EA Forum Podcast (All audio)

Making good decisions under significant uncertainty is a central skill for anyone trying to understand how they can maximise the positive impact of their work or donations. This is something we look for at AIM for both our Charity Entrepreneurship program and when hiring new staff, such as in the four roles we currently are advertising.

In good news, decision-making is a skill that you can improve. Thinking about decisions as a weighted factor model (WFM) is a useful framework I’ve found for improving my decision-making.

I think about a lot of decisions in terms of WFM: I ask myself, what would make a great model versus a bad model for thinking about this issue? I think a good WFM, and thus a good decision, often comes down to two major factors I see a lot of variance in:

• Number of meaningfully different heuristics considered (the criteria [...]

---

Outline:

(01:18) Considering many heuristics

(03:13) Generating more options: numbers and divergence

(06:27) Bringing it all together

(07:44) Why does this matter?

The original text contained 1 image which was described by AI.

---

First published:

October 30th, 2024

Source:

https://forum.effectivealtruism.org/posts/ctm5iKzP9EcoNtBxB/thinking-about-decisions-as-a-weighted-factor-model

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

This is a link post.

We’re delighted to announce the winners of the Essay competition on the Automation of Wisdom and Philosophy.

Overview

The competition attracted 90 entries in total (only one of which was obviously just the work of an LLM!), taking a wide variety of angles on the topic. The judges awarded the top four prizes as follows:

• $7,000 to Rudolf Laine for essays on wisdom, amortised optimisation, and AI:
  • Part I: Wisdom, amortised optimisation, and AI
  • Part II: Growth and amortised optimisation
  • Part III: AI effects on amortised optimisation
• $6,000 to Thane Ruthenis for Towards the operationalization of philosophy & wisdom
• $4,000 to Chris Leong for essays on training wise AI systems:
  • An Overview of “Obvious” Approaches to Training Wise AI Advisors
  • Some Preliminary Notes on the Promise of a Wisdom Explosion
• $3,000 to Gabriel Recchia for Should we just be building [...]

---

Outline:

(00:25) Overview

(02:18) Judge introductions

(03:54) Top Prize Winners

(04:11) Wisdom, amortised optimisation, and AI — Rudolf Laine — $7,000

(04:40) Summary

(06:04) BS's comments

(06:42) CN's comments

(08:43) DM's comments

(12:10) Towards the operationalization of philosophy and wisdom — Thane Ruthenis — $6,000

(12:18) Summary

(13:56) DM's comments

(16:59) CN's comments

(19:21) BS's comments

(21:35) Essays on training wise AI systems — Chris Leong — $4,000

(22:06) Summary

(23:14) CN's comments

(24:28) BS's comments

(25:05) Should we just be building more datasets? — Gabriel Recchia — $3,000

(25:13) Summary

(26:01) CN's comments

(27:50) BS's comments

(28:38) Runner-up prizes

(28:42) Designing Artificial Wisdom: The Wise Workflow Research Organization — Jordan Arel

(29:12) Summary

(30:23) CN's comments

(30:43) Philosophys Digital Future — Richard Yetter Chappell

(30:48) Summary

(31:37) BS's comments

(32:36) CN's comments

(32:51) Synthetic Socrates and the Philosophers of the Future — Jimmy Alfonso Licon

(32:58) Summary

(33:44) CN's comments

(34:00) BS's comments

(34:21) The Web of Belief — Paal Fredrik S. Kvarberg

(34:26) Summary

(35:49) CN's comments

(36:14) Wise AI support for government decision-making — Ashwin Acharya and Michaelah Gertz-Billingsley

(36:21) Summary

(37:27) CN's comments

(37:57) Machines and moral judgement — Jacob Sparks

(38:02) Summary

(38:54) CN's comments

(39:52) The purpose of philosophical AI will be: to orient ourselves in thinking — Maximilian Noichl

(40:01) Summary

(40:41) CN's comments

(41:04) BS's comments

(41:32) Cross-context deduction: on the capability necessary for LLM-philosophers — Rio Popper and Clem von Stengel

(41:42) Summary

(42:33) CN's comments

(43:32) DM's comments

(46:13) Evolutionary perspectives on AI values — Maria Avramidou

(46:20) Summary

(47:08) CN's comments

(47:57) BS's comments

(48:07) Tentatively against making AIs wise — Oscar Delaney

(48:13) Summary

(49:11) CN's comments

(49:46) Concluding thoughts

(50:13) BS's thoughts

(53:16) DM's thoughts

(55:40) AS's thoughts

(57:22) Acknowledgements

The original text contained 2 footnotes which were omitted from this narration.

