The Nonlinear Library The Nonlinear Fund

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 intuition with spaced repetition systems, published by Jacob G-W on May 14, 2024 on LessWrong.
Do you ever go to a lecture, follow it thinking it makes total sense, then look back at your notes later and realize it makes no sense? This used to happen to me, but I've learned how to use spaced repetition to fully avoid this if I want. I'm going to try to convey this method in this post.
Much of my understanding of how to create flashcards comes from "Using spaced repetition systems to see through a piece of mathematics" by Michael Nielsen and "How to write good prompts: using spaced repetition to create understanding" by Andy Matuschak, but I think my method falls in between both, in terms of abstraction. Finally, I want to credit Quantum Country for being an amazing example of flashcards created to develop intuition in users.
My method is more abstract than Michael Nielsen's approach, since it does not only apply to mathematics, but to any subject. Yet it is less abstract than Andy Matuschak's approach because I specifically use it for 'academic subjects' that require deep intuition of (causal or other) relationships between concepts.
Many of Matuschak's principles in his essay apply here (I want to make sure to give him credit), but I'm looking at it through the 'how can we develop deep intuition in an academic subject in the fastest possible time?' lens.
Minimize Inferential Distance on Flashcards
A method that I like to repeat to myself while making flashcards that I haven't seen in other places is that each flashcard should only have one inferential step on it. I'm using 'inferential step' here to mean a step such as remembering a fact, making a logical deduction, visualizing something, or anything that requires thinking. It's necessary that a flashcard only have a single inferential step on it. Anki trains the mind to do these steps.
If you learn all the inferential steps, you will be able to fully re-create any mathematical deduction, historical story, or scientific argument. Knowing (and continually remembering) the full story with spaced repetition builds intuition.
I'm going to illustrate this point by sharing some flashcards that I made while trying to understand how Transformers (GPT-2) worked. I made these flashcards while implementing a transformer based on Neel Nanda's tutorials and these two blog posts.
Understanding Attention
The first step in my method is to learn or read enough so that you have part of the whole loaded into your head. For me, this looked like picking the
attention step of a transformer and then reading about it in the two blog posts and watching the section of the video on it. It's really important to learn about something from multiple perspectives. Even when I'm making flashcards from a lecture, I have my web browser open and I'm looking up things that I thought were confusing while making flashcards.
My next step is to understand that intuition is fake! Really good resources make you feel like you understand something, but to actually understand something, you need to engage with it. This engagement can take many forms. For technical topics, it usually looks like solving problems or coding, and this is good! I did this for transformers! But I also wanted to not forget it long term, so I used spaced repetition to cement my intuition.
Enough talk, here are some flashcards about attention in a transformer. For each flashcard, I'll explain why I made it. Feel free to scroll through.
Examples
I start with a distillation of the key points of the article.
I wanted to make sure that I knew what the attention operation was actually doing, as the blog posts emphasized this.
When building intuition, I find it helpful to know "the shape" or constraints about something so that I can build a more accurate mental model. In this case, th

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: Monthly Roundup #18: May 2024, published by Zvi on May 14, 2024 on LessWrong.
As I note in the third section, I will be attending LessOnline at month's end at Lighthaven in Berkeley. If that is your kind of event, then consider going, and buy your ticket today before prices go up.
This month's edition was an opportunity to finish off some things that got left out of previous editions or where events have left many of the issues behind, including the question of TikTok.
Oh No
All of this has happened before. And all of this shall happen again.
Alex Tabarrok: I regret to inform you that the CDC is at it again.
Marc Johnson: We developed an assay for testing for H5N1 from wastewater over a year ago. (I wasn't expecting it in milk, but I figured it was going to poke up somewhere.)
However, I was just on a call with the CDC and they are advising us NOT to use it.
I need a drink.
They say it will only add to the confusion because we won't know where it is coming from. I'm part of a team. I don't get to make those decisions myself.
Ben Hardisty: The usual institute, or did they have a good reason?
Marc Johnson: They say it would only add to the confusion since we don't know precisely where it is coming from. But then they said 2 minutes later that they aren't sure this isn't just regular influenza appearing late. We can answer that, so why don't we??? I don't get it.
Alex: Are your team members considering bucking the CDC advice or has the decision been made to acquiesce? I understand them not wanting panic but man if that's not self serving advice I don't know what is.
Marc Johnson: The CDC will come around.
