Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: ChatGPT struggles to respond to the real world, published by Alex Flint on January 12, 2023 on The AI Alignment Forum.

I wanted to test whether ChatGPT could respond to situations as they arise in the real world, so I asked it to guide me through making a turmeric latte. It had no problem giving me a nicely formatted recipe but it really struggled to guide me through the process of making the latte.

The rules I set for this experiment was that I would perform any instruction that I could perform in less than 10 seconds, so "fetch a saucepan" is fine, "go to the fridge" is fine, "put one cup of milk into the saucepan" is fine if I have the milk and the saucepan already in front of me, but if the milk or saucepan were not in front of me I would respond with something like "I don’t have the milk in front of me. Can you guide me through finding some milk?"

I insisted that I be given just one next instruction, not a list of numbered instructions, unless I could perform the whole set in less than 10 seconds total. The reason is that I wanted to test whether ChatGPT could respond to difficulties that arose organically in the process of making the latte (I did not synthetically create any difficulties). In order to do that, I wanted ChatGPT to give me one instruction and then I wanted to respond with the outcome. For example, here is a part of a dialog in which I had a (human) friend guide me through the same task via text:

Friend: Select a small pot.

Me: Ok I have chosen the pot but I am still standing looking at the pot while it is sitting on the bottom shelf

Friend: Bend down and retrieve the pot.

Me: Done!

Friend: turn around and place the pot on an available burner.

Me: Done!

Friend: turn to your right, walk three steps, turn to your left, go to the metal fridge on the right.

Me: Done!

Overall, ChatGPT really struggled to give one instruction at a time. I spent much of the dialog insisting on this:

Hello! I’d like to make a turmeric latte using a turmeric and cinnamon mix that my friend left when she visited me recently. Can you guide me step by step in how to make a latte using this?

Sure! Here's a recipe for a turmeric latte that you can make at home:

[11 lines of nicely formatted recipe]

I hope you enjoy your latte! Let me know if you have any questions or need further guidance.

Ok thank you for that but I was wondering if you could walk me through one step at a time

Certainly! Here are the steps for making a turmeric latte, one step at a time:

[6 steps in a numbered list]

I hope this helps! Let me know if you have any questions or need further guidance.

Ok that’s amazing but could you give me just the first step?

Certainly! The first step to making a turmeric latte is to gather all of your ingredients. You will need:

[5 bullet points with the ingredients]

Once you have all of your ingredients, you are ready to begin making your turmeric latte. Let me know if you have any questions or need further guidance.

Ok great but how do I find the ingredients?

You should be able to find all of the ingredients needed for a turmeric latte at most supermarkets or grocery stores.

[bullet points with guidance on finding the ingredients in a supermarket]

I hope this helps! Let me know if you have any other questions.

I’m pretty sure everything in that list is in the kitchen where I am right now. Let’s go one ingredient at a time. Can you walk me through finding the ingredients?

Certainly! Here is some guidance on finding each of the ingredients you will need for your turmeric latte:

[more bullet points]

[about 5 more iterations of trying to get it to give me instructions that I could perform in 10 seconds or less]

That’s 6 steps, not one. Can you give me just the first step?

Certainly! Here is the first step for making a turmeric latte:

Step 1: Gather all of your ingredients. You wi...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Building Effective Altruism in Africa: My Experience Running a University Group, published by Tim K. Sankara on January 12, 2023 on The Effective Altruism Forum.

Welcome to my adventure of creating an effective altruism movement in Africa. My journey started when I started a student organisation. I'll relate my experiences, obstacles I faced, and triumphs I had while establishing EA in Africa in this post, along with how I handled them. I'll also discuss my plans for creating a thriving university community that will have a beneficial, long-lasting impact on this area.I'm Tim, an Information Technology graduate of the Jomo Kenyatta University of Agriculture and Technology. In July this year, I celebrated the successful completion of my undergrad! My varied interests in the EA movement have been documented in brief here.

My official EA journey commenced in early 2022 when I joined the introductory program, then later, the in-depth program. However, I had already been engaging with various EA content prior to this, without even knowing that it was closely linked to EA.

