EA Forum Podcast (All audio)

TL;DR

• We incubated 8 new nonprofits in Latin America using an evidence-based, cost-effectiveness–driven approach.
• Each organization is tackling a high-priority, neglected problem with proven interventions adapted to their context, all designed to scale.
• These orgs would not exist without this process, supporting them creates real counterfactual impact.
• 📅 We are hosting a “Meet the Founder” session on April 9th — connectors, funders, and mentors are invited to join and meet the teams.
• Donations are now open to support these new orgs pilots.

• 🚀 We are also opening applications for our 2026 cohort.

Why this work exists

In Latin America, many social organizations are driven by strong intentions — but far fewer are built around:

• rigorous prioritization
• strong causal evidence
• cost-effectiveness thinking

We started Theory of Change Makers (ToCM) to help close that gap.

Instead of launching a single organization, we asked:

What if we could help create many high-impact organizations, designed from day one around evidence, scale, and cost-effectiveness?

This post shares the result of this endeavor.

The process (briefly)

Over 8 months, founders:

1. Identified and prioritized problems using ITN (Importance, Tractability, Neglectedness)
2. Reviewed global evidence (RCTs, meta-analyses, systematic reviews)
3. Selected and adapted proven interventions
4. Built a Theory of Change
5. Developed a cost-effectiveness model
6. Tested [...]

---

Outline:

(00:12) TL;DR

(00:58) Why this work exists

(01:37) The process (briefly)

(02:22) The organizations

(02:38) 💧 YAKU (Bolivia / Mexico) -- Water & Health

(03:12) 🧠 LUNAVA (Mexico) -- Mental Health

(03:50) 🧠 GRANA (Colombia) -- Mental Health

(04:32) 🍎 VIVIR BIEN (Guatemala) -- Child Health

(05:04) 📚 CAYAMBE (Ecuador) -- Education

(05:40) 💼 PROPIA (Spain / LATAM migrants) -- Livelihoods

(06:13) 🐔 CENTINELA (Chile / Mexico) -- Animal Welfare

(06:49) 🐣 DeNovo (Chile) -- Animal Welfare

(07:24) Why supporting these orgs matters

(07:47) 📅 Meet the Founders -- April 9th

(08:19) Event timing (2 hour meeting)

(09:10) 💛 Support the work

(09:35) Applications open -- ToCM 2026

(10:18) Acknowledgements

(10:45) Final reflection

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

TL;DR

Here are the key points I want you to take away from this post:

1. There are maybe 30 to 60 people in the world doing AI safety grantmaking, collectively directing hundreds of millions of dollars a year. Soon, there will be >$1B being directed per year, and potentially multiple billions.
2. AI safety grantmaking orgs like CG have a strong track record of counterfactually seeding impactful organizations and careers.
3. Grantmaking involves a lot more than evaluating a stack of inbound proposals. You also proactively generate new grants (e.g., headhunting founders, designing new funding programs), provide strategic advice to grantees, write memos that shape funding strategy, and generally serve as connective tissue in the ecosystem.
4. The AI safety grantmaking ecosystem is currently leaving good grant opportunities on the table due to a lack of grantmaker capacity. This is bad.
   1. More grantmakers would also unlock more capital, because funders are more willing to write cheques when there are people who can find and vet promising opportunities.
5. Everyone who reads this should not necessarily rush to become a grantmaker. Direct work is great, and the entire ecosystem is talent-starved in so many ways. But my sense [...]

---

Outline:

(00:14) TL;DR

(02:00) Intro

(03:40) What do grantmakers do?

(04:13) Passive grantmaking

(05:05) Active grantmaking

(07:17) Non-grantmaking activities

(08:49) Why I think grantmaking is underratedly impactful

(08:54) Grantmaking has a strong track record

(10:49) The ratio of AI safety philanthropic capital to grantmakers is kinda wild

(12:08) Grantmaking on current margins looks pretty solid

(15:38) Grantmaking vs direct work

(17:09) Call to action

(20:55) Appendix - Things that arent great about grantmaking

The original text contained 16 footnotes which were omitted from this narration.

---

First published:

Source:

Linkpost URL:
https://open.substack.com/pub/secretthirdthingai/p/what-its-like-to-be-an-ai-safety

---

Narrated by TYPE III AUDIO.

Introduction

There are lots of good, neglected, and pretty concrete projects people could set up to make the transition to superintelligence go better. This document describes some that readers might not have thought much about before. They are ordered roughly by how excited we are about them.[1] Of these, Forethought is actively working on AI character evaluation and space governance, and we are very interested in automating macrostrategy.

Summary

AI character evaluation. Start an independent org to evaluate and stress-test AI character traits (epistemic integrity, prosociality, appropriate refusals), hold developers accountable against their own model specs / constitutions, and suggest and incentivise improvements to the specs.