The original text contained 1 image which was described by AI.

---

First published:

October 28th, 2024

Source:

https://forum.effectivealtruism.org/posts/vYbgxrTGy3rYKZkcE/winners-of-the-essay-competition-on-the-automation-of-wisdom

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

I’ve written this in my personal capacity and all views are my own. I didn’t even integrate Claude's suggested improvements. Thanks to Zoe Sigle for the feedback on an earlier draft of this post.

Summary

• Many animal advocates frame the goal of the movement as "ending factory farming".
  • I see why it's a tempting message, both to hold onto internally, and when pitching to people new to the movement.
  • Yet, I think the reality is that we might never get there.
• I think the framing therefore leads to the following problems:
  • Unrealistic hope leads to disillusionment and burnout.
  • You should count counterfactual wins, not the absolute numbers. 
  • A lack of strategic clarity when developing a theory of change. 
  • Leads to a poor allocation of resources.
• I think a better approach is to focus on reducing suffering for as many individual animals as possible.
  • Helping millions [...]

---

Outline:

(00:19) Summary

(01:29) What's the problem?

(02:48) Why is the current framing bad?

(05:06) So what's the solution?

---

First published:

October 30th, 2024

Source:

https://forum.effectivealtruism.org/posts/z2j5gEMKyuvknLd8j/the-goal-isn-t-end-factory-farming-but-reduce-as-much

---

Narrated by TYPE III AUDIO.

This post introduces the Tech Governance Project (TGov), a new nonprofit initiative for emerging technologies risk governance in Africa. The post presents our theory of change, progress, plans, and opportunities to get involved.

Background 

Africa has traditionally been overlooked and underrepresented in global catastrophic risk and emerging technologies governance. African institutions can be significant actors in these fields, for example, due to their expanding development in biotechnology and biocontainment research activities, initiatives on AI governance, and their role in broader international affairs. However, due to insufficient investment and initiative, opportunities to advance global catastrophic risk reduction and emerging technologies governance in Africa have not been fully realized. The Tech Governance Project (Tgov), a new nonprofit initiative, seeks to build a risk-conscious field of emerging technologies governance across Africa.

Like other countries worldwide, African states also grapple with adopting and using emerging technologies. However, a widespread governance gap and [...]

---

Outline:

(00:20) Background

(03:04) Our Approach and Solution

(04:01) Theory of Change

(04:21) AI Governance

(08:53) Biorisk Governance

(12:08) Our Progress Since Our launch in June 2024

(14:16) Our Goal for the Next Seven (7) Months

(14:46) Our Structure

(15:54) Opportunities to Get Involved

The original text contained 1 footnote which was omitted from this narration.

The original text contained 1 image which was described by AI.

---

First published:

October 29th, 2024

Source:

https://forum.effectivealtruism.org/posts/AiKhGtpkWsAzwuP3s/introducing-tech-governance-project

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

Summary of key arguments

At present, there are no rules around the creation of artificial sentient beings. Anyone can create them, own them, make them do whatever they want, and treat them however they want to.

This is bad, and could lead to a lot of suffering for these beings.

The creation of artificial sentience ought not be left as an unregulated free-for-all. It ought to be regulated by governments, on behalf of society.

The moral stakes are so high that we should wait before creating artificial sentience, rather than rush into creating it, whether deliberately or accidentally. Very possibly, it may be best not to create it at all.

Responsible researchers and companies should not seek to deliberately create artificial sentience.

And governments should take steps now to prevent the creation of artificial sentience, at least in the short- and medium- term.

Even [...]

---

Outline:

(00:04) Summary of key arguments

(01:27) Introduction

(04:38) Pros and cons of different broad options for the regulation of artificial sentience

(05:33) Option 1: Free-for-all

(10:06) Option 2: Voluntary codes-of-conduct

(13:21) Option 3: Government-mandated regulation

(14:33) Option 4: Banning the creation of artificial sentience (permanently, or temporarily)

(20:50) Option 5: Banning the creation of artificial suffering

(23:11) Tactical considerations

(26:27) Conclusion

(27:53) Potential next steps

---

First published:

October 29th, 2024

Source:

https://forum.effectivealtruism.org/posts/9adaExTiSDA3o3ipL/we-should-prevent-the-creation-of-artificial-sentience

---

Narrated by TYPE III AUDIO.

