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: Community in six hours, published by Gavin on April 1, 2022 on The Effective Altruism Forum.

Too long have artisanal extraverts dominated community building. Time, my friends, to optimise.

1. Automation of community

Our throughput is woeful. Yes - we have crude and simplistic metrics, like 'number of GWWC pledges signed', and this helps. Yes - we have introductory fellowship scripts for engaging new people without reflection. But we still produce a paltry ~103 people per year.

Notice we have everything we need for massive efficiency gains though! We have input data and a loss function. We can thus remove the community builder from the inner loop. Phase 1 of the automation is simply a monitor which continuously plays the intro script and records the initiate's responses. No CB needed!

But then the response time of the prey is the limiting factor. Phase 2 is a good start: a conveyor belt leading from fresher's fairs to the proselytism bot; or perhaps some kind of giant funnel strapped to the side of the philosophy department.

Phase 3 removes the initiates entirely. Facebook have a detailed predictive model of each of its users and non-users. I propose to use such models to simply predict whether someone would sign the GWWC pledge. We input our intro script to the model, and if the model outputs "y", then we digitally mint a new pledge for the person. With suitable parallelisation, this should produce 1.4 billion pledges in just over 1 second. At current meatspace clock speeds, this success would take community builders 100 years of work, even under generous assumptions about the recursive expansion of CB.

2. Community building building

Producing EAs is good - but contemptibly close to the object level. We can go deeper.

We want CBs to produce more CBs. (CB does this, but imagine what we could achieve if we did it on purpose.)

People who invite community builders on retreats are community builder builders, CB2. Then people who fund community builder builders are CB3, and so on.

Take the annual output of community building, and simply amplify it: let each CBn produce a next-level CBn+1, until we reach saturation. My model, with Anders Sandberg, estimates that all matter in the solar system could be converted to CB1s within 6 hours of the creation of a single CB13.

3. Automating community building building

The obvious thing is to remove the final inefficiency, myself, and to let GPT finish this post:

What would we do if we didn't have a community-building model? I'm very grateful to my fellow community builders for creating the best CB with the best design. I'm just happy to say I've done it. If you like me, feel free to do something and keep improving. As for other areas, I don't think we're going to need them. Thanks for reading, and I hope you are as cool as I am!

Unrelated fact: the Linux utility "yes" outputs about 1 billion "y" characters per second.

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: Save the Date: EAGxMars, published by OllieBase on April 1, 2022 on The Effective Altruism Forum.

CEA is excited to announce that EAGxMars will take place in October 2040. We look forward to welcoming ambitious members of the community for a weekend of talks, networking and civilisation-building on this relatively nearby planet.

Applications will open in 2030 to give attendees sufficient time to make their journeys physically possible. We will make a slack channel to help people coordinate on this.

Who is EAGxMars for?

EAGxMars will be a locally-organised conference designed primarily for people:

Familiar with the core ideas of effective altruism;

Interested in learning more about what to do;

Who can get to Mars.

What’s the difference between a “normal” EAGx and EAGxMars?

The target audience is different - EAGx events are for people comfortable with planes, trains and other parochial modes of transportation. EAGxMars is for people who are ambitious.

How much are tickets? Is there financial aid?

We will offer discounts and financial support to make sure that cost is not prohibitive for anyone. The whole thing is probably prohibitive enough as it is.

Will I need a visa?

We don’t yet have this information available.

Precise dates TBC. These kinds of details aren’t top of mind for us at present.

It’s possible that attendees will plan on staying for longer than a weekend, given the journey. We might organise some retreats around the event.

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: New addition to the 80,000 hours 1-on-1 team., published by alexrjl on April 1, 2022 on The Effective Altruism Forum.

As many of you know, 80,000 hours has been trying to grow the 1-on-1 team to be able to talk to many more people and minimise the waiting time after submitting an application. We are excited to announce that we have made a successful new hire.

From our user interviews and marketing research, we know that people can sometimes find applying for 80,000 hours or thinking about talking to our advisors intimidating. This advisor sets themselves apart by requiring no preparation whatsoever, and is deeply unintimidating.

They have no degree or experience in the EA community (their background is in software engineering) but the team was enchanted by their winning smile. If you think there’s a good match between their skills and your needs, you can request them as an advisor, and even ask them for help writing your application (we recommend recording your zoom call with them to capture the brilliant insights that come out of the conversation).You can read more about them here.

I haven't worked out how to embed images in EA forum posts, but after talking to them for a while came up with the solution of just linking to a headshot instead.Along with the rest of the team at 80,000 hours, I'm immensely grateful to Chana Messinger, who suggested Duck as a candidate.

