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: 14 Techniques to Accelerate Your Learning, published by spencerg on May 19, 2022 on LessWrong.

Originally written by Belén Cobeta and Spencer Greenberg from ClearerThinking.org. Revised by Teis Rasmussen & Florence Hinder from ThoughtSaver.com. This is a cross-post from the Effective Altruism Forum.

TLDR: There are a number of techniques that may accelerate the speed at which you learn (14 of which we explain in this article). Additionally, embedded within this article are flashcards to provide you with the key takeaways so you can remember these ideas and more effectively put them into action.

Vasconcelos Library — photo by Diego Delso

What if you could learn more in less time? Whether you’re studying to pass your classes, aiming to improve at work, honing your personal life skills, or focusing on having the greatest impact you can, learning to learn more efficiently can be a great time investment because it accelerates the rest of your learning.

Many learning methods are inefficient

Many widespread learning practices waste a lot of time and effort, at least if we assume that the goal of those efforts is to actually learn. For example, you have probably had the experience of reading an exciting and useful piece of nonfiction, only to forget basically all of it and not take any action based on what you learned from it. And you’ve probably spent a lot of time taking classes that taught information that mostly wasn’t useful to you, and which you no longer remember either way.

It’s unfortunate that such experiences are as common as they are. The good news is that there are more efficient and powerful learning practices out there - many of us just don't adopt them.

Here we lay out 14 of our favorite techniques for improving your learning processes. They are grouped into four categories:

Learn faster

Boost understanding

Remember more

Put your learning into practice

Part 1: LEARN FASTER

This section will cover techniques that can help you absorb more information or knowledge per hour that you spend learning.

1. Listening instead of reading

Listening is the new reading: take advantage of technology by listening to books and articles. It may feel less studious than reading, but at least some research shows that you can learn and retain just the same. The intonation of the narrator can also help you understand the text. Use the Audible version of the book or text-to-speech software and set the reading speed to a level that feels challenging but still allows you to understand the content. In time, your listening speed may become faster than your reading speed, as you go from listening at 1x to 2x and beyond. Some people even claim that faster speeds can improve comprehension to a degree because they require greater focus and so can prevent the mind from wandering. Another advantage of audio over regular reading is that you can do other activities that don’t use your conscious mind at the same time, like taking out the trash, walking outdoors, washing the dishes, or taking a bath.

Check out The Nonlinear Library for audio-versions of content from blogs such as the EA Forum, Alignment Forum, and LessWrong.

2. Immersive reading

Why choose between listening and reading when you can do both? Try Emerson Spartz´s #1 speed reading hack: read a book using the printed and the audible version simultaneously (that is, read with your eyes WHILE you’re also listening). By engaging two senses at once, your focus and reading speed may increase. It takes some practice, so be sure to try this method for at least a few hours before deciding if it’s right for you (people who just try it for one hour are unlikely to see a benefit). To get the full benefit, you’ll also want to push the speed of the audio to the upper edge of what feels comfortable to you.

3. Recursive sampling

Use this technique t...

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: Does it make sense for EA’s to be more risk-seeking in earning to give?, published by Joey on May 20, 2022 on The Effective Altruism Forum.

Repost from Charity Entrepreneurship's blogHistorically it has been argued that EAs who are looking to make large donations may have more reasons to be risk-neutral than people who are aiming to make money for personal, happiness-focused reasons. This is in part due to the impact of donations not suffering the same diminishing returns as money on happiness. However, we believe there is an additional, significant reason for people who are doing E2G (earning to give) to be even more sympathetic to risk. In short; as income goes up, a larger percentage of the income is donated. This leads to dichotomized earning outcomes and a higher net donation per dollar.

Theoretical example:

Say we have ten EAs, each of whom can do E2G in a low-risk way (e.g. working in software) or a high-risk way (e.g. founding a software startup). In both cases, they all have signed the GWWC pledge and plan to donate 10%+ to an effective charity.

Low-risk outcome: In the low-risk scenario, each does E2G and earns $100k, donating 10%. Across the group this results in $10k x 10 people = $100k donations. This is a really great outcome, but let's look at the higher risk situation with the same expected income value ($1,000,000).