This is a link post.

We've published this article on stable totalitarianism, which we rank as an "emerging challenge" on our list of pressing world problems at 80,000 Hours. Emerging challenges are potentially extremely pressing, but their nature is unclear and the solutions to them are not very developed.

We have already updated this article in response to feedback we received after publishing it. Since not a lot has been written about the topic, there's a lot more to learn, and we could easily update our views further. We welcome any additional input.

The Khmer Rouge ruled Cambodia for just four years, yet in that time they murdered about one-quarter of Cambodia's population.

Even short-lived totalitarian regimes can inflict enormous harm. In the 20th century alone, they committed some of the most horrific crimes against humanity in history. In addition to the Cambodian Genocide, these include:

• The Holocaust
• The [...]

---

Outline:

(04:18) Summary

(04:49) Our overall view

(07:37) Why might the risk of stable totalitarianism be an especially pressing problem?

(08:02) Could totalitarianism be an existential risk?

(09:11) Is any of this remotely plausible?

(12:30) Will totalitarian regimes arise in future?

(13:33) Could a totalitarian regime dominate the world?

(19:51) Could a totalitarian regime last forever?

(23:39) The chance of stable totalitarianism

(28:21) Preventing long-term totalitarianism in particular seems pretty neglected

(30:44) Why might you choose not to work on this problem?

(31:37) What can you do to help?

(32:05) AI Governance

(33:43) Researching risks of global coordination

(35:03) Working on defensive technologies

(37:13) Protecting democratic institutions

(38:49) Learn more about risks of stable totalitarianism

The original text contained 13 footnotes which were omitted from this narration.

---

First published:

October 29th, 2024

Source:

https://forum.effectivealtruism.org/posts/Ljae6jJEwifD3QCr2/stable-totalitarianism-an-overview

---

Narrated by TYPE III AUDIO.

What do I mean by “homegrown”? These projects are:

• Local: Creators have a good track record in the EA or AI Safety community
• Modest: The amount requested is not large; $5k would be meaningful
• Overlooked: Not already backed by large institutional funders like OpenPhil

If you’re a small donor or earn to give, consider giving to projects like these:

1. Feature-length documentary on SB 1047

By Michael Trazzi — $16k raised of $55k

Michael has already recorded interviews with the main characters of SB 1047: sponsors like Scott Wiener and Dan Hendrycks, proponents like Zvi Mowshowitz and Holly Elmore, and opponents like Dean Ball and Timothy B Lee. Now he needs the funding to turn it into a 1-hour feature documentary. This is a rare chance to sponsor a high-quality video narrative, and share it beyond our existing ecosystem. I’ve personally donated $10k towards this, and expect [...]

---

Outline:

(00:35) 1. Feature-length documentary on SB 1047

(01:30) 2. Overview of AI Safety in 2024

(02:31) 3. Podcast series on Effective Altruism's values

(03:05) 4. Sentinel, a foresight and emergency response team

(03:40) 5. Research on co-occurence of sparse autoencoder latents

(04:13) What else have we been up to?

The original text contained 2 images which were described by AI.

---

First published:

October 28th, 2024

Source:

https://forum.effectivealtruism.org/posts/CdHrwEiGsJqC8RAat/5-homegrown-ea-projects-seeking-small-donors

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

This is a link post.

Executive Summary

Who should read this report?

• Desk researchers attempting to Fermi estimate which wild animal interventions are cost-effective.
• Field scientists who want to know how much value of information there would be in conducting research on the abundance of wild animal populations.
• Grant evaluators who need to assess whether a proposed empirical study to estimate population size is adopting rigorous methodology.
• Animal advocates who need practical guidance on whether to implement interventions when there is considerable uncertainty about the number of potential beneficiaries.

What we did

• There is an interest in identifying cost-effective interventions to help wild animals. A major factor affecting cost-effectiveness is the number of target beneficiaries.
• We searched the academic and gray literature for abundance estimates of populations for which there are already technically feasible interventions: Urban pigeons, urban rats, and birds who collide into windows.
• For each population [...]