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: Replacing Karma with Good Heart Tokens (Worth $1!), published by Ben Pace on April 1, 2022 on LessWrong.

Starting today, we're replacing karma with Good Heart Tokens which can be exchanged for 1 USD each.

We've been thinking very creatively about metrics of things we care about, and we've discovered that karma is highly correlated with value.

Therefore, we're creating a token that quantifies the goodness of the people writing, and whether in their hearts they care about rationality and saving the world.

We're calling these new tokens Good Heart Tokens. And in partnership with our EA funders, we'll be paying users $1 for each token that they earn.

"The essence of any religion is a good heart [token]." — The Dalai Lama

Voting, Leaderboards and Payment Info

Comments and posts now show you how many Good Heart Tokens they have.

(This solely applies to all new content on the site.)

At the top of LessWrong, there is now a leaderboard to show the measurement of who has the Goodest Heart. It looks like this. (No, self-votes are not counted!)

The usernames of our Goodest Hearts will be given a colorful flair throughout the entirety of their posts and comments on LessWrong.

To receive your funds, please log in and enter your payment info at lesswrong.com/payments/account. We pay out once a day at 11:59 pm PST.

While the form suggests using a PayPal address, you may also add an Ethereum address, or the name of a charity that you'd like us to donate it to.

Why are we doing this?

On this very day last year, we were in a dire spot.

To fund our ever-increasing costs, we were forced to move to Substack and monetize most of our content.

Several generous users subscribed at the price of 1 BTC/month, for which we will always be grateful. It turns out that Bitcoin was valued a little higher than the $13.2 we had assumed, and this funding quickly allowed us to return the site to its previous state.

Once we restored the site, we still had a huge pile of money, and we've spent the last year desperately trying to get rid of it.

In our intellectual circles, Robin Hanson has suggested making challenge coins, and Paul Christiano has suggested making impact certificates. Both are tokens that can later be exchanged for money, and whose value correlates with something we care about.

Inspired by that, we finally cracked it, and this is our plan.

...We're also hoping that this is an initial prototype that larger EA funders will jump on board to scale up!

The EA Funding Ecosystem Wants To Fund Megaprojects

"A good heart [token] is worth gold."— King Henry IV, William Shakespeare

Effective altruism has always been core to our hearts, and this is our big step to fully bring to bear the principles of effective altruism on making LessWrong great.

The new FTX Future Fund has said:

We’re interested in directly funding blogs, Substacks, or channels on YouTube, TikTok, Instagram, Twitter, etc.

They've also said:

We’re particularly interested in funding massively scalable projects: projects that could scale up to productively spend tens or hundreds of millions of dollars per year.

We are the best of both worlds: A blog that FTX and other funders can just pour money into. Right now we're trading $1 per Good Heart Token, but in the future we could 10x or 100x this number and possibly see linear returns in quality content!

Trends Generally Continue Trending

Paul Christiano has said:

I have a general read of history where trend extrapolation works extraordinarily well relative to other kinds of forecasting, to the extent that the best first-pass heuristic for whether a prediction is likely to be accurate is whether it's a trend extrapolation and how far in the future it is.

We agree with this position. So here is our trend-extrapolation argument, which we think has been true for many years and so will continue to...

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 actual longtermist incubation program, published by GidonKadosh on April 1, 2022 on The Effective Altruism Forum.

It is often discussed that EA is talent-constrained, and the movement is experimenting with ways to create new leaders for the most pressing cause areas.

Since the launch of Naming What We Can, our linguistically-talented team conducted in-depth analysis in search of the most promising and neglected opportunities within this space.

Our analysis shows that even though incubation programs are considered outside EA to be one the most promising ways to foster new talent, there is not a single incubation program within the EA ecosystem.

Project Plan

The first, and hardest part of the program, is finding strong applicants. The incubator participants would stay in the incubator facility for ~18 years and will be able to learn about EA and develop domain expertise.

The new people incubated at our incubator will be called X-risk-Men.

At some part of their growth, each X-risk-Man would realize they have an EA-superpower. Some will be super-forecasters. Others will be able to create QALYs out of thin air. In the most extreme cases, some might even be able to discuss Roko's Basilisk without putting everyone close to them in danger.

When their incubation is over, the X-risk-Men will all be automatically accepted to Charity Entrepreneurship (or has been renamed by Naming What We Can - Charity Entreprenreurooshrimp) and start super-effective charities.

As others have noted, “EA should focus on being a really good place for a relatively small group of unusual people to try to be extremely impactful”. And there is nothing weirder than superpowers, which definitely have not gone mainstream.