These sessions introduced me to the EA forum, through which I learned about the University Groups Accelerator Program, which aims to increase support for EA cause areas by providing support to university groups around the globe.

The application procedure was simple, and after a few interviews, training sessions, and mentorship meetings, I was finally prepared to launch the Uni group.

Running a university group involved a variety of tasks, including but not limited to:

Outreach & group growth strategy

Curriculum development

Event planning (dinners, socials, retreat 1-on-1s)

Community health

I'll stick to a few of the semi-distinctive aspects that make founding a university organisation alone, and especially among the first few Uni groups on my continent, something worth writing about and perhaps someone will borrow a leaf from my experience. Although this advise may occasionally be particular or generic, it is hopefully still useful. So here goes.

1. The importance of making the group a place where people want to be

I recently attended an EA related conference and give a brief talk on my experience and one of the attendees asked me if I had to pick the one thing that made my group successful and I picked: "make the group a fun place to be".I can't insist enough how much this helps to build a sense of community. It is important to provide a variety of activities and experiences that allow members to have fun and make memories together. By creating a place where people want to be, the group can create a strong sense of community and foster a culture of collaboration, creativity, and progress.

Some activities that worked for us included:

Having some unique ice-breakers before each session like this one here.

Finding a conducive space where the fellows can be away from distractions and can just relax and unwind with free pizza and soft drinks.

Holding a game-night. I asked around for some games that may be fun among my group and settled on a few board games, a quiz night of sorts as well as some digital games too.

But apart from the fun activities, the group was a place where intellectually stimulating discussions could be had and meaningful connections could be made. We sought to create an atmosphere of respect, inclusiveness and collaboration, which allowed for people to learn from each other, share ideas and grow. This was an essential part of making the group a place where people not only wanted to be, but wanted to stay and contribute.

2. Watch for community health

In order to make the group inviting and engaging, it is important to create a safe space where people can express their thoughts and ideas without fear of judgment. This can be achieved by having a code of conduct, providing resources and supp...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: What you can do to help stop violence against women and girls, published by Akhil on January 12, 2023 on The Effective Altruism Forum.

Introduction

I previously wrote an entry for the Open Philanthropy Cause Exploration Prize on why preventing violence against women and girls is a global priority. For an introduction to the area, I have written a brief summary below. In this post, I will extend that work, diving deeper into the literature and the landscape of organisations in the field, as well as creating a cost-effectiveness model for some of the most promising preventative interventions. Based on this, I will offer some concrete recommendations that different stakeholders should take - from individuals looking to donate, to funders, to charity evaluators and incubators.

The key recommendations I make, in order of importance, are:

Support community-based interventions that seek to shift harmful gender norms and reduce violence- they have a high quality of evidence, and cost $180/DALY (disability adjusted life years) or $150 for a woman to live a year free from violence. Two particularly promising organisations are CEDOVIP and Raising Voices.

Fund and conduct a well-designed randomised control trial of radio or television dramas to shift harmful gender norms and reduce violence- they have a startling cost-effectiveness of $13/DALY or $11 for a woman to live a year free from violence, but currently lacks a well-established evidence base.

Fund organisations undertaking economic programs supporting women (e.g. microfinancing, cash transfers, village savings and loans association) to add on social empowerment programs focused on reducing violence; they have a cost-effectiveness of $180/DALY, or $145 for a woman to live a year free from violence.

Found new charities focused on community-based interventions that seek to shift harmful gender norms and reduce violence, particularly in neglected geographies and populations

Consider supporting self-defence training programs - although analysis suggests that self-defence training may not be as cost-effective as other interventions for VAWG, it is nevertheless a relatively cost-effective (at $260/DALY or $215 for a woman to live a year free from violence) and potentially scalable intervention. No Means No Worldwide and Ujamaa appear to be two organisations scaling this intervention well.