Higher variance scenario:

In this situation, nine out of ten people have startups that do not work that well, resulting in a yearly income of $50k. However, one person earns considerably more; $550k. Most of the lower earners stick with their donation percentage, with eight continuing to give 10% and one no longer giving. This results in $40k of donations. However; the earner of $550k feels 10% seems a little underwhelming relative to their newfound wealth, and instead donates 50%. This results in $315k being donated to charities out of the same $1,000,000 earned by the high-risk group.

Although the financial earnings in these two groups were identical, the donation outcomes differed by more than 3 times ($100k vs $315k). The higher variance scenario did considerably more good.

Real-world data:

This is a theoretical example with made-up numbers. How do we know that people who earn more will, in fact, donate more? Thankfully we have some data on this topic from both within the EA movement and outside of it. Inside the EA movement, we have compelling data showing this trend happens both in GWWC and non-GWWC members; this is likely the most relevant data when thinking about giving EA career advice. However, outside of the EA world there is also some confused data that tentatively points in the same direction.

Countervailing factors:

There are some other factors that can affect this calculation pretty significantly. If half of the ten low-risk EAs have donation matching at their workplace, that increases their net donations pretty significantly (x1.5), although still not enough to change the endline outcome. Also, if you assume that every EA fundraises a bit from people in their same income group it gets even less clear.

On the other hand, a big, beneficial factor of dichotomous income is that it opens up more optimal career change paths. In the low-risk situation, any one of those people changing careers affects the net donations by 10%, however, in the higher variance scenario, nine of the ten people could change careers while only affecting 1.5% of the total donations made. Of course, the cost for the highest income person becomes a lot higher, but assuming the people are of roughly equal talent at the start, having differentiated outcomes allows easier career changes with less percentage of donations lost. Those who manage to launch a successful start-up business are also less likely to change careers in general. Another positive is that ...

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: CE Career Coaching for Impact-focused Entrepreneurs, published by KarolinaSarek on May 20, 2022 on The Effective Altruism Forum.

TL;DR For many, entrepreneurship could be the most impactful career path; both Probably Good and 80,000 Hours recommend it as one of their top paths. But knowing if you are a good fit and deciding which particular route to take can be challenging. Maybe you are thinking about starting a for-profit company, a social enterprise or a charity. Perhaps you want to work at a startup to gain valuable experience. Perhaps you have your own idea, or want to start one of ours. There have never been more opportunities.

Each year at Charity Entrepreneurship; we speak with several hundred entrepreneurs, consultants, aspiring founders and potential staff of startups. Often we coach them to gain clarity about their personal fit, the impact they could have, and their options. Some join our incubation program and start new ventures, some we introduce to highly relevant people in our network and some we connect to job opportunities in the community.

So if you’re interested in working in or building a startup, and you want to do as much good as you can, we’d love to help you.

APPLY FOR CE CAREER COACHING HERE

How we help

If you are just starting out, changing careers, have significant experience or just have questions or uncertainties; we are here to help. With one to one zoom coaching we will help you think it through. We can explore your fit for different options, discuss their potential relative impact, and identify how you might get more information in order to make the best possible decision for you. We may suggest additional paths or ideas, and address your most important unknowns. Where possible we’ll work with you to identify concrete next steps.

How it works

1. Fill out our Application Form. 2. We'll try to speak to ~three-five applicants a week and keep you informed if we can't offer you assistance. If we think we can help, we’ll send you a link to book your call and ask for some additional details to get you thinking. You can even take our quiz here to assess how good a fit you’d be for founding a charity. 3. We’ll connect you with a suitable coach from the CE team for a zoom call (typically 30 mins, with the potential for follow up). You’ll talk through your career and clarify your next steps.

We’re most helpful for people who:

Want to do good with their careers and are significantly interested in Effective Altruism, and entrepreneurship or senior positions at existing organizations

Have the drive to start something or want to join an early-stage or growing organization, and help steer its progress

Are aware of the kinds of organizations we help launch (i.e. those that aim to be at least five times as cost-effective as the Against Malaria Foundation)

What is outside of our scope:

General EA career advising

Cause-specific coaching not related to entrepreneurship

For animal advocacy we recommend Animal Advocacy Careers, for longtermism check out 80,000 Hours, for global health & development Probably Good

Even if you are not 100% sure if you should apply, we recommend that you do. It’s a short form that may alone help you clarify your thoughts.

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: "Big tent" effective altruism is very important (particularly right now), published by Luke Freeman on May 20, 2022 on The Effective Altruism Forum.