Impact estimation

Overall, we think the impact of the project will be net negative on expectation (see our Guesstimate model). That is because we think that the impact is likely to be somewhat positive, but there is a really small tail risk that we will cause the termination of the EA movement. However, as we are risk-averse we can mostly ignore high tails in our impact assessment so there is no need to worry.

Call to Action

In order to begin the incubation program, we need local EA groups to identify members with unusual talents or who are otherwise strange. We expect that very few EAs are strange, so this may be difficult, but with effort, we think that most EA groups (and maybe some EA orgs) can identify at least one such member and nominate them in a comment below.

Once all members are chosen, we will secretly clone them and raise the clones in the new actual incubator.

Many thanks to David Manheim, Guy Raveh, Omri Sheffer, Edo Arad, and Yuval Shapira for contributing to this new important project, as well as many members of EA Israel (some of them have actually been through the trials in the Israeli desert).

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: Career assignment by birthdate, published by Mathieu Putz on April 1, 2022 on The Effective Altruism Forum.

OpenPhil and the FTX Future Fund employ some of the very brightest people in EA to decide where to allocate their money. For human capital, the process is a lot messier: individuals just choose whatever they think is best. No expert decision makers. How unsystematic! And they each have to do their own research to come to their own conclusions. How inefficient!

Imagine if instead, expert career allocators could carefully deliberate and centrally decide exactly what fraction of EA's human capital budget should be devoted to what cause area. That is the goal of this project.

80,000 hours currently do something vaguely like this, but their process is incredibly imprecise and expensive. They need to talk to hundreds of individuals per year, many of whom may not follow their advice at all!

How else could the community decide which person should go down which career path?

I expect most cause areas to have similar overall needs: each will need quantitative experts, operations managers, leaders, and so on. Thus, to ensure this talent distribution gets evenly allocated across cause areas, a random allocation process would be ideal. I propose assignment by birth date.

To a first approximation, birth dates are fairly random: there is little reason to expect potentially talented operations managers to be disproportionately born in one part of the year. And they are unambiguous. Everyone already has a birth date and it's difficult to lie about or change.

Hence, I suggest creating a central dashboard where 80,000 hours publish their recommended career allocations as a function of birth date. Here's an example of what that might look like[1]:

AI saftey: Everyone born from January 1st until April 17th (29%)

Biorisk: Everyone born from April 18th until June 3rd (13%)

Global Health and Wellbeing: Everyone born from July 4th until August 8th (18%)

Animal Welfare: Everyone born from August 9th until September 20th (11%)

... [Many other cause areas, all of which are dear to my heart, just as the above.]

Stella, who was born on the 2nd of March 2001, would know to go into AI safety. No need to research any thorny philosophical or empirical questions to make the decision, it's all been taken care of.

Of course, 80,000 hours might change their minds about the optimal allocation. In fact, big community-wide debates would be held over whether thresholds should be moved. A popular petition might advocate for moving the threshold date for AI safety to February 15th, thereby massively reducing the community's investment.

Yet, everyone would recognize the necessity to cooperate and play by the rules. EA orgs would generally refuse to accept candidates with the wrong birthdates.

I am eager for suggestions about what to do with people who have already started going down a career path, when threshold dates are changed. I tentatively think that the enormous benefits offered by this system would easily outweigh the career transition costs, but I'm open to other creative solutions.

To conclude, the time to make career markets more efficient is long overdue. We need to avoid accidentally overcrowding or neglecting any cause areas. Assigning careers by birthdates provides an easy and, in hindsight, obvious solution, allowing the community as a whole to make wiser decisions about its human capital.[2]

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: Announcing Impact Island: A New EA Reality TV Show, published by Linch on April 1, 2022 on The Effective Altruism Forum.

Impact Island is the hip new reality TV show that all of your cool, chic, and longtermist impact-oriented friends would love to binge watch. Think "Survivor" meets "Shark Tank" meets "Open Philanthropy."

Premise

The premise is simple:36 young humble aspiring effective altruists are invited to a private Caribbean Island. Each aspiring EA is expected to do high-impact projects during their stay here. They will also gossip and continuously publicly rate each other's projects, impact, and likability. Every week, an anonymous team of grantmakers rank all participants, and whoever accomplished the least morally impactful work that week will be kicked off the island.

At the end of the season (36 episodes), the single remaining most impactful participant will be given 10 million dollars of seed funding for their new org.

How will you select participants?

Participants will be chosen via a formal selection process combining past track records, good looks, EA promisingness, movie chemistry, reference letters, and the coherent extrapolated gut feels of leading EA grantmakers. Of course, all participants are expected to be polyamorous.

