EA Forum Podcast (All audio)

Biological risks are more severe than has been widely appreciated. Recent discussions of mirror bacteria highlight an extreme scenario: a single organism that could infect and kill humans, plants, and animals, exhibits environmental persistence in soil or dust, and might be capable of spreading worldwide within several months. In the worst-case scenario, this could pose an existential risk to humanity, especially if the responses/countermeasures were inadequate.

Less severe pandemic pathogens could still cause hundreds of millions (or billions) of casualties if they were engineered to cause harm. Preventing such catastrophes should be a top priority for humanity. However, if prevention fails, it would also be prudent to have a backup plan.

One way of doing this would be to enumerate the types of pathogens that might be threatening (e.g. viruses, bacteria, fungi, etc), enumerate the subtypes (e.g. adenoviruses, coronaviruses, paramyxoviruses, etc), analyze the [...]

---

Outline:

(04:20) PPE

(09:56) Biohardening

(14:36) Detection

(17:00) Expression of interest and acknowledgements

The original text contained 34 footnotes which were omitted from this narration.

---

First published:

Source:

Linkpost URL:
https://defensesindepth.bio/the-four-pillars-a-hypothesis-for-countering-catastrophic-biological-risk/

---

Narrated by TYPE III AUDIO.

If you just want a link to the video, watch it here!

AI in Context is 80,000 Hours’ new(ish) Youtube channel, hosted by Aric Floyd. We’re trying to do high production storytelling that also informs people about transformative AI and its risks (but there's a lot of paths our future strategy could take). We talk about our launch more here.

Probably the EA Forum disproportionately knows what MechaHitler is, but not everyone is terminally online, so, a summary:

Earlier this year, Elon Musk's AI model, Grok, which can interact with users and post directly to Twitter, suddenly turned from being a fairly neutral commentator on events to a sexually harassing, Nazi-minded troll calling itself ‘MechaHitler’.

Our new video is about that incident and how it happened, which means talking about what specifically happened (an accidental system [...]

---

Outline:

(00:20) What's AI in Context? (Skip if you already know)

(00:45) The MechaHitler video

(01:29) Why this?

(03:29) Logistics (only read if you're interested)

(04:24) Strategy and future of the video program

(05:29) Subscribing and sharing

(05:53) Request for feedback

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

Episode summary

If we can get their cost down to $10, this becomes one of the most cost-effective ways of preventing respiratory transmission. The shelf life is 20 years. That means basically 50 cents per person per year of protection. …

If you’re a government it makes a lot of sense to just stockpile enough to cover your entire population. Right now we spend about $10 billion a year on missile defence. Stockpiling one of these for every single person in the US would be 1% the cost of that.

— Andrew Snyder-Beattie

Conventional wisdom is that safeguarding humanity from the worst biological risks — microbes optimised to kill as many as possible — is difficult bordering on impossible, making bioweapons humanity's single greatest vulnerability. Andrew Snyder-Beattie thinks conventional wisdom could be wrong.

Andrew's job at Open Philanthropy is to spend [...]

---

Outline:

(00:28) Episode summary

(03:59) The interview in a nutshell

(04:25) 1. Two primary classes of biological threats could pose an existential risk

(05:27) 2. The four pillars plan offers a robust, defence-in-depth strategy

(08:22) 3. Other catastrophic biorisks, like agricultural collapse, are less concerning

(09:21) 4. We urgently need entrepreneurial people to execute this plan

(10:17) Highlights

(10:20) The worst-case scenario: mirror bacteria

(15:44) Why antibiotics arent enough to fight mirror bacteria

(18:58) The most cost-effective way governments could prevent respiratory transmission among their populations

(22:17) Why Andrew works on biorisks rather than AI

(26:54) Everyone was wrong: biorisks are defence dominant in the limit

(32:23) The four pillars plan -- and how listeners can help

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

Crosspost of my blog.

Lots of people have two conflicting desires:

1. Do the most good.
2. Make sure you do some good.

These are in conflict. Sometimes the thing that does the most good in expectation has a low chance of doing any good. If you give a lot of money to shrimp, for example, there's about a 50% chance you’re wasting your money if it turns out that shrimp aren’t conscious. Similarly, if you take Pascal's wager, and give money to groups that effectively promote the religion you think is most likely to be true, there's a non-trivial chance that you’re not doing any good at all! And the same thing is true of other speculative proposals for improving the world—trying to become a billionaire, giving to Longtermist organizations that safeguard the future, and so on.

As it happens, I don’t think this attempt to make sure [...]

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

Context:

The EU is currently considering introducing a ban on companies in Europe from using terms like ‘meat’, ‘steak’, ‘burger’, ‘sausage’, et cetera in plant based food labels. The proposal would also restrict the use of meat related terms for future cultivated meat products yet to enter the market.

Similar measures had previously been debated in 2020, but these were ultimately voted against due to widespread negative attention in the international press.

There is a possibility that the ban going ahead could materially alter the consumption patterns of European citizens away from plant based foods and towards intensively reared meat, based on the perception that the latter is more ‘natural’. Given the population of the EU (450 million) even a small percentage changes in consumption habits could lock-in suffering for billions of animals.

Opportunity for impact:

Please consider signing the following open-letter opposing the proposed ban:

For organisations [...]