Fund academic research to understand what types of culture change programs targeting boys and men are most effective at reducing violence against women and girls

You can find a 2 page summary of my initial post and this post here

Why VAWG is an important cause area

Nearly one third of women and girls aged 15 years of age or older have experienced either physical or sexual intimate partner violence (IPV) or non-partner sexual violence globally, with 13% (10–16%) experiencing it in 2018 alone (Sardinha et al 2022). It is one of the leading burdens of disease globally, responsible for 8.5 million DALYs and 68 500 deaths annually.

In several countries, violence against women is in the top 3-5 leading causes of death for young women aged between 15 and 29 (Mendoza et al 2018).

VAWG has wide-ranging effects on women’s physical, sexual and mental health- in fact, it is responsible for 11% of the DALY burden of depressive disorders and 14% of the DALY burden of HIV in women (IHME 2019)

Globally, the rates of VAWG are both alarmingly high and have increased over the last 30 years, despite gains in other areas of women’s health, such as maternal care (Think Global Health).

In 2016, the global economic cost of violence against women was estimated by the UN to be US$1.5 trillion, equivalent to approximately 2% of the global GDP (UN Women 2016).

Although there are many groups working to stop VAWG, it is fairly neglected relative to the scale of harm that it...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Announcing the awardees for Open Philanthropy's $150M Regranting Challenge, published by ChrisSmith on January 12, 2023 on The Effective Altruism Forum.

In February, we launched the Regranting Challenge, aiming to add $150 million in funding to the budgets of outstanding grantmakers at other foundations.

We saw the Challenge as an opportunity to maximize our impact, by:

Adding funding to high-impact work that was already underway. We’ve long been inspired by the work of other grantmakers, and we believe there are highly effective grantmaking organizations doing better work than we could in their respective spaces. Instead of reinventing the wheel, we’ve used the Regranting Challenge to give some of those organizations additional funding, so they can increase the scale and scope of their own work.

Piloting a new approach to growing highly effective grantmaking programs. The lack of feedback mechanisms that ensure effective grantmakers get more money to allocate is a major shortcoming in the existing philanthropic ecosystem. The Regranting Challenge gave us a chance to experiment with changing that dynamic.

Learning from a wide range of grantmakers with different approaches. By creating an open call, we were able to identify highly effective foundations and program areas to support that we wouldn’t have known about otherwise.

Results

After eight months, three selection rounds, and evaluations from dozens of experts (inside and outside Open Philanthropy), we’ve chosen the awardees!

We’ll share a brief description of each program here, but you can learn more about them on the Regranting Challenge website.

Development Innovation Ventures: $45,000,000

Development Innovation Ventures is a program inside the United States Agency for International Development (USAID). They invest in early-stage organizations and projects in global health and development that have the potential to be highly impactful and cost-effective. They’ve supported programs in water sanitation, early childhood education, and routine immunization (among others).

(Read more)

Eleanor Crook Foundation: $25,000,000

The Eleanor Crook Foundation funds research and advocacy to end global malnutrition. They have a track record of successfully advocating for the increased use of the “Power 4” malnutrition interventions — prenatal vitamins for pregnant women, breastfeeding support for mothers, vitamin A supplementation, and ready-to-use therapeutic food to treat wasting.

(Read more)

Global Education, Bill & Melinda Gates Foundation: $5,000,000

The Global Education program at the Gates Foundation makes grants to organizations that aredeveloping and improving highly effective education interventions. These interventions — namely remediation, structured pedagogy, and teaching at the right level — have been shown to improve foundational literacy and numeracy in low- and middle-income countries.

(Read more)

Global Health Innovation, Bill & Melinda Gates Foundation: $65,000,000

We are supporting two global health initiatives at the Gates Foundation. The first ($40,000,000) will fund grantees who are advancing a new vaccine through efficacy trials against tuberculosis in adults and adolescents. The second ($25,000,000) will fund grantees who are helping a new vaccine manufacturer supply oral cholera vaccine — which will diversify the vaccine’s manufacturing base and increase vaccine supply to better meet global demand.