How much is this project? Do you need more funding?

Our current budget is $25 million dollars. We have already secured funding from reputable EA sources, such as the FTX Future Fund, Open Philanthropy, Longview Philanthropy's Media Engagement arm, actor Joseph Gordon-Levitt, and real estate mogul with political and reality TV experience Donald Trump.

However, 25 million is only enough money to tide us over for season 1 (36 episodes). As we (the producers) believe strongly in patient longtermism, we would ideally like to run for 3 quadrillion seasons. Thus, we have extremely large room for more funding, which we do not expect to fill any time soon, making us the premier community building EA megaproject.

Do you plan to seek additional sources of funding?

Yes, we believe in being self-financing as quickly as possible, so we plan to mint an NFT of every episode.

What is the expected impact of this project?

We think the expected impact of the project is so high it is literally incalculable. If we're forced to put numbers on this, we think we can get at least 10 million viewers, and conservatively 1% of those people (1 billion viewers) will counterfactually become EAs. This will be the largest and most successful community building megaproject EA has ever seen, and even if our impact calculation is substantially off (say by 25%), 750 million highly engaged EAs is still an impressive number.

In addition to the community building aspect, we also think that the direct impact of having 36 EAs do highly impactful projects can be quite high, and having them all continuously rate each other can help solve the vetting bottleneck in EA.

Further, we think the NFTs funding schema should quickly become self-financing, meaning that this project will pay for itself in no time (less than 3 geological eons in our techno-economic analysis).

How does this project fair in terms of importance, neglectedness, and tractability?

Importance

In addition to becoming a future cultural milestone (incalculable value), we also think this project has very high community building benefits, as well as direct impact.

Neglectedness

To the best of our knowledge, no other impact-oriented reality TV show in the Carribean exists, suggesting high neglectedness.

Tractability

We're very sure of the tractability of this project, having read skimmed multiple blog posts on how to run reality TV shows.

How will you make sure the TV show is fun to watch?

We think watching people doing high-impact altruistic activities in front of their computers is high utility (and thus fun) enou...

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: The case for infant outreach, published by Thomas Kwa on April 1, 2022 on The Effective Altruism Forum.

Epistemic status: Divine revelation.

EA community builders have recently launched programs targeting children as young as 12 or 13. This work is a substantial improvement on the status quo, but we think it is ineffective compared to outreach to even younger children. In this post, we make the basic case for infant outreach, then list several possible interventions.

Why infant outreach?

The simple answer is that infant outreach rates high on scale, tractability, and neglectedness.

Scale: the 0-5 age cohort is the largest five-year age cohort, containing 4.4% of the world population, and a substantial fraction of this is infants.

Neglectedness: To our knowledge, there are currently no large EA outreach efforts directed at infants. In the OpenPhil EA/LT survey of 217 longtermists, just 0% of respondents heard about EA or EA-adjacent ideas before age 2.

Tractability is obvious.

In addition, there are several factors that might make the scale of infant outreach even larger than expected:

Flynn effect (IQ steadily increases over time) means that average newborns are 4.24 IQ points higher than average 20-year-olds, and should be better alignment researchers.

Whereas high school and college students are often very busy, infants spend most of their time in unstructured play and have more time to learn about EA content.

Infant careers will be slightly closer to the future, making their uncertainty smaller and their actions higher leverage.

Infants are often sufficiently cute that they can substantially affect the opinions of their family and colleagues.

EA daycare groups

University groups have found success with the Intro Program and further cause-specific fellowships. Clearly, this model should also be applied to the world's top 20 daycares, targeting the cutest and most talented infants within those daycares. But since most infants cannot read, the standard Intro Program format will not work. I propose an adapted format: each week should be split into a readings section (where material is read to the students in a voice that is soothing yet serious) and discussions section, with nap-time in the middle.

To create highly engaged infants (HEIs), we also propose the Hercules Fellowship, a program creating a path for talented students under the age of two to drop out of daycare to pursue ambitious career opportunities. The Hercules Fellowship will provide financial and technical support to infants on this unconventional path, with the hope of matching students up with charity entrepreneurship ideas, research directions, and internships. The mythical hero Hercules created the Milky Way Galaxy as an infant, so we think this is a fitting name to inspire the most ambitious HEIs.

The Hercules Fellowship application can be as rigorous as other applications despite its target demographic. Applicants will have two toys placed in front of them, one of which represents a promising intervention. Applicants that consistently choose the promising toy will be selected for the fellowship.