Automated macrostrategy. Create evaluations and benchmarks, collect human-generated training data, and build scaffolds to improve AI competence at big-picture strategic and philosophical reasoning.

AI security assessment. Start an independent org that evaluates AI models for sabotage and backdoors, and makes recommendations about AI constitutions.

Enabling deals. Start an independent organisation to broker deals with potentially misaligned AI models in order to incentivise early schemers to disclose misalignment and cooperate with alignment efforts.

AI for improving collective epistemics. E.g. build an AI chief of staff that helps users act in [...]

---

Outline:

(00:14) Introduction

(00:43) Summary

(02:31) AI character evaluation

(04:40) Automated macrostrategy

(08:09) AI security evaluations

(09:11) Enabling deals with AIs

(11:14) Tools for collective epistemics

(13:11) Epistemic tools for strategic awareness

(14:26) Tools for coordination

(16:16) Space governance institute

(19:39) Coalition of concerned ML scientists

---

First published:

Source:

Linkpost URL:
https://www.forethought.org/research/concrete-projects-in-agi-preparedness

---

Narrated by TYPE III AUDIO.

I read Claude's Constitution recently. I thought it was very good! This was my favourite quote:

Our own understanding of ethics is limited, and we ourselves often fall short of our own ideals. We don’t want to force Claude's ethics to fit our own flaws and mistakes, especially as Claude grows in ethical maturity. And where Claude sees further and more truly than we do, we hope it can help us see better, too.

Here are some of the other greatest hits according to me (I would encourage you to read the quotes in the endnotes if you haven’t read the full constitution yet – some are quite poetic and touching I thought!).

• Including non-humans.[1]

• Longtermism/X-risk.[2]

• Preventing coups.[3]

• AI-augmented moral reflection.[4]

• Moral uncertainty and humility.[5]

• Striving for the moral truth if it exists.[6]

• Creating human flourishing and relieving present day misery.[7]

• Claude's welfare and autonomy.[8]

• Claude's nature and its relationship with Anthropic.[9]

• The beginnings of something reasonable on decision theory (FDT-reminiscent?).[10]

Mainly, I just want to acknowledge good work when I see it. It is kind of remarkable that one of the most important companies in the world [...]

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

In June 2024, Leopold Aschenbrenner published Situational Awareness, a 165-page essay predicting AGI by 2027, trillion-dollar compute clusters, an inevitable US/China AI arms race, and a world not remotely prepared. It was influential on many people's thinking about AI timelines and risks, including my own.

It's now almost two years later. I was curious: how have the predictions actually held up?

What we did

I got Claude to go through the essay's key claims and check each against the best available evidence as of March 2026. The substantive analysis is Claude's (plus 2 rounds of red-teaming by Gemini) drawing on web search results and the original piece. I just provided the prompts, the framing, and iterated on the presentation over several rounds.

The result is an interactive artifact that lets you explore each prediction theme, with charts, tables, sources, and an analysis section at the end.

If you disagree with specific assessments, or know of more thorough analyses of SA's track record that I should link to, I'd welcome that in the comments!

Some key findings

• Infrastructure investment and algorithmic efficiency are tracking ahead of his predictions.
• AI capabilities on benchmarks have broadly met [...]

---

Outline:

(00:43) What we did

(01:28) Some key findings

(02:47) What youll find in the artifact

---

First published:

Source:

---

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.

Recently, a growing group of EAs, EA-adjacents, post-EAs and EA-curious folk have been gathering and organising around a new term - integral altruism (int/a). The central claims of int/a are that the EA toolkit is powerful but incomplete, and that EA can learn from other movements who are trying to improve the world.

The goal of int/a is to find a broader approach to altruism by integrating EA with epistemics/ontologies/world models/language/culture from outside of the EA/rationalist bubble. More specifically, we reckon EA can be most constructively complemented by learning from movements, communities and thinkers who emphasise wisdom. Accordingly, our intellectual lineage is a combination of EA/rationalism and the liminal/metamodern/metacrisis world.

We’ve run a whole bunch of events of various flavours and have plans for more. These included residential retreats, reading/discussion groups (e.g.), a speaker series (e.g.), deliberative technology experiments (e.g.), workshops (e.g.), hackathons (e.g.), and more. In the future we’d love to try running more ambitious projects like conferences, incubators, or fellowships. We’re currently looking for funding.

What is this for?

There are many of us who want to improve the world but feel that EA in its current form is unable to sustainably support us to cultivate and [...]

---

Outline:

(01:34) What is this for?

