EA Forum Podcast (All audio)

Background

Ilya Sutskever is a renowned machine learning researcher who co-authored the AlexNet paper in 2012 that helped kick off the deep learning revolution. Sutskever co-founded OpenAI and served as Chief Scientist until 2024. Sutskever participated in the temporary ouster of CEO Sam Altman and subsequently left the company. He founded his own company, Safe Superintelligence, which has raised $3 billion, but about which little else is known.

Sutskever's second-ever interview with Dwarkesh Patel was released today, November 25, 2025. Here are the highlights from my point of view, along with my commentary.

Research vs. scaling

Sutskever says that (plus or minus a few years) 2012 to 2020 was an age of research, 2020 to 2025 was an age of scaling, and 2026 onward will be another age of research. This reiterates previous comments he's made. Sutskever specifically predicts that another 100x scaling of AI models would make a difference, but would not transform AI capabilities.

Self-supervised pre-training vs. reinforcement learning-based training for LLMs

Sutskever says “based on what people say on Twitter” — which is a vague, sketchy source — companies “spend more compute on RL than on pre-training at this point, because RL can actually consume quite [...]

---

Outline:

(00:16) Background

(00:59) Research vs. scaling

(01:28) Self-supervised pre-training vs. reinforcement learning-based training for LLMs

(02:42) Confusion about LLMs' lack of economic impact

(03:41) A lot of seat-of-the-pants speculation about the nature of intelligence

(05:51) The role of emotion in cognition

(08:45) Conclusion

---

First published:

November 25th, 2025

Source:

https://forum.effectivealtruism.org/posts/iuKa2iPg7vD9BdZna/highlights-from-ilya-sutskever-s-november-2025-interview

---

Narrated by TYPE III AUDIO.

The UPenn Effective Altruism group has been inactive since spring 2024. A small group and I are now reviving it!

I'm Hazem, a freshman in Mechanical Engineering at UPenn and one of two current organizers.

Current status:

• 2 organizers, 2 advisors, ~10 interested non-organizing members
• 3 socials held
• access restored to club infrastructure (website, logins, PennClubs.com, etc.)
• starting to pursue external support (i.e. CEA)

If you're at UPenn and are up for long-term board/organizing work (expect 4-10 hrs/week), PLEASE reach out.

Also very helpful:
- pointers to EA-familiar students at UPenn
- advice for reviving uni EA groups

You can DM me here or email [email protected].

---

First published:

November 26th, 2025

Source:

https://forum.effectivealtruism.org/posts/CYBJ8pTD2WMMQ96jS/help-revive-ea-at-upenn-looking-for-board-members

---

Narrated by TYPE III AUDIO.

The UPenn Effective Altruism group has been inactive since spring 2024. A small group and I are now reviving it!

I'm Hazem, a freshman in Mechanical Engineering at UPenn and one of two current organizers.

Current status:

• 2 organizers, 2 advisors, ~10 interested non-organizing members
• 3 socials held
• access restored to club infrastructure (website, logins, PennClubs.com, etc.)
• starting to pursue external support (i.e. CEA)

If you're at UPenn and are up for long-term board/organizing work (expect 4-10 hrs/week), PLEASE reach out.

Also very helpful:
- pointers to EA-familiar students at UPenn
- advice for reviving uni EA groups

You can DM me here or email [email protected].

---

First published:

November 26th, 2025

Source:

https://forum.effectivealtruism.org/posts/CYBJ8pTD2WMMQ96jS/help-revive-ea-at-upenn-looking-for-board-members

---

Narrated by TYPE III AUDIO.

Johannes Haushofer and Richard Nerland

Summary

Malengo is a charity that facilitates international educational migration in the pursuit of poverty alleviation. Its flagship program supports Ugandan high school graduates in applying for and completing a Bachelor's degree in Germany. Four years into Malengo's operations and two years following GiveWell's initial cost-effectiveness analysis, we present early results from our ongoing Randomized Controlled Trial (RCT), and an updated estimate of our impact based on these results. Early RCT results suggest that program beneficiaries experience a USD 843 increase in monthly labor income already in their first year in the program, adjusted for purchasing power. They also report substantially improved psychological well-being. There is no evidence of negative economic or psychological spillovers on parents and siblings at home; in fact, parents also report a substantial increase in psychological well-being. Updated program metrics suggest higher retention rates than originally assumed, with only 2 students out of 250 having returned home so far. We use these results to provide an updated Back-of-the-Envelope Calculation (BOTEC) in the style of GiveWell. The program is 28 times more effective than cash transfers in a baseline scenario, and 3,732 times in an optimistic scenario.

Introduction

Malengo was [...]