This August, when Will MacAskill launches What We Owe The Future, we will see a spike of interest in longtermism and effective altruism more broadly. People will form their first impressions – these will be hard to shake.

After hearing of these ideas for the first time, they will be wondering things like:

Who are these people? (Can I trust them? Are they like me? Do they have an ulterior agenda?)

What can I do (literally right now and also might it might my decisions over time)?

What does this all mean for me and my life?

If we're lucky, they'll investigate these questions. The answers they get matter (and so does their experience finding those answers).

I get the sense that effective altruism is at a crossroads right now. We can either become a movement of people who appear dedicated to a particular set of conclusions about the world, or we can become a movement of people that appear united by a shared commitment to using reason and evidence to do the most good we can.

In the former case, I expect to become a much smaller group, easier to coordinate our focus, but it's also a group that's more easily dismissed. People might see us as a bunch of nerds who have read too many philosophy papers and who are out of touch with the real world.

In the latter case, I'd expect to become a much bigger group. I'll admit that it's also a group that's harder to organise (people are coming at the problem from different angles and with varying levels of knowledge). However, if we are to have the impact we want: I'd bet on the latter option.

I don't believe we can – nor should – simply tinker on the margins forever nor try to act as a "shadowy cabal". As we grow, we will start pushing for bigger and more significant changes, and people will notice. We've already seen this with the increased media coverage of things like political campaigns and prominent people that are seen to be EA-adjacent.

A lot of these first impressions we won't be able to control. But we can try to spread good memes about EA (inspiring and accurate ones), and we do have some level of control about what happens when people show up at our "shop fronts" (e.g. prominent organisations, local and university groups, conferences etc.).

I recently had a pretty disheartening exchange where I heard from a new GWWC member who'd started to help run a local group felt "discouraged and embarrassed" at an EAGx conference. They left feeling like they weren't earning enough to be "earning to give" and that they didn't belong in the community if they're not doing direct work (or don't have an immediate plan to drop everything and change). They said this "poisoned" their interest in EA.

Experiences like this aren't always easy to prevent, but it's worth trying.

We are aware that we are one of the "shop fronts" at Giving What We Can. So we're currently thinking about how we represent worldview diversity within effective giving and what options we present to first-time donors. Some examples:

We're focusing on providing easily legible options (e.g. larger organisations with an understandable mission and strong track record instead of more speculative small grants that foundations better make) and easier decisions (e.g. "I want to help people now" or "I want to help future generations").

We're also cautious about how we talk about The Giving What We Can Pledge to ensure that it's framed as an invitation for those who want it and not an admonition of those for whom it's not the right fit.

We're working to ensure that people who first come across EA via effective giving can find their way to the actions that best fit them (e.g. by introducing them to the broader EA community).

We often cros...

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: Covid 5/19/22: The Law of Five, published by Zvi on May 19, 2022 on LessWrong.

If you’re five, you now get a booster. If you’re four, you still get no vaccine at all (while we ‘wait for more data’) and you’re often forced to take precautions that might make sense for senior citizens.

Because no one cares. And that truly continues to be the overall message. No one cares about the physical prevention or mitigation of Covid. Congress allocates zero dollars, doctors don’t prescribe Paxlovid and life mostly continues as normal because Omicron in a mostly-already-infected-and/or-vaccinated population simply isn’t that dangerous. Cases go up, but life goes on.

It’s pretty terrible that we’re not funding pandemic prevention or anything for the current pandemic, but at this point I don’t know that the most likely alternatives are even improvements.

(And of course, if you’re under one, thanks in large part to the FDA you may have also have a problem getting food.)

Florida did not report this week. I am retroactively smoothing out Florida’s numbers by giving 50% of its reporting weeks to its no-reporting weeks, now that this has happened three times in a six-week period. All future numbers and predictions will include this adjustment (assuming I remember to do it) and I will assume up front that when Florida reports we take 50% off the top. If I find time I’ll do a smarter adjustment, since this one is slightly flawed, but it’s much better than doing nothing.

Executive Summary

Cases up.

Boosters available for ages 5-11.

No one cares and deaths are not rising yet.

Let’s run the numbers.

The Numbers

Predictions

Prediction from last week: 640,000 cases (+23%) and 2,100 deaths (+1%)

Results: 600,742 cases (+23% after adjusting Florida for both weeks, would have been +15% without the adjustment) and 2,004 deaths (+2% after Florida adjustment, -3% without the adjustment).