ZzippyCorgi11: Marc, can private entities ask you to test wastewater around their locations? Is the CDC effectively shutting down any and all testing of wastewater for H5N1?
Marc Johnson: No, if people want to send me wastewater I can test them with other funding. I just can't test the samples I get from state surveillance.
JH: This is ridiculous. Do it anyway!
Marc Johnson: It's not my call. I got burned once for finding Polio somewhere I wasn't supposed to find it. It fizzled, fortunately.
Ross Rheingans-Yoo: It's a societal mistake that we're not always monitoring for outbreaks of the dozen greatest threats, given how cheap wastewater testing can get.
Active intervention by the CDC to stop new testing for a new strain of influenza circulating in mammals on farms is unconscionable.
I strongly agree with Ross here. Of all the lessons to not have learned from Covid, this seems like the dumbest one to not have learned. How hard is 'tests help you identify what is going on even when they are imperfect, so use them'?
I am not so worried, yet, that something too terrible is that likely to happen. But we are doing our best to change that.
We have a pattern of failing to prepare for such easily foreseeable disasters. Another potential example I saw today would be the high-voltage transformers, where we do not make them, we not have backups available and if we lost the ones we have our grid plausibly collapses. The worry in the thread is primarily storms but also what about sabotage?
Oh No: Betting on Elections
I am proud to live in an information environment where 100% of the people, no matter their other differences, understand that 'ban all prediction markets on elections' is a deeply evil and counterproductive act of epistemic sabotage.
And yet that is exactly what the CFTC is planning to do, with about a 60% chance they will manage to make this stick.
Maxim Lott: This afternoon, the government bureaucrats at the CFTC announced that they plan to ban all election betting (aka "prediction markets on elections", aka "event contracts") in the United States. They will also ban trading on events in general - for example, on who will win an Oscar.
The decision was 3-2, with 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: Environmentalism in the United States Is Unusually Partisan, published by Jeffrey Heninger on May 13, 2024 on LessWrong.
This is the first in a sequence of four posts taken from my recent report: Why Did Environmentalism Become Partisan?
Introduction
In the United States, environmentalism is extremely partisan.
It might feel like this was inevitable. Caring about the environment, and supporting government action to protect the environment, might seem like they are inherently left-leaning. Partisanship has increased for many issues, so it might not be surprising that environmentalism became partisan too.
Looking at the public opinion polls more closely makes it more surprising. Environmentalism in the United States is unusually partisan, compared to other issues, compared to other countries, and compared to the United States itself at other times.
The partisanship of environmentalism was not inevitable.
Compared to Other Issues
Environmentalism is one of the, if not the, most partisan issues in the US.
The most recent data demonstrating this comes from a Gallup poll from 2023.[1] Of the 24 issues surveyed, "Protecting the Environment Has Priority Over Energy Development" was tied for the largest partisan gap with "Government Should Ensure That Everyone Has Healthcare." Of the top 5 most partisan issues, 3 were related to environmentalism. The amount this gap has widened since 2003 is also above average for these environmental issues.
Figure 1: The percentages of Republicans and Democrats who agree with each statement shown, 2003-2023. Reprinted from Gallup (2023).
Pew also has some recent relevant data.[2] They ask whether 21 particular policies "should be a top priority for the president and Congress to address this year." The largest partisan gap is for "protecting the environment" (47 p.p.), followed by "dealing with global climate change" (46 p.p.). These are ten percentage points higher than the next most partisan priority. These issues are less specific than the ones Gallup asked about, and so might not reveal as much of the underlying partisanship.
For example, most Democrats and most Republicans agree that strengthening the economy is important, but they might disagree about how this should be done.
Figure 2: The percentages of Republicans and Democrats who believe that each issue should be a top priority. Reprinted from Pew (2023).
Guber's analysis of Gallup polls from 1990, 2000, & 2010 also shows that environmentalism is unusually partisan.[3] Concern about "the quality of the environment" has a similar partisan gap as concern about "illegal immigration," and larger than concern about any other political issue. If we hone in on concern about "global warming" within overall environmental concern, the partisan gap doubles, making it a clear outlier.
Figure 3: Difference between the mean response on a four point scale for party identifiers on concern for various national problems in 2010. "I'm going to read you a list of problems facing the country. For each one, please tell me if you personally worry about this problem a great deal, a fair amount, only a little, or not at all." Reprinted from Guber (2013).