(Read more)

Tara Climate Foundation: $10,000,000

Tara Climate Foundation focuses on climate change mitigation in South, Southeast, and East Asia (excluding China and India). They help found new nonprofits and grow the climate movement across this region. In addition to climate impacts, their work could also lead to substantial improvements in air quality.

(Read more)

Process

Here’s more on how we arrived at...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Iron deficiencies are very bad and you should treat them, published by Elizabeth on January 12, 2023 on LessWrong.

In brief

Recently I became interested in what kind of costs were inflicted by iron deficiency, so I looked up studies until I got tired. This was not an exhaustive search, but the results are so striking that even with wide error bars I found them compelling. So compelling I wrote up a post with an algorithm for treating iron deficiency while minimizing the chance of poisoning yourself. I’ve put the algorithm and a summary of potential gains first to get your attention, but if you’re considering acting on this I strongly encourage you to continue reading to the rest of the post where I provide the evidence for my beliefs.

Tl;dr: If you are vegan or menstruate regularly, there’s a 10-50% chance you are iron deficient. Excess iron is dangerous so you shouldn’t supplement blindly, but deficiency is easy and cheap to diagnose with a common blood test. If you are deficient, iron supplementation is also easy and cheap and could give you a half standard deviation boost on multiple cognitive metrics (plus any exercise will be more effective). Due to the many uses of iron in the body, I expect moderate improvements in many areas, although how much and where will vary by person.

Note that I’m not a doctor and even if I was there isn’t good data on this, so it’s all pretty fuzzy. The following is an algorithm for treating iron deficiency that I’ve kludged together from various doctors. I strongly believe it is a lot better than nothing on average, but individuals vary a lot and you might be unlucky.

Take a serum ferritin test. If you have a doctor they will almost certainly say yes to a request, or you can order for yourself at walkinlab.com

If your results show a deficiency (<20ug/L), increase iron intake through diet or supplements such as Ferrochel, taking the default dose once per day, with a meal.

The definition of deficiency can vary by study, lab and goal. I picked <20ug/L because it’s the highest level I have concrete evidence is insufficient, but personally believe people are likely to benefit from iron beyond that and am taking pills accordingly.

If you experience negative effects after taking the pills, stop immediately. Give yourself a week to recover, then you can try other brands, be more careful to eat with a full meal, etc.

If you are experiencing the symptoms of iron poisoning (listed below), stop pills and see a doctor now. Iron poisoning is a very big deal, which is why step 1 of this algorithm is “get tested” not “gobble pills”. Unfortunately several of these are pretty generic, but I’m never going to feel bad about telling people with seizures to seak medical attention:

Nausea

Vomiting

Abdominal pain

Dizziness

Low blood pressure and a fast or weak pulse

Headache

Fever

Shortness of breath and fluid in the lungs

Grayish or bluish color in the skin

Jaundice (yellowing of the skin due to liver damage)

Seizures

Black or bloody stools

Retest at 8-12 weeks, ideally at the same lab as before.

Continue to retest every 8-12 weeks.

If you increase by 20ug from your starting value without noticing any improvements to your cognition or overall energy levels; low ferritin is probably not your bottleneck.

If you believe it’s not a problem at all, quit.

If you believe it is a problem but another problem is limiting your gains, stay on a maintenance dosage but don’t put more time into managing this. Verrrrry roughly, divide your current dosage by your currently monthly gains (so If you take one RDA/day and gain 10ug/month, your result is 0.1), and take that much. This hopefully keeps you from losing ground, without gaining so quickly it could become a problem.

If you’re getting improvements, keep going until those taper off. I personally would exercise caution ...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Announcing: SERI Biosecurity Interventions Technical Seminar (BITS), published by James Lin on January 12, 2023 on The Effective Altruism Forum.