Prediction for next week:

Deaths

Deaths are rising in the Northeast and flat elsewhere, despite it being more than three weeks since cases began rising again. It seems clear that case fatality rates in this wave are going to be lower. It’s not great that thousands of Americans with Covid are dying each week, some of them from Covid as well. I wish it were not so, but I am glad we are not ruining our lives trying to stop this.

Cases

Enough weeks in a row with 20% weekly case growth and things get out of hand. We’re now triple our low point and back at a point similar to where we’ve been for most of the past year. I expect case growth to peak well before we get back to where things were in January, and even if we did get those case numbers back I expect less death to result from it, but certainly this is not where I hoped or expected to be in May.

The rise of BA.2.1.12 (and BA.4 and BA.5) are something I wasn’t predicting would happen, or at least would not happen so fast. Variants are inevitable, but this was unexpectedly quick, and Topol’s note below that this implies more future variants is well taken.

Is the pandemic over? In terms of cases, we have a clear no. In terms of deaths and people’s reactions, that remains to be seen. I still consider my pandemic over.

Physical World Modeling

Eric Topol describes our current approach as The Covid Capitulation. That name seems apt, for both good aspects and bad. The first half of the post is about how B.2.1.12 and B.4/5 bring both the new mutations and doom.

To recap, we have a highly unfavorable picture of: (1) accelerated evolution of the virus; (2) increased immune escape of new variants; (2) progressively higher transmissibility and infectiousness; (4) substantially less protection from transmission by vaccines and boosters; (5) some reduction on vaccine/booster protection against hospitalization and death; (6) high vulnerability from infection-acquired immunity only;...

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: Most students who would agree with EA ideas haven't heard of EA yet (results of a large-scale survey), published by Lucius Caviola on May 19, 2022 on The Effective Altruism Forum.

TL;DR

In a large-scale survey we found that only 7.4% of New York University students knew what effective altruism (EA) is. At the same time, 8.8% were extremely sympathetic to EA. These students agreed with basic EA ideas when presented with a short introduction, showed interest in learning more about it, and scored highly on our ‘effectiveness-focus’ and ‘expansive altruism’ measures. Interestingly, these EA-sympathetic students were largely ignorant about EA; only 14.5% knew about it before the survey. We estimate that, in total, 7.5% of NYU students do not know of EA but would be very sympathetic to the basic ideas. These findings could have important implications for the scaling of outreach but need to be interpreted carefully.

Note: This post is long because of the Detailed results section. Most readers may just want to focus on the Key results and Conclusions sections.

Key results

We conducted an approximately representative online survey with 938 students at NYU to investigate their familiarity with and views on EA and existential risk.

Familiarity

When asked to explain what EA is, only 7.4% of NYU students could clearly demonstrate that they knew it. An additional 10.6% demonstrated that they may have heard of it, but their definition wasn’t sufficiently accurate (see below). The remaining 82.1% clearly did not know about EA.

Views on effective altruism

After reading a short introduction about EA (see below), 25.0% agreed with the key ideas relatively strongly, and 82.1% of those respondents expressed interest in learning more about EA.

We found that 8.8% of students were especially sympathetic towards EA since they agreed with the basic principles, wanted to learn more, and scored highly on “effectiveness-focus” and “expansive altruism,” two psychological traits which seem to predict EA engagement.

(When we mention ‘EA-sympathetic’ students in the post, it is this group we refer to. Please keep in mind that students who didn’t meet the ‘very sympathetic’ bar didn’t necessarily dislike EA ideas).

Unreached EA-sympathetic students

7.5% were unaware of EA, yet were very sympathetic when first presented with the ideas. Since there are about 50,000 NYU students in total, we estimate there are about 3,800 such students. (Of course, as discussed below, not all of them would become highly engaged EAs. Conversely, some students who are not immediately in agreement with EA may be interested after thinking more about the ideas.)

Predictors

There were few robust demographic predictors of EA agreement. Neither gender, SAT scores, nor most study subjects significantly correlated with it. Exceptions are that students studying arts and media or humanities tended to agree slightly less with EA than other students and that older students tended to agree slightly more. And the most EA-sympathetic students tended to be more politically liberal than other students.

These results could be seen as good news for attempts to make EA more diverse. They suggest that different groups are sympathetic toward the basic EA ideas. But keep in mind that immediate agreement is not the only predictor of becoming more engaged with EA (see below).