The partisanship of environmentalism cannot be explained entirely by the processes that made other issues partisan. It is more partisan than those other issues. At least this extra partisan gap wants an explanation.
Compared to Other Countries
The United States is more partisan than any other country on environmentalism, by a wide margin.
The best data comes from a Pew survey of "17 advanced economies" in 2021.[4] It found that 7 of them had no significant partisan gap, and that the US had a partisan gap that was almost twice as large as any other country.
Figure 4: Percentages of people with different ideologies who would be willing to make a lot of or som

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: In DC, a new wave of AI lobbyists gains the upper hand, published by Chris Leong on May 13, 2024 on The Effective Altruism Forum.
"The new influence web is pushing the argument that AI is less an existential danger than a crucial business opportunity, and arguing that strict safety rules would hand America's AI edge to China. It has already caused key lawmakers to back off some of their more worried rhetoric about the technology.
... The effort, a loosely coordinated campaign led by tech giants IBM and Meta, includes wealthy new players in the AI lobbying space such as top chipmaker Nvidia, as well as smaller AI startups, the influential venture capital firm Andreessen Horowitz and libertarian billionaire Charles Koch.
... Last year, Rep. Ted Lieu (D-Calif.) declared himself "freaked out" by cutting-edge AI systems, also known as frontier models, and called for regulation to ward off several scary scenarios. Today, Lieu co-chairs the House AI Task Force and says he's unconvinced by claims that Congress must crack down on advanced AI.
"If you just say, 'We're scared of frontier models' - okay, maybe we should be scared," Lieu told POLITICO. "But I would need something beyond that to do legislation. I would need to know what is the threat or the harm that we're trying to stop."
... After months of conversations with IBM and its allies, Rep. Jay Obernolte (R-Calif.), chair of the House AI Task Force, says more lawmakers are now openly questioning whether advanced AI models are really that dangerous.
In an April interview, Obernolte called it "the wrong path" for Washington to require licenses for frontier AI. And he said skepticism of that approach seems to be spreading.
"I think the people I serve with are much more realistic now about the fact that AI - I mean, it has very consequential negative impacts, potentially, but those do not include an army of evil robots rising up to take over the world," said Obernolte."
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: Impact Accelerator Program for EA Professionals, published by High Impact Professionals on May 13, 2024 on The Effective Altruism Forum.
High Impact Professionals is excited to announce that applications are now open for the Summer 2024 round of our
Impact Accelerator Program (IAP). The IAP is a 6-week program designed to equip experienced EA-aligned professionals (not currently working at an EA organization) with the knowledge and tools necessary to make a meaningful impact and empower them to start taking actionable steps right away.
To date, the program has been a success, with several alumni having already changed careers to new impactful roles and nearly all participants planning to do the same in the next ~6 months. IAP alumni are also volunteering an average of ~100 hours at EA-aligned orgs/projects and donating on average more than 7% of their annual salary to effective charities. We are currently running the Spring 2024 round, which features a larger number of participants and cohorts.
We're pleased to open up this new Summer 2024 program round, which will start the week of July 15. More information is available below and
here. Please
apply here by May 23.
Program Objectives
The IAP is set up to help participants:
identify paths to impact,
take concrete, impactful actions, and
join a network of like-minded, experienced, and supportive EA professionals.
At the end of the program, a participant should have a good answer to the question "How can I have the most impact with my career, and what are my next steps?", and they should have taken the first steps in that direction.
Program Overview
Important Dates
Deadline to apply: May 23
Apply here
Program duration: 6 weeks (week of July 15 - week of August 19, 2024)
Format
Weekly individual work (2-3 hours). A mix of:
Learning: Reading resources on how to think about and prioritize different options for impact
Doing: Taking impactful actions, such as developing your own personalized impact plan and taking concrete steps to begin implementing it
Virtual group sessions (1.5 hours of discussions and coaching)
Includes mastermind sessions where each member of the cohort has the opportunity to present their plans, obstacles, and uncertainties and get in-depth, tailored feedback
1-on-1 sessions with IAP facilitators
Extracurricular sessions: The possibility of extra sessions on topics defined by the needs of the cohort (e.g., financial considerations, networking)
Post-program support sessions: Access additional group sessions in the months following the program to maintain momentum and continue implementing your plan
Topics covered
Values and mission - determine your motivations / guiding principles, strengths, and weaknesses to set a clear starting point for your journey
Paths to impact for professionals - explore the landscape of career possibilities, address your key uncertainties, and identify your best career and impact options
Develop an action plan - put all you've discovered into a roadmap with actionable steps to guide you to your impact goals
Implement your plan - Begin taking active steps to turn your plan into an impactful reality
Led by: Select members of the High Impact Professionals team/community
Why Should You Apply?