1. Introduction

The biosecurity space has promising technical interventions, such as UV-C and metagenomic sequencing. Often, students in college and community-builders are interested in biosecurity, but don’t have an in-depth understanding of the solutions. This is concerning for 3 central reasons:

Poor explanations are presented for biosecurity solutions, and may contain inaccuracies

The reputation of the biosecurity community is damaged

Public perception of biosecurity interventions will be off the mark relative to the intervention’s true effectiveness (e.g. “far UVC will end pandemics by 2025”)

As a consequence, people are turned away from biosecurity

Promising people transition to other fields of work

When someone doesn’t understand the specifics, it’s highly unlikely that they’ll pursue direct work, e.g. building a startup for that technology

Instead, they’ll likely transition to community-building or traditional career paths

Biosecurity culture becomes diluted and can’t attract technically bright researchers or entrepreneurs

Competence is magnetic, and promoting a culture of technical learning and deep dives is crucial to doing impactful work in a domain as challenging as biosecurity

A technical biosecurity program can serve as a way for students to become much more knowledgeable about biosecurity solutions.

2. Context

There have been biosecurity discussion groups run in the past, but most of these groups focus on the top-of-the-funnel content—the introductory information and context that helps to form a foundation. To our knowledge, there doesn’t yet exist a deep dive group, one that specifically aims to equip students with technical understanding of various biosecurity interventions.

When it comes to precedent, there have been similar discussion and research mentorship programs in the field of AI alignment, including AGISF, SERI MATS, MLSS, and the CERI Fellowship for biosecurity, nuclear risk, climate change, and AI safety. These programs have been largely successful, providing hundreds of students with hands-on technical experience or in-depth discussions about particular domains of AI safety.

3. Program

About

Our proposal is to establish an online 8-week-long technical deep dive group. This group will explore many crucial aspects of technical biosecurity interventions while ensuring that we maintain an emphasis on defensive interventions. Examples of session topics include far-UVC, SotA for PPE, and the NAO (syllabus). The program will culminate in a research presentation for a chosen aspect of related topics and interventions (more details here). Each group will meet once per week and will have ~5 people. The expected weekly time commitment is 2-3 hours.

There will be a virtual and in-person pilot of this program. Everyone is eligible to apply for the virtual program. The in-person program seminar group will be held at Stanford, though depending on demand we may also run an in-person group at Berkeley.

There are a few reasons for running a primarily virtual program:

It’s important to have an experienced biosecurity person facilitating, which usually limits these interactions to virtual

Virtual programs are much more scalable and flexible

It’s easier to reach a critical mass of students

If you would like to join, please feel free to apply! The application deadline is Friday, Jan 20th, and the program is expected to start in 2-3 weeks. We expect to start by running a few groups in parallel at different times during the week to accommodate those in different time zones.

Syllabus

This is the syllabus with the content we will be covering. There will also be an optional research poster to be submitted 2 weeks after the...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: EA could use better internal communications infrastructure, published by Ozzie Gooen on January 12, 2023 on The Effective Altruism Forum.

(This was a quick post, written in around 30 min. It was originally posted on Facebook, where it generated some good discussion.)

I really wish EA had better internal communications.

If I wanted to make a blog post / message / recording accessible to a "large subset of effective altruist professionals", I'm not sure how I'd do that.

I don't think we yet have:

One accepted chat system

An internal blogging system

Any internal email lists (for a very wide net of EA professionals)

It's nice to encourage people to communicate publicly, but there's a lot of communication that's really not meant for that.

Generally, the existing options are:

Post to your internal org slack/emails (note: many EA orgs are tiny)

Share with people in your office

Post to one of a few domain-specific and idiosyncratic Slacks/Discords

Post publicly, for everyone to see

I think the SBF situation might have shown some substantial vulnerability here. It was a crisis where public statements were taken as serious legal statements. This meant that EA leadership essentially didn’t have a real method of communicating with most EAs.

I feel like much of EA is a lot like one big org that tries really hard not to be one big org. This gives us some advantages of being decentralized, but we are missing a lot of the advantages of centralization. If "Professional EAs" were looked at as one large org, I'd expect that we'd look fairly amateur, compared to other sizeable organizations.

A very simple way to make progress on internal communications is to separate the issue into a few clusters, and then attack each one separately.