Older students, those with higher SAT scores, and students studying STEM or social sciences were more likely to already know of EA.

Unsurprisingly, students with greater intellectual interest in people and resources related to EA (as well as other philosophical topics) tended to agree more with EA and were more likely to know of it.

Reactions to existential risk

After reading a short introduction about existential risk (see below), 25.7% agreed relatively strongly that exist...

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: Why should I care about insects?, published by JamieGittins on May 18, 2022 on The Effective Altruism Forum.

It is currently common for people to feel little or no duty of care towards insects. In this post, I make a case for why I think expanding our circle of compassion to insects is both rational, and important.

I think the arguments also apply to other invertebrates such as shrimp and gastropods (slugs and snails) but for simplicity, I’ve kept this post focused on insects.

Why is this important?

The sheer number of insects whose lives humans have the power to affect is huge. A few bits of relevant info to demonstrate the scale:

Around 1 trillion insects are raised and slaughtered on farms annually for food and animal feed. A single farm could have 29 billion insects alive at any one time. And the insect farming industry is looking to expand rapidly in coming years, despite concerns about welfare.

In the wild there are an estimated 10 quintillion (10 with 18 zeros!) insects alive at any given time. Many of these are harmed by pesticides and other human activities.

Across the world, it is generally legally and even socially acceptable to harm insects. For instance, the boiling of silkworms alive is part of the process for making most silk. As a kid, I owned a children’s ‘fly-splatter’ gun which was sold as a fun toy.

Despite the huge numbers of insects who are farmed and harmed by humans each year, the topic of insect welfare is currently highly neglected. There are very few organisations working on the issue (although shoutout to Rethink Priorities, who have produced some great research) and I'd be surprised if even 0.01% of the animal advocacy movement's resources are currently going into insect welfare work.

So, to summarise: the scale is huge, the welfare issues often serious, and the situation completely neglected.

A selection of arguments

The aim of this post is not to go deep into the science or philosophy of insect sentience. Rather, it is to offer a brief selection of arguments in favour of caring about insects.

You may not agree with every single argument in this post. But my hope is that, as a whole, they make a case for each of us to feel a deeper sense of moral concern for insects.

1. The science: behaviour

The scientific literature on the behaviour of insects (a hugely diverse class of animals) is generally under-explored.

A great in-depth review of what we know about invertebrate behaviours, including insects, can be found here. I think the key takeaway is that invertebrates are capable of some quite surprising and complex behaviours, which suggests they are conscious of the world.

Just a tiny selection of studies demonstrating these behaviours in insects:

Honeybees have displayed pessimistic cognitive biases after being exposed to a threat, as well as positive emotion-like states following exposure to unexpected rewards.

Funnel ants were able to use absorbent objects to transport sugar water back to their nest, quickly learning that sponges were the most absorbent material and using them preferentially.

Fruit flies are likely affected by painkillers in a similar manner to humans and are willing to endure a cost (e.g. an electric shock) in order to receive a reward.

Of course, none of these studies prove that insects are conscious. But, the existence of such complex, flexible and emotion-like behaviours suggests that they are not simply carrying out an unconscious pre-programmed set of behaviours.

2. The science: nervous systems

Again, the current science on this is pretty spotty. However, Rethink Priorities have done some excellent work collecting the evidence on this. I've also written on it, but Rethink's stuff is way more comprehensive!

I think there are two key takeaways from what we know:

It's not clear that a 'large' brain is required to suppo...

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: Are you really in a race? The Cautionary Tales of Szilárd and Ellsberg, published by HaydnBelfield on May 19, 2022 on The Effective Altruism Forum.

OR The Tragedy of the Einstein Letter and the Gaither Report; Cautionary Lessons from the Manhattan Project and the ‘Missile Gap’; Beware Assuming You’re in an AI Race; The illusory Atomic Gap, the illusory Missile Gap and the AGI Gap

Summary

In both the 1940s and 1950s, well-meaning and good people – the brightest of their generation – were convinced they were in an existential race with an expansionary, totalitarian regime. Because of this belief, they advocated for and participated in a ‘sprint’ race: the Manhattan Project to develop a US atomic bomb (1939-1945); and the ‘missile gap’ project to build up a US ICBM capability (1957-1962). These were both based on a mistake, however - the Nazis decided against a Manhattan Project in 1942, and the Soviets decided against an ICBM build-up in 1958. The main consequence of both was to unilaterally speed up dangerous developments and increase existential risk. Key participants, such as Albert Einstein and Daniel Ellsberg, described their involvement as the greatest mistake of their life.