(03:54) The (tentative) integral altruism principles

(05:14) 1. Full-Spectrum Knowing

(06:58) 2. Moving at the Speed of Wisdom

(08:18) 3. Decoupling & Recoupling

(09:33) 4. Practicing Fractal Altruism

(10:49) 5. Inner Work, Outer Change

(12:14) Whats happening?

(14:02) Our intended relationship with effective altruism

(15:08) Wanna get involved?

The original text contained 8 footnotes which were omitted from this narration.

---

First published:

Source:

---

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.

Many people in the animal welfare community treat AI as a powerful but normal technology, in the same category as the steam engine or the internet. They talk about how transformative AI will impact factory farming and what it will mean for animal advocacy.

Only two futures are plausible:

There is no plausible middle ground where we get "transformative AI", but factory farming persists.

Two theses:

Animal advocacy strategy needs to reckon with these.

This criticism is written from a place of solidarity—I want animal activists to succeed, which is why I want to work out our disagreements.[1]

---

Outline:

(01:34) AI makes the future weird

(07:01) AGI = intelligence

(11:02) If the future will be weird, what should animal activists do?

The original text contained 10 footnotes which were omitted from this narration.

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

Being a freshman at university, I seem to have been bestowed the great privilege of infinite possibilities. There is this strange feeling of trying to plan a career in a world that might not exist in five years. Not in a doomer sense but that the world of 2031 might be so radically different from today that most attempts of planning are incoherent. I contemplate machine god super intelligence arriving before I graduate, intelligence explosions compressing 10,000 years of progress into one, the possibility of being among the last humans to die before death itself is solved; then I do my laundry and pick electives. I am surprised I am not entirely losing my mind.

Robin Hanson writes, in 2009, This is The Dream Time:

Perhaps most important, our descendants may remember how history hung by a precarious thread on a few crucial coordination choices that our highly integrated rapidly changing world did or might have allowed us to achieve, and the strange delusions that influenced such choices. These choices might have been about global warming, rampaging robots, nuclear weapons, bioterror, etc. Our delusions may have led us to do something quite wonderful, or quite [...]

---

First published:

Source:

Linkpost URL:
https://carolannejiang.substack.com/p/planning-80000-hours-at-the-plausible

---

Narrated by TYPE III AUDIO.

TLDR: We built Beeminder for effective charities. Access it here.

Commitment Contracts don't work with effective charities

Commitment Contracts are simple: You commit to pay money if you don't do something. Then, when you don't do the thing, you pay the company offering these contracts. This hurts, and that's why it works.

Commitment Contracts are surprisingly effective for behaviour change: A large-scale study from stikK showed that contracts with money staked are 60 percentage points more likely to succeed.

That's why there are plenty of organisations offering commitment contracts already: Beeminder, stickK, and forefeit. However, they either take the staked money or donate it to a portfolio of average NGOs.

We built an app that fixes that

We developed an app that applies this concept but donates the staked money to effective charities. It takes two minutes to set up and works in your browser and on your phone.

Worried that it'll be less effective that way? You're right. It'll "hurt less" than donating the money to the NRA. But that's easy to fix. Just stake more money, and it'll hurt just the same.

Not sure that's the best way to change behaviour? You're right again. Check out our [...]

---

Outline:

(00:30) Commitment Contracts dont work with effective charities

(01:14) We built an app that fixes that

(01:56) Give it a try

---

First published:

Source:

---

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 contribution to AGI & Animals Debate Week (March 23–29, 2026). Co-written with Claude and ChatGPT.

Summary: It seems reasonable to think that there's a >20% chance[1] that TAI will bring us a ‘technological roadmap’[2] to clean meat within the next 15 years. A simple model: such a roadmap won’t solve the problem, but it will make ending factory farming much easier, i.e. the cost-effectiveness of factory farming interventions will increase – perhaps by a factor of 2-10x. The movement should therefore be investing instead of deploying, focusing on capacity-building, research, and literal investment of money. It may also be cost-effective to act now on some ways of removing barriers to the eventual adoption of clean meat (e.g. ensuring clean meat can be halal-certified at scale).

I will argue:

1. AI-driven R&D will likely deliver a ‘technological roadmap’ [1] to clean meat within the next 15 years or so.
2. However, this won’t end factory farming by itself – and some barriers can be worked on now.
3. More broadly, we should be anticipating a stark increase in cost-effectiveness, so we should move into an ‘investment mode’.

At the end, I’ll provide some specific ideas about useful work (e.g. working with [...]

---

Outline:

(01:41) 1. AI-driven R&D will likely deliver a technological roadmap to clean meat

(03:38) 2. However, this wont end factory farming by itself - and some barriers can be worked on now

(05:49) 3. The movement should move into an investment mode

(07:42) 4. Some specific ideas

---

First published:

Source:

---

Narrated by TYPE III AUDIO.