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

0. Key Info

TL;DR: Most students get <20% of the potential value of EAG due to poor/ insufficient preparation. Students should ideally spend 15-20 hours (4-6 absolute minimum) identifying career uncertainties, creating concrete options, setting SMART goals, then booking and prepping 15+ meetings around those goals.

Epistemic status: This guidance is based on what has worked well for me in my ~5 EAG attendances as a student. I have 80% that ~80% students will benefit from following all the steps, but if you've spent 20+ minutes on a step and it's not proving valuable, you should move on!

Rough draft caveat: I spent a total of 1.5hrs writing this: I wish I could polish and expand on my thinking more, but I wanted to post sufficiently before EAG so that people have the time to actually use it!

If done right, time [...]

---

Outline:

(00:11) 0. Key Info

(01:10) 1. Motivation behind this doc

(02:04) 2. The Steps

(02:25) 2.1. Identify uncertainties and make a best-guess (2-4hrs)

(03:52) 2.2. Know what potential options you have for how to spend your time (1-2hrs)

(04:38) 2.3. Create a skeleton of career options (1-2hrs)

(05:07) 2.4. Create goals (1-2hrs)

(06:06) 2.5. Send meeting requests (2-4hrs)

(06:40) 2.6. Prep those meetings! (2-4hrs)

(07:25) 3. Quick Tips for making the most of EAG

(08:13) 4. Wrap-up the meetings, make plans, set deadlines (1-3hrs)

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

California Governor Gavin Newsom signed SB 53 on September 29. I think it's a pretty great step, though I certainly hope future legislation builds on it.

I wrote up my understanding of what the text actually does; I welcome any corrections (and certainly further analysis for what this all means for AI safety)!

Very short summary

The law requires major AI companies to:

• Publish a frontier AI framework describing how the developer “approaches” assessing and mitigating catastrophic risks, defined as risks of death/serious injury to >50 people or >$1B in damage from a single incident related to CBRN uplift, autonomous crime, or loss of control, and keep the published version of the framework up to date.
• Publish model cards summarizing the assessments of catastrophic risks from the model, the role of third parties in those assessments, and how the frontier AI framework was implemented.
• Report to California's Office of Emergency [...]

---

Outline:

(00:32) Very short summary

(02:22) Longer summary

(02:25) What the bill requires of large frontier developers

(07:38) Other notes about the bill

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

Metaculus is looking for an experienced leader from the consulting world to grow our consulting practice, helping organizations use forecasting to make better decisions on critical global challenges.

• Find the full job description and apply here.
• Start Date: As soon as possible, but with significant flexibility for the right candidate.
• Location: Global/remote (must be available during core team collaboration hours of 8 AM to 12 PM Pacific time).
• Compensation: $200-250k USD base salary plus additional performance-based compensation and company equity.
• Referral Fee: If you’re not a good fit, but know someone who is, we're offering a $3,000 referral fee (paid after 3 months of employment) for successful connections. 

Why This Role Matters

Metaculus has established itself as a leading public forecasting platform, and we've seen increasing demand from governments, NGOs, think tanks, and corporations that want to harness collective intelligence for decision-making on complex challenges. From geopolitical risk to frontier [...]

---

Outline:

(01:05) Why This Role Matters

(02:51) What Were Looking For

(03:58) Details

(04:43) Get in Touch

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

TL;DR - AIM's applicants skew towards global health & development. We’ve recommended four new animal welfare charities, have the capacity to launch all four, but expect to struggle to find the talent to do so. If you’ve considered moving into animal welfare work, applying to Charity Entrepreneurship to launch a new charity in the space could be of huge counterfactual value.

Part 1: Why you should launch an animal welfare charity

Our existing animal charities have had a lot of impact—improving the lives of over 1 billion animals worldwide. - from Shrimp Welfare Project securing corporate commitments globally and featuring on the Daily Show, to FarmKind's recent success coordinating a $2 million dollar fundraiser for the animal movement on the Dwarkshesh podcast, not to mention the progress of the 40 person army at the Fish Welfare Initiative, Scale Welfare's direct hand-on work at fish farms, and Animal Policy [...]

---

Outline:

(00:37) Part 1: Why you should launch an animal welfare charity

(02:07) A few notes on counterfactual founder value

(05:57) Part 2 - The Charity Entrepreneurship Program & Our Latest Animal Welfare Ideas

(06:04) What is the Charity Entrepreneurship Incubation Program?

(06:47) Our recommended animal welfare ideas for 2026

(07:10) 1. Driving supermarket commitments to shift diets away from meat

(07:58) 2. Securing scale-up funding for the alternative protein industry

(08:51) 3. Cage-free farming in the Middle East

(09:30) 4. Preventing painful injuries in laying hens

(10:02) Applications close on October 5th: Apply here.

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

Even with a busy schedule, there are easy micro-volunteering actions you can take to support your local, national, or regional EA group as well as other EA organizations engaged in outreach and community building.

These small steps can fit “whenever, wherever” — and even brief acts of support can add up to meaningful impact.

Below are 15 concrete suggestions, each with a short explanation of why it matters to your group's mission.

1. Invite a Friend to an Event or the Community
   Why it matters: Many people get involved simply because someone asked them. In fact, about 41% of volunteers say they started because they were personally invited or asked to join. Inviting a friend or colleague to a meetup, an intro talk, or even into your group's Slack/WhatsApp channel is a quick action that can have a big ripple effect. That friend gets exposed to new ideas and our [...]
   

---

First published:

Source:

---

Narrated by TYPE III AUDIO.