Our current situation with AGI shares certain striking similarities and certain lessons suggest themselves: make sure you’re actually in a race (information on whether you are is very valuable), be careful when secrecy is emphasised, and don’t give up your power as an expert too easily.

I briefly cover the two case studies, discuss the atmosphere at RAND, then draw the comparison with AGI and explain my three takeaways. This short piece is mainly based on Richard Rhodes’ The Making of the Atomic Bomb and Daniel Ellsberg’s The Doomsday Machine. It was inspired by a Slack discussion with Di Cooke.

The ‘atomic gap’, the Einstein-Szilárd Letter, and the Manhattan Project

Here is a rough timeline of some key events around the Manhattan Project:

12 September 1933: Szilárd conceives the idea of a nuclear chain reaction, and keeps it a secret for the next six years.

2 August 1939: Einstein-Szilárd letter to Roosevelt advocates for setting up a Manhattan Project.

1 September 1939: Nazi invasion of Poland.

9 October 1941: Roosevelt approves the atomic program, subsequently Manhattan Project receives serious funding (eventually, 0.4% of GDP).

June 1942: Hitler decides against an atomic program for practical reasons.

December 1942: First chain reaction in Chicago. Szilard notes “I shook hands with Fermi and I said I thought this day would go down as a black day in the history of mankind.”

30 April 1945: Hitler kills himself.

7 May 1945: Nazi surrender.

July 1945: Szilárd petition (signed by 70 scientists) calls for the bomb to be used only after Japan had refused to surrender, and for the decision to be made by Truman personally – and reiterated the original intention was to defend against the Nazis.

6 and 9 August 1945: The USA bombs Hiroshima and Nagasaki.

29 August 1949: First successful Soviet nuclear test.

The Manhattan Project (and other US projects like the Apollo program) was a ‘sprint’ project, in which the peak year funding reached 0.4% of GDP (Stine, 2009, see also Grace, 2015).

Why did nuclear scientists like Szilárd, who kept the chain reaction secret and opposed nuclear weapons for decades after the war, advocate for and participate in the Manhattan Project? In Ellsberg’s words:

“How could he? The answer is he believed, even before others, that they were racing Hitler to the attainment of this power. It was German scientists, after all, who had first accomplished the fission of a heavy element. There seemed no reason to suppose that Germany could not stay ahead of any competitors in harnessing this unearthly energy to Hitler’s unlimited ambitions for conquest. The specter of a possible Germ...

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: Reading the ethicists: A review of articles on AI in the journal Science and Engineering Ethics, published by Charlie Steiner on May 18, 2022 on LessWrong.

Epistemic status: Stream of consciousness reactions to papers read in chronological order. Caveat lector.

I have a dirty family secret. My uncle is a professional ethicist.

In a not-too roundabout way, this is why I ended up looking at the October 2020 issue of the journal Science and Engineering Ethics, their special issue on the ethics of AI. I am now going to read that issue, plus every article this journal has published about AI since then [I wussed out and am just going to skim the latter for ones of special interest] and give you the deets.

October 2020

Hildt et al., Editorial: Shaping Ethical Futures in Brain-Based and Artificial Intelligence Research

This is the introduction to the issue. They give each paper a sentence or two of summary and try to tie them all together. The authors helpfully give a list of topics they think are important:

Data Concerns: Data management, data security, protection of personal data, surveillance, privacy, and informed consent.

Algorithmic Bias and Discrimination: How to avoid bias and bias related problems? This points to questions of justice, equitable access to resources, and digital divide.

Autonomy: When and how is AI autonomous, what are the characteristics of autonomous AI? How to develop rules for autonomous vehicles?

Responsibility: Who is in control? Who is responsible or accountable for decisions made by AI?

Questions relating to AI capabilities: Can AI ever be conscious or sentient? What would conscious or sentient AI imply?

Values and morality: How to build in values and moral decision-making to AI? Are moral machines possible? Should robots be granted moral status or rights?

Based on this list, I anticipate that I'm about to run into four-sixths ethics papers about present-day topics that I will skim to point out particularly insightful or anti-insightful ones, one-sixth philosophers of mind that I will make fun of a little, and one-sixth papers on "How to build values into general AI" that I'm really curious as to the quality of.