---

Outline:

(00:15) Summary

(01:40) Introduction

(03:46) Early RCT Results

(08:04) Ongoing Survey Results

(11:52) Impact Model Updates

(26:58) Relation to GiveWell's updated estimates of GiveDirectly's effectiveness

(28:19) Outlook

---

First published:

November 26th, 2025

Source:

https://forum.effectivealtruism.org/posts/CpZYHkqLKAmLZSRbz/malengo-impact-model-update-fall-2025

---

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.

AI In Group Discourse

I wanted to programmatically analyze the AI in-group ecosystem and discourse using AI as an exploration in sensemaking during my time in the “AI for Human Reasoning” FLF fellowship.

I also analyzed the EA Forum itself in an earlier fellowship sensemaking MVP.

Sidenote: I’m very interested in the potential for “AI uplift” to increase the impact of EA-aligned orgs by removing bottlenecks, automating flows, processing unstructured data, etc. I’ve founded a successful B2B SaaS company in the past and am now looking to pivot to high impact work using my skills. If you have ideas, questions, or may need some consulting / contracting work, please DM me.

This first thing I I needed to do was find the high authority network of people talking about AI to find the relevant information and sources they posted / shared.

I started with a hand picked seed of ~50 twitter accounts I considered firmly focused on high quality AI discourse and then used twitter following data to create an ever expanding following / authority / page-rank graph.

I also used AI to analyze the past 1000 tweets for each account to score how relevant their profile was to [...]

---

Outline:

(00:14) AI In Group Discourse

(02:39) en-US-AvaMultilingualNeural__ Leaderboard showing Top 41 AI Organizations - Authority Leaderboard with rankings and scores.

(03:10) Worldview Clusters

(08:35) en-US-AvaMultilingualNeural__ Chat interface discussing AI existential risk with doom scenario framing.

(09:07) Popular domains

(11:12) AI Trajectory Analysis

(17:32) Bluesky vs Truth Social

(18:02) Pipeline

(19:44) Results

---

First published:

November 26th, 2025

Source:

https://forum.effectivealtruism.org/posts/JxiEyRtaApvqAZ2cQ/ai-discourse-analyzed-we-looked-at-essays-twitter-bluesky

---

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.

Cross-posted from the @High Impact Professionals blog - see the original here.

--

In my years of working with people on their financial plans for giving, I’ve met many wonderful individuals who feel morally compelled to take a giving pledge. Intellectually, they are fully aware that they are in a position to help others. But emotionally, they aren’t ready to commit to a pledge.

The sticking point here is a common thought many of us have: “Do I have enough?”

• Do I have enough to keep my 🔸10% Pledge now that we’re having a baby?
• Do I have enough to take the 🔹Trial Pledge when I have student loans to pay off?
• Do I have enough to increase my pledge if I’m moving to a high-cost-of-living city?

In my coaching program for aspiring philanthropists, it's my job to help people explore this question using their very own financial data.

For many of us, the root of “Do I have enough?” comes from a scarcity mindset: we focus on our limited resources rather than our abundance. I very much understand where these fears come from. When we compare ourselves to peers who earn more or dwell on goals we [...]

---

Outline:

(01:27) 1. How rich are you?

(02:45) 2. Build a true financial awareness

(06:29) 3. Uncover your spending values

(08:38) 4. Challenge yourself

(09:51) 5. Build confidence by reframing

(10:55) About Me

---

First published:

November 25th, 2025

Source:

https://forum.effectivealtruism.org/posts/zegJkmZAxFFQwSWsG/scared-to-pledge-5-financial-steps-for-confident-giving

---

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 keep thinking about what kind of identity would be useful for building a powerful animal advocacy movement. Here are 3 features of veganism that I often think about which make me doubt its usefulness.

Too maximalist

The official definition of veganism by the inventors of the term is the following:

“Veganism is a philosophy and way of living which seeks to exclude—as far as is possible and practicable—all forms of exploitation of, and cruelty to, animals for food, clothing or any other purpose”

This basically amounts to "avoid doing bad things as far as possible." The threshold sits right below what is impossible. I think that is way too ambitious. Doing the best possible thing at every circumstance shouldn’t be the criterion for inclusion to a social movement. We don't expect human rights activists to avoid all forms of exploitation and cruelty as far as possible to qualify as human rights activists.

Some activists respond "No, veganism is the bare minimum. The 'as far as possible and practicable' part means it's not about being perfect.". But when I ask for examples of gratuitously harmful actions that veganism doesn't forbid, at most I hear about instances of accidental uses [...]

---

Outline:

(00:22) Too maximalist

(03:37) No space for believers to sin

(04:29) Too behaviour-focused

---

First published:

November 26th, 2025

Source:

https://forum.effectivealtruism.org/posts/BX8hPeye2QRcyftRk/3-doubts-about-veganism

---

Narrated by TYPE III AUDIO.

TLDR: Just pick up the phone. Agree on permissions for information before you share it (see Step 4).