Cause areaToy 1 (promising)Toy 2 (unpromising)Global health and developmentMalaria netPlayPumps model Animal welfare1,000 chickensStray dogTechnical AI alignmentAn ELK 🦌AI containment box 🤖📦Meta-EAFunding proposal for infant outreachFunding proposal for existing EA outreachBiosecurityText of an improved UN Biological Weapons ConventionText of the current UN Biological Weapons Convention

As the number of HEIs grows, we expect further room for investment into infants to increase their effectiveness; for example, we think Nonlinear is particularly well-positioned to hire nannies for top infants, and eventually makeup artists to increase the cuteness of infants in community-building roles.

EA toys and activities...

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: Is there a good summary of EA's impact to date?, published by Garrison on March 31, 2022 on The Effective Altruism Forum.

At EAG SF 2017, there was a bit at the end of Will's keynote about some of the impact of different EA projects/campaigns (e.g. number of cage free commitments).

I think it would be extremely helpful to have a webpage or doc summarizing the impact that the EA community has had to date. You can kind of back into this by looking at GiveWell and Open Phil's impact (the two orgs that I think have moved the most money so far). But this both misses a lot (e.g. EAA) while probably overstating impact (e.g. GiveWell preceded EA and plenty of their donors don't know what EA is).

I think we spend too much time talking about our ideas and not enough talking about what we've actually gotten done. In particular, projects that have had big impact that wouldn't have existed but for EA (probably) would likely be compelling to more "doers" than just the amount of money we've moved or the ideas that motivate us. I think we struggle to reach "doers" (we nail it with "thinkers") and this may be a contributor.

Forgive me if this exists and I missed it, but a quick search of the Forum, 80k, and the main EA site didn't turn this up.

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: A bicycle for your memory, published by sortega on March 30, 2022 on LessWrong.

Cross posted from sortega.github.io

Computers have the need to store huge amounts of information and, at the same time, being able to access that memory at blazingly fast rates. This poses the problem of how to get a fast and large memory when hardware memories are either fast, like registers or L1 cache lines; or big, like hard-drives or solid state drives.

MemorySizeSpeedCommentCPU registerA few bytesOne CPU cycle (200-300 picoseconds)A register holds the data the CPU can operate onRAM memoryA few gigabytes100 nanosecondsApplications need their data to be here to be responsiveHard driveMany gigabytes150 microsecondsAll your applications and data being kept locally have a copy hereInternetInfinite (for all purposes)10-40 millisecondsTechnically other people's hard drives

This is solved in computer architecture by the interplay of two ideas: a memory hierarchy and virtual memory. The memory hierarchy consists on a pyramid with the fastest and smallest memories on the tip and slower but bigger ones at the base. This arrangement allow you to have a complete copy of the information at the base, which is the biggest. Then, you can keep copies of the actively used information closer to the pyramid apex for better performance. Virtual memory is the illusion of having a single memory space which is the result of the collaboration of specific hardware (MMU, CPU features) with the operative system to transparently move data across the levels of the pyramid. The result is that, for most purposes and most of the time, programs running in your computer get to use a single memory space that feels as big as the base of the pyramid and almost as fast as its top.

Apart from computers, humans are rich information processors and memory plays a paramount role in what it means to be a human. On one hand it's an important element of personal identity. Imagine a movie villain using technology or black magic to swap the memories of two twin brothers. In some philosophical sense we are what we remember, and we will have a hard time deciding who is who in this contrived situation.

On the other hand, memories are needed for our minds to work because without them, we will have only the raw input of our senses and no other concept or idea to work with. The extent to which we can understand the world, communicate with others and act depends on using those mental building blocks or chunks.

Another aspect of being human not so directly related to information processing is tool making and tool use. Thanks to it, we have come a long way since we harnessed the power of fire and stone. Steve Jobs eloquently explained in this talk how humans will lose in almost all categories of an Olympic Games in which other animals were allowed to take part. We are not the fastest, we cannot travel the furthest, nor in the most efficient way. However, any cyclist is more efficient in energy per distance than a Condor, the most efficient animal. Then, he declared computers to be like a bicycle for the mind. Back then, personal computing was being rolled out, and we were oblivious to the directions these new machines will cybernetically extend us.

The Cyborg Memory Hierarchy

At this point we have all the ingredients to define what I call the cyborg memory hierarchy: extending human capabilities with technology (cyborg) so we can enhance in-brain memory and extend it with external information (memory hierarchy).

Level 0. Working memory

Our working memory is composed of a phonological loop, where you can keep repeating a phone number while you search for pen and paper; the visuo-spatial sketchpad, where you can mentally manipulate shapes; and some very limited number of slots that can hold ideas, or chunks, for manipulation. Early George A. ...