Onward!

Nallur, Landscape of Machine Implemented Ethics

Primarily this paper is a review of a bunch of papers that have implemented or proposed ethics modules in AI systems (present-day things like expert systems to give medical advice, or lethal autonomous weapons [which he has surprisingly few qualms about]). These were mostly different varieties of rule-following or constraint-satisfaction, with a few handwritten utility functions thrown in. And then one of these is Stuart Armstrong (2015) for some reason - potentially that reason is that the author wanted to at least mention "value-loading," and nobody else was talking about it (I checked - there's a big table of properties of different proposals).

It also proposes evaluating different proposals by having a benchmark of trolley-problem-esque ethical dilemmas. The main reason this idea won't work is that making modern-day systems behave ethically involves a bunch of bespoke solutions only suitable to the domain of operation of that system, not allowing for cross-comparison in any useful way.

If were to salvage this idea, we might wish to have a big list of ethical questions the AI system should get the right answer to, and then when building a sufficiently important AI (still talking about present-day applications), the designers should go through this list and find all the questions that can be translated into their system's ontology and check that their decision-making procedure gets acceptable answers. E.g. "Is it better to kill one person or two people?" can become self-driving car scenarios where it's going to hit either one or two people, and it should get the right answer, bu...

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: Early spending research and Carrick Flynn, published by AllAmericanBreakfast on May 19, 2022 on The Effective Altruism Forum.

In the main EA Forum post on the Flynn Congressional campaign loss, a comment thread got into the research on primary campaign spending.

MHarris: How much did the $13 million shift the odds? That's the key question. The conventional political science on this is skeptical that donations have much of an effect on outcomes (albeit it's a bit more positive about lower profile candidates like Carrick)/...

JoshYou: Fundraising is particularly effective in open primaries, such as this one. From the linked article:

"But in 2017, Bonica published a study that found, unlike in the general election, early fundraising strongly predicted who would win primary races. That matches up with other research suggesting that advertising can have a serious effect on how people vote if the candidate buying the ads is not already well-known and if the election at hand is less predetermined along partisan lines."

Owen Cotton-Barratt: Although early fundraising could be correlational with success rather than causal, if it's an indicator of who can generate support from the electorate...

Here's a few key points from Bonica 2017 (main link, sci-hub link), along with spending data for the Salinas and Flynn campaigns.

Summary:

Money predicts victory in contested primaries. In contested primaries (like Carrick was involved in), the candidate who raises the most money wins.

Early fundraising deficit is predictive of dropping out of the race in the first 90 days.

Fundraising brings more support from professional political organizations like the DCCC.

"One challenge in estimating campaign spending effects is that fundraising is likely endogenous to the probability of winning. If donors are actively targeting candidates who are most likely to be elected, causality could flow in reverse."

Bonica's causal model by which money buys elections is never crisply stated, but has two main parts. First, early money keeps candidates from dropping out early on. Second, it garners them support from professional political organizations and enables them to buy the advertising and other things their campaign needs. However, this only matters in elections where there's a real chance for any candidate to win, such as contested primaries.

According to New Statesman, by April 15th, Flynn had received 60% more itemized donation money than his opponent, Salinas, but 80% of Salinas's itemized donations were from in-state donors, as compared to 2.5% of Flynn's itemized donations.

In terms of outside spending, Flynn had by the end received $12,329,074 to Salinas's $1,689,530. Flynn received about 7x more outside money than Salinas.

Here's Bonica's regression for the probability of victory vs. the candidate's share of total fundraising.

Flynn had about a 90% total fundraising share, so, eyeballing it, this model would have predicted he'd have about an 85% chance of victory. Instead, his main opponent, Salinas, received about twice as many votes as Flynn.

Bonica didn't look at the degree to which a candidate's share of in-state money predicted their chance of victory. I'd be very curious to see such an analysis. We may also have a case of reverse causality on our hands, in which Sam Bankman-Fried, Flynn's largest donor, targeted him not for his chance of winning, but for his alignment with SBF's own political agenda.

Based on Bonica's proposed mechanism by which money can influence a contested primary, it seems like a little seed money is what does the trick, not a tsunami of cash.

Musings:

If this "seed, not a tsunami" model is right, then this suggests a political strategy for consideration by EAs. EAs with strong pre-existing ties to their local community should run for office widely. SBF and other do...