Overcome Barriers: We'll guide you in exploring your personal obstacles to impact and assist you in taking real-world, impactful actions.
Understand Impact: Delve into the complex nature of creating positive change and discover the diverse opportunities available to professionals.
Develop and Implement a Solid Impact Plan: Acquire tools to assess and plan for impact, and integrate them into your own circumstances to create a personal roadmap for maximizing your positive influence. Then, begin taking concrete steps to put your plan into action.
Connect with Like-Minded People: Embark on this j

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: "Cool Things Our GHW Grantees Have Done in 2023" - Open Philanthropy, published by Lizka on May 13, 2024 on The Effective Altruism Forum.
Open Philanthropy[1] recently shared a blog post with a list of some cool things accomplished in 2023 by grantees of their Global Health and Wellbeing (GHW) programs (including farm animal welfare). The post "aims to highlight just a few updates on what our grantees accomplished in 2023, to showcase their impact and make [OP's] work a little more tangible."
I'm link-posting because I found it valuable to read about these projects, several of which I hadn't heard of. And I like that despite its brevity, the post manages to include a lot of relevant information (and links), along with explanations of the key relevant theories of change and opportunity.
For people who don't want to click through to the post itself, I'm including an overview of what's included and a selection of excerpts below.
Overview
The post introduces each program with a little blurb, and then provides 1-2 examples of projects and one of their updates from 2023.
Here's the table of contents:
1. Global Public Health Policy
1. Dr. Sachchida Tripathi (air quality sensors)
2. Lead Exposure Elimination Project (LEEP)
2. Global Health R&D
1. Cures Within Reach
2. SAVAC
3. Scientific Research
1. Dr. Caitlin Howell (catheters)
2. Dr. Allan Basbaum (pain research)
4. Land Use Reform
1. Sightline Institute
5. Innovation Policy
1. Institute for Progress
2. Institute for Replication
6. Farm Animal Welfare
1. Open Wing Alliance
2. Aquaculture Stewardship Council
7. Global Aid Policy
1. PoliPoli
8. Effective Altruism (Global Health and Wellbeing)
1. Charity Entrepreneurship
9. How you can support our grantees
Examples/excerpts from the post
I've chosen some examples (pretty arbitrarily - I'm really excited about many of the other examples, but wanted to limit myself here), and am including quotes from the original post.
1.1 Dr. Sachchida Tripathi (air quality sensors)
Sachchida Tripathi is a professor at IIT Kanpur, one of India's leading universities, where he focuses on civil engineering and sustainable energy.
Dr. Tripathi used an Open Philanthropy grant to purchase 1,400 low-cost air quality sensors and place them in every block[2] in rural Uttar Pradesh and Bihar. Using low-cost sensors involved procuring and calibrating them (see photo).
These sensors now provide much more accurate and reliable data for these rural areas than was previously available to the air quality community.
This work has two main routes to impact. First, these sensors make the problem of rural air pollution legible. Because air quality in India is assumed to be a largely urban issue, most ground-based sensors are in urban areas. Second, proving the value of these low-cost sensors and getting operational experience can encourage buy-in from stakeholders (e.g., local governments) who may fund additional sensors or other air quality interventions.
Air quality monitoring is a major theme of our South Asian Air Quality grantmaking. We are actively exploring opportunities in new geographic areas, both within and beyond India, without high-quality, ground-based monitoring. Santosh Harish, who leads our grantmaking on environmental health, recently spoke to the 80,000 Hours podcast about this grant as well as air quality in India more generally.
2.2. SAVAC (accelerating the development and implementation of strep A vaccines)
The Strep A Vaccine Global Consortium (SAVAC) is working to accelerate the development and implementation of safe and effective strep A vaccines.
Open Philanthropy is one of very few funders supporting the development of a group A strep (GAS) vaccine (we've funded two projects to test new vaccines). GAS kills over 500,000 people per year, mostly by causing rheumatic heart disease.[3]
Wh