Thanks so much to everyone who took the time to play through this and provide such thoughtful feedback! I really appreciate it, and apologies for the delay in implementing these changes.

Here's what I've updated based on your suggestions:

Bug Fixes:

• URL (@BrianTan ): Thanks for flagging! I think this should be fixed globally now.
• Perching bug (@Ben Stewart ): You can now press P again to exit perching - no more getting stuck!
• Arrow keys (@Sanjay): Fixed - they now work
• Battery cage crowding (Ben): Adjusted the spacing to show the realistic density - you should now see 6-8 hens in the cage matching the "5-10 hens" description

UX Improvements:

• Upfront explanation (Sanjay): Added a "Think you know the difference?" challenge box on the welcome screen so players understand they'll be quizzed afterward
• Homepage clarity: Changed from "cage-free or pasture-raised" to "cage-free or free-range" since those are more ambiguous and better convey the challenge
• Visual cage surroundings (Ben): Enhanced the battery cage scenario - all surrounding cages now show detailed, properly-sized chickens with more subtle movement to better convey the industrial scale and crowding without as much weird flashing. (I didn't fully fix it, but I think it's a lot better)

Accuracy: Pasture-raised labeling (@david_reinstein): Added "Certified Humane" specification throughout and included a disclaimer that "pasture-raised" isn't USDA-regulated for eggs, so uncertified products may vary significantly

I do agree with several of you (Sanjay, Ben) that full gamification would make it a lot better. This just seems like a change too far for my meagre vibe-coding capabilities and limited time availability for a spare time/fun project. If someone wanted to take it this on and run with it further and make it actually good, I'd be excited about that though! I'd be happy to hand over code etc. 

The updated version is live at the same link. Thanks again for helping make this better! I also just posted on LinkedIn if anyone wants to share etc from there.

I agree with the first half of your comment. Do you think that the EA community (or the EA Forum) should solely focus on cause prioritisation though?

I feel excited about scope sensitive decision-making and evidence-based prioritisation being used at various levels of abstraction/concreteness, e.g. cause prioritisation, intervention prioritisation, organisation prioritisation. 

I welcome/encourage this being done carefully and well (and discussed on the Forum) even if I disagree with someone's prioritisation at another level of abstraction, and I don't think we should expect someone to justify their prioritisation at other levels.

I.e. I think this is a cool post, even though I don't prioritise biodiversity and the post doesn't explain why the authors prioritise it.

Seems like a cool study! I tried to check the full report (because I wasn't sure if unde stood the methodology from this summary) and FYI I got an error message: "this site can't be reached. Osf.io unexpectedly closed the connection." Maybe worth checking if the link is down 

Thank you for sharing Allegra! Welcome to the Forum, and congrats on writing and sharing this.

I think this is well written and engaging! I agree it seems a real shame for these people and for the world that the existing services have been cut. And I do think that your bullet point list suggests it's worth considering/evaluating.

I think a stronger case would delve into more detail on these claims, which aren't currently substantiated: "Sudan ranks extraordinarily high on scale, neglect, and tractability", and "Emergency networks, women's coalitions, and independent media have proven they can reach people effectively", then compare explicitly and quantitatively to other candidates for top cause areas and interventions. Is the problem area more promising than others? Are the available interventions more cost-effective? (Or in the same ballpark?)

I built an interactive chicken welfare experience - try it and let me know what you think

Ever wondered what "cage-free" actually means versus "free-range"? I just launched A Chicken's World - a 5-minute interactive game where you experience four different farming systems from an egg-laying hen's perspective, then guess which one you just lived through and how common that system is.

Reading "67 square inches per hen" is one thing, but actually trying to move around in that space is another. My hope is that the interactive format makes welfare conditions visceral in a way that statistics don't capture.

The experience includes:

• Walking through battery cage, cage-free, free-range, and pasture-raised systems
• Cost-effectiveness data based on Rethink Priorities' research on corporate campaigns
• A willingness-to-pay element leading to an optional donation to THL via Farmkind

I'd welcome feedback:

• Any factual errors I should correct? (The comparative advantage of early adopters here! Most of the fact-finding and red-teaming was done by LLMs.)
• What would make it more useful to you personally? (You'll probably give me more useful feedback this way than if you try to model other users.)
• What would make it work better as an outreach tool? (I built this with non-EA audiences in mind.)

Try it: https://achickens.world/. (Backup link here if that doesn't work.)

PS thanks Claude for the code, plus THL, RP, Farmkind for doing the actual important work; I'm just making a fun tool. This was a misc personal project, nothing to do with my employer.

Very cool! No reply needed/expected, just sharing a few misc reflections:

I didn't follow the reason for excluding Dwarkesh; you already have quality adjustments multipliers so you can just include him and apply the adjustments. (Id be interested to see, since I think it's relevant, he has an influential audience, and he has solid revenue, which I think will lead to high cost-effectiveness in your model.)

In the other direction: I'm not sure if your goal was to compare most cost-effective opportunities among current/established YouTubers, but if you're trying to evaluate the cost-effectiveness of the overall intervention instead/as well, then you'll of course need to account for efforts that started but collapsed or didn't gain much traction. 

Cool to see this analysis here though, and will likely be a useful reference/comparison for some apps that the EA Infrastructure Fund gets!

On 2., this is a kind offer! Is there some way you'd be able and comfortable sharing some of these with participants in CEA's ongoing career bootcamps?

We have a fair few software engineers, current or former C-suite types, etc. 

(Or is there some way we could connect individuals with you?)

Thanks!

The brand new episode with Kyle Fish from Anthropic (released since you wrote this comment) discusses some reasons why AI Safety and AI welfare efforts might conflict or be mutually beneficial, if you're interested!

Make your high-impact career pivot: online bootcamp (apply by Sept 14)

Many accomplished professionals want to make a bigger difference with their career, but don’t always know how to turn their skills into real-world impact.

We (the Centre for Effective Altruism) have just launched a new, free, 4-day online career bootcamp designed to help with that.

How it works:

• Runs Sept 20–21 & 27–28 (weekends) or Oct 6–9 (weekdays)
• Online, 6–8 hours/day for 4 days
• For accomplished professionals (most participants mid-career, 5+ years’ experience, but not a hard requirement)

What you’ll get:

• Evaluate your options: identify high-impact career paths that match your skills and challenge blind spots
• Build your network: meet other experienced professionals pivoting into impact-focused roles
• Feedback on CVs: draft, get feedback, and iterate on applications
• Make real progress: send applications, make introductions, or scope projects during the bootcamp itself

“It’s short, intense, and gives you a focused burst of momentum. You’ll come away with practical tools to clarify your next steps and job search direction.” — Cibeles, pilot participant

Applications take ~30 mins and close Sept 14.

If you’re interested yourself, please do apply! And if anyone comes to mind — colleagues, university friends, or others who’ve built strong skills and might be open to higher-impact work — we’d be grateful if you shared this with them.

Thanks Karen! Interested if you have specific things in mind for implications of the economic angle? I can certainly see it playing into some of the "Predict how AI will change things, and try to make that go well for animals" predictions, or leading to more of an emphasis on "Shift towards all-inclusive AI safety".