Access/Onboarding/OffboardingMake official lists that cover "professional/trusted members". You could start with simple criteria like "works at an org funded by an EA funder" or "went to 2+ EAGs".

Negotiation and Moderation"EA Professionals" might basically be an "enterprise", and need "enterprise tools". These often are expensive and require negotiation.

A Responsible IndividualMy preference would be that we find someone who did a good job at this sort of thing in other sizeable companies and try to get them to do it here.

I bet with $200k/year for the talent, plus maybe $200k-$1k/year, we could have a decent setup, assuming we could find good enough talent. That said, this would definitely be work to establish, so I wouldn't expect anything soon.

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Abolitionist in the Streets, Pragmatist in the Sheets: New Ideas for Effective Animal Advocacy, published by Dhruv Makwana on January 11, 2023 on The Effective Altruism Forum.

This series of posts is not intended to re-ignite the never-ending philosophical debate about welfarism vs abolitionism. Its main point is simply to point out that (broadly speaking) animal advocacy within Effective Altruism is uniform in its welfarist thinking and approach, and that it has assumed with insufficient reason that all abolitionist thinking and approaches are ineffective. This assumption is partly based on (1) poor evidence (2) a narrow conception of what it means to take an abolitionist approach to animal advocacy (3) with one notable exception, abolitionists’ poor/absent engagement with animal advocacy on pragmatic terms.

It is important to note I am not claiming that alternative approaches mentioned are (cost-)effective, though I do state reasons this might be true. In logic terms, I am simply saying that currently, the status quo is,

EA Animal Advocacy = Welfarist approaches & not (Abolitionist approaches).

What I am proposing is that it should at least be

EA Animal Advocacy = Welfarist approaches & unknown(Abolitionist approaches).

That is, the community does not have strong enough evidence and arguments to rule out abolitionist thinking and approaches. This opens up new lines of thinking, new questions for research and new advocacy strategies to test and refine. I also suggest how advocates might measure such strategies’ effectiveness, in ways appropriate to abolitionism (Social Change Dynamics).

Epistemics

I am a pragmatic, abolitionist-leaning animal advocate and vegan and I wrote this series voluntarily, in rare spare time, with help from three other vegans, all of whom have been familiar with EA for a few years. Two of us have a moderate amount of experience (a few years) volunteering with local abolitionist activism of various sorts, and we recognise this could lead to motivated reasoning. Two of us would identify as EAs, and have attended EA conferences. One of us is studying for a PhD in wild animal suffering; I am studying for a PhD in Computer Science. None of us have any formal training or professional experience in EA-related research, or criticism. Most importantly, this series was written over the course of more than 14 months; the sheer length of time means research moves on - I’ve done my best to keep up, and have even been scooped on multiple points, but I’ve probably missed things.

This series is intended to be a big-picture piece, surfacing and investigating common beliefs within EA animal advocacy. It covers a huge amount of ground, more akin to setting a research agenda than answering specific questions: each paragraph could be the subject of an investigation. Necessarily, I deal with generalisations of views, which will not cover all organisations, variations, or advocates. I have made my best efforts to avoid accidentally strawmanning, by, wherever possible, checking my intuition on what the mainstream views are against any available evidence (funding allocations, forum posts, surveys, reports from prominent EA organisations, searches on the EA forum).

Summary

A Case for AbolitionI present a short, pragmatic, and EA-targeted case for complete abolition of animal exploitation, and for using abolitionist approaches to achieve this. I show that (1) a longtermist perspective leads one to aim for complete abolition as a goal, and with one key assumption, using abolitionist approaches to get there and (2) contrary to prior work, abolition is helpful to reducing wild animal suffering (and conversely, welfarism could hinder such efforts).

Limitations with Current EA Animal AdvocacyI argue that the current major strands of EA Animal Advocacy (corporate welfare ca...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: We don’t trade with ants, published by Katja Grace on January 12, 2023 on The Effective Altruism Forum.

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: What's the deal with AI consciousness?, published by ThomasW on January 11, 2023 on The Effective Altruism Forum.

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.