The original text contained 1 image which was described by AI.

---

First published:

October 28th, 2024

Source:

https://forum.effectivealtruism.org/posts/BAb5XHTDiaGnCZC6h/abundance-estimates-of-three-wild-populations

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

(Crossposted from Manifund)

Summary of the proposal

We will produce a one-hour feature documentary about SB-1047, which would:

• Serve as a comprehensive reference on the bill's history and implications.
• Present a balanced view of perspectives from both proponents and opponents, bridging ideological divides.
• Contribute to informed discussions about future AI Policy by providing in-depth, impartial analysis.
• Enhance public understanding of AI regulation challenges, fostering more informed public discourse.
• Offer policymakers, researchers, and the public a nuanced resource on various stakeholder perspectives, supporting well-informed decision-making processes.

Scott Wiener, Member of the California State Senate, SB-1047 sponsor
“Congress has not passed major tech regulation in more than a quarter century. And so in the absence of congressional action, California has a responsibility to lead.”

Funding

• We are seeking $35,000 to finish the documentary before January 2nd, 2025 (10 weeks).
• With $55,000, we would be able to finish the project in [...]

---

Outline:

(00:22) Summary of the proposal

(01:26) Funding

(02:09) Interviewees

(02:18) Proponents of the bill

(03:44) Opponents of the bill

(04:58) Project goals

(08:08) More links

The original text contained 1 image which was described by AI.

---

First published:

October 28th, 2024

Source:

https://forum.effectivealtruism.org/posts/urYYPiXW55nqRDP6k/finishing-the-sb-1047-documentary-in-6-weeks-1

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

I became the CEO of EV US in January 2023. I worked alongside the EV US and UK teams, former EV UK CEO Howie Lempel, and current EV UK CEO Rob Gledhill to recover and reform Effective Ventures and improve the robustness of the EA ecosystem in the aftermath of FTX's collapse. Amidst these efforts, I and others learned or fortified lessons that I think aren’t unique to EV and could be valuable to the wider EA community. Being able to look at hard problems, discuss them with candor, and update based on what we learn are values that I admire, and I see them as a positive and necessary mechanism for doing good. I want to act on those values here.

Goals of this post

• Provide an update to create communal knowledge: Clarify what reforms have taken place at EV in recent years and the reasoning behind [...]

---

Outline:

(00:50) Goals of this post

(01:48) Some high-level notes on FTX-related reflection

(06:04) Scope of this post

(11:53) Summary of reforms and actions taken at EV

(14:54) Background on EV

(18:42) Reforms and other actions taken

(18:46) Hiring CEOs (and other non-board personnel) for EV US and EV UK

(24:00) Changes to EV US and EV UK board

(30:04) FTX-related investigations

(38:43) Instituting financial reforms

(41:23) Improving donor due diligence

(45:36) Adopting a restrictive communications policy

(57:19) Streamlining whistleblowing policies

(59:44) Updating anti-harassment and misconduct policies

(01:03:04) Improving the COI policy

(01:06:22) Clarifying the level of separation between the EV US and EV UK entities

(01:10:44) Initiating EV shut down

(01:15:11) Some of the lessons EA can learn from the EV experience

(01:15:31) Brief summary of these lessons

(01:16:10) The lessons

(01:16:14) Organizational governance and compliance can have serious implications

(01:17:08) It's important to carefully think through and explicitly communicate your organizational risk strategy, and pay attention to it as your organization develops. This is particularly true for a fiscal sponsor

(01:19:10) EA organizations often underrate experience relative to “intelligence” and “value alignment”

(01:22:09) Vetting external counsel is important

(01:23:02) Crisis prep is underrated relative to crisis response

(01:24:43) Invest in capacity building early

(01:26:34) Communicate early (and have the resources to do so)

(01:27:31) Acknowledgments

(01:29:12) Appendix: other content about and reflections on FTX on the EA Forum

(01:29:47) Information

(01:31:24) Reflections

(01:33:18) Investigation-related

(01:33:55) Statements

The original text contained 12 footnotes which were omitted from this narration.

The original text contained 1 image which was described by AI.

---

First published:

October 28th, 2024

Source:

https://forum.effectivealtruism.org/posts/AuSah98NtR5qv8zQA/reflections-and-lessons-from-effective-ventures-1

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.