I have worked as a professional journalist covering AI for over a year, and during that time, multiple people working in AI safety have asked me for advice on engaging with journalists. At this point, I've converged on some core lessons, so I figured I should share them more widely.

I've also reviewed some of the prior art about how to talk to journalists on LessWrong and found it unsatisfying. The answer to the question often seems to be "Don't."[1]

Unsurprisingly then, I think many people feel like they are not prepared to talk to journalists. They often have a sense that they are waiting to receive some arcane knowledge that would equip them for such a treacherous conversation, the sort of knowledge acquired at ritual ceremonies called "media trainings."

In fact, there is very little such secret knowledge. It's easy to talk to journalists—it mostly involves picking up the phone—and the most important lessons can fit in a quick post. This post provides those lessons so you no longer have any excuses for not talking to journalists if you want [...]

---

Outline:

(01:31) Before the Call

(01:34) Step 1: Understand the Interaction

(02:40) Step 2: Research Your Target

(03:31) Step 3: Sus Out the Topics

(03:50) Step 4: Set Permissions

(05:53) Changing Permissions

(06:34) On Background

(08:11) Deep Background

(08:59) Off The Record

(09:50) Anonymity

(10:45) Embargoes

(11:39) Step 5: Figure Out Key Points

(12:37) During the Call

(12:41) Step 6: Be Nosy

(14:11) Step 7: Hammer Your Key Points

(15:27) Step 8: Defense Against the Dark Arts

(18:01) After the Call

(18:05) Step 9: Follow Up

(18:26) Step 10: Keep in Touch

(18:59) Whats Next

---

First published:

November 25th, 2025

Source:

https://forum.effectivealtruism.org/posts/Ax7sMzZxsvqLgzdTn/how-to-talk-to-journalists-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.

Overview

The Unjournal is a 501(c)(3) that commissions rigorous public expert evaluations of impactful research. We've built a strong team, completed more than 55 evaluation packages, built a database of impactful research, launched a Pivotal Questions initiative, and are systematically evaluating research from EA-aligned organizations

We argue that

1. EA-driven research would benefit from external expert scrutiny to improve decision-making.

2. Academic research contains valuable insights but needs steering toward impact and usefulness.

3. The Unjournal bridges this gap through rigorous public evaluations of both EA-driven and academic research and ‘pivotal questions’. Our approach is more transparent and efficient than traditional academic journal peer review, but more rigorous and structured than crowdsourcing feedback, e.g., on the EA Forum.

4. We’re having some success. Our work has contributed to methodology and funding considerations at GiveWell, brought senior and prominent academics into the global priorities conversation, shaped animal welfare funding priorities, and provided quality control for EA organizations that use and produce research.

The Unjournal commissions public evaluation packages: referee-style assessments, quantitative ratings, and "merited journal tier" recommendations. See our output and ratings dashboard, and claim assessment.

We're at an inflection point. Future grant funding will determine whether we can [...]

---

Outline:

(00:17) Overview

(00:53) We argue that

(02:45) Why academic rigor matters for EA (and EA for academia)

(02:52) The problem were addressing

(03:31) Some examples

(03:35) Cultured meat forecasts

(04:03) Water Treatment

(05:06) Promoting methodological rigor -  when can we trust research cited on this forum?

(06:25) Pivotal methodological questions

(07:02) Nudging academia towards impact

(07:38) Our approach

(09:01) Funding

(09:04) Current Funding Situation

(09:49) Using Marginal Funding

(11:27) What were building: Infrastructure for better decision-making

(12:38) Our theory of change

(12:50) How to support and engage with us

---

First published:

November 21st, 2025

Source:

https://forum.effectivealtruism.org/posts/FaBBakJn8ZKvC2HSn/the-unjournal-bridging-the-rigor-impact-gaps-for-ea-relevant

---

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.

Crosspost.

(I think this is a pretty important post to get the word out about, so I’d really appreciate you restacking it).

The EU is taking input into their farm animal welfare policies until December 12. Animals in the EU are kept confined in a space about the size of a piece of paper, never able to express their natural behaviors or move around much, constantly standing against wire meshing. They suffer constantly because of inadequate animal welfare laws. These animals deserve better than a brief and hellish life of the kind currently afforded to them.

Usually, a shockingly small number of people submit input, which means that by providing feedback (which you can do whether or not you’re an EU citizen), you can be one of a relatively small number of people impacting decisions which affect hundreds of millions of farm animals. Specifically, you can influence whether hundreds of millions of animals have to languish in a tiny cage, and whether baby male chicks can be ground up alive right after birth.

Filling out the questionnaire should take only a few minutes. If every one of you filled it out, then my blog readers personally would make [...]

---

First published:

November 21st, 2025

Source:

https://forum.effectivealtruism.org/posts/Fa4SdaMBDTweEDnLi/urgent-easy-opportunity-to-help-many-animals

---

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.