

In March, Rethink Priorities (RP) launched their first research internship program, inviting candidates interested in animal welfare, longtermism, and EA movement research to apply. In result, RP received 665 applications. RP is now also hiring a Research Project and Hiring Manager. The application deadline is 13 May 2021. 

One of RP’s staff members, Daniela Waldhorn, joined the Aquatic Animal Advocacy Panel on shrimp farming and welfare, organized by the Aquatic Life Institute. A recording of this session can be found here.

Wild Animal Initiative

Wild Animal Initiative is currently interviewing candidates for their Director of Scientific Affairs position (applications are closed). This position is part of Wild Animal Initiative’s plan to expand and optimize the services they provide to academics conducting wild animal welfare research. 

Wild Animal Initiative submitted a public comment urging the EPA to consider an avian pesticide’s impact on wild animal welfare. 

WAI researcher Luke Hecht summarized the historical context and welfare implications of the UK’s new gray squirrel contraceptive plan.

Add your own update

If your organization isn't shown here, you can provide an update in a comment.

You can also email me if you'd like to be one of the organizations I ask for updates each month. (I may not accept all such requests. Whether I include an org depends on its size, age, focus, track record, etc.)