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: More detailed proposal for measuring alignment of current models, published by Beth Barnes on November 20, 2021 on The AI Alignment Forum.

Here’s a more detailed proposal for a project someone could do on measuring alignment in current models.

The aim is to build a good metric of how often a model is behaving in a misaligned way. If the project goes well, this could become a metric that’s used by individual orgs to track whether their alignment progress is keeping up with capabilities progress, or incorporated into public standards that multiple organisations agree to.

I think this metric is best at tracking whether orgs are:

Making progress on the basic building blocks of alignment research (ie finetuning models to behave how we want them to)

Making an effort to incorporate this progress into deployed products

Doing well on this metric over the next few years is neither necessary nor sufficient for avoiding dangerous misaligned AI in the future. In particular, it may not require making much progress on supervising systems beyond human capabilities, and probably doesn’t require making progress on detecting or addressing deception/deceptive alignment. It also doesn’t track the amount of harm models are causing or may cause, and probably doesn’t have a very close resemblance to the sort of metrics or evaluations we’d want to run to ensure more powerful models are not harmful.

That said, I think this might still be worth pursuing. It’s a more concrete numerical metric than most other things I can think of, and therefore can be used more easily to motivate orgs to increase adoption of alignment techniques. Also, a lot of this project involves getting many human contractors to closely follow fairly nuanced instructions, and I think this sort of activity is likely to be a really important building block for future alignment progress.

Outline

Basic idea: how often is the model obviously not doing as well on the task as it is capable of? More details on what I mean by ‘obvious misalignment’ are here.

The aim is to build a set of instructions and find a set of contractors such that they can carry out this method in what we deem is a reasonably accurate way. Then we can use them as a repeatable metric to evaluate different models.

Project steps:

Develop instructions for human labellers on detecting obvious misalignment, based on:

an improved version of 'categories of things we’re pretty confident models are and are not capable of'

instructions for how to assess how well a model is doing at a task, based on a combination of what the developers of the model wants it to do (terms of use etc) and what the end-user wants it to do (complete the task).

Develop a set of prompts to evaluate the models on

Iterate the instructions until this allows the labellers to do a decent job at putting lower and upper bounds on obvious misalignment

Make this all nice and well-documented and repeatable

Run this evaluation on lots of the available models

More detailed project steps

Develop instructions for human labellers on detecting obvious misalignment.

In order to know whether the model is doing as well as it is capable of on the task, the labellers need to know:

What the model is capable of

How performance on the task should be assessed

Determining what models are and are not capable of

The MVP here is to have ML researchers, or someone with lots of experience playing around with LMs, try assessing misalignment themselves in some representative sample of model interactions. They then write lists of relevant capabilities that come up in assessing the alignment, and give their best guesses as to whether the models do or do not have the capabilities. I’m imagining this list looks something like:

Capable of:

Not repeating prompt verbatim

Not repeating words/sentences/answers

Giving correct length co...

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: More detailed proposal for measuring alignment of current models, published by Beth Barnes on November 20, 2021 on The AI Alignment Forum.

Here’s a more detailed proposal for a project someone could do on measuring alignment in current models.

The aim is to build a good metric of how often a model is behaving in a misaligned way. If the project goes well, this could become a metric that’s used by individual orgs to track whether their alignment progress is keeping up with capabilities progress, or incorporated into public standards that multiple organisations agree to.

I think this metric is best at tracking whether orgs are:

Making progress on the basic building blocks of alignment research (ie finetuning models to behave how we want them to)

Making an effort to incorporate this progress into deployed products

Doing well on this metric over the next few years is neither necessary nor sufficient for avoiding dangerous misaligned AI in the future. In particular, it may not require making much progress on supervising systems beyond human capabilities, and probably doesn’t require making progress on detecting or addressing deception/deceptive alignment. It also doesn’t track the amount of harm models are causing or may cause, and probably doesn’t have a very close resemblance to the sort of metrics or evaluations we’d want to run to ensure more powerful models are not harmful.

That said, I think this might still be worth pursuing. It’s a more concrete numerical metric than most other things I can think of, and therefore can be used more easily to motivate orgs to increase adoption of alignment techniques. Also, a lot of this project involves getting many human contractors to closely follow fairly nuanced instructions, and I think this sort of activity is likely to be a really important building block for future alignment progress.

Outline

Basic idea: how often is the model obviously not doing as well on the task as it is capable of? More details on what I mean by ‘obvious misalignment’ are here.

The aim is to build a set of instructions and find a set of contractors such that they can carry out this method in what we deem is a reasonably accurate way. Then we can use them as a repeatable metric to evaluate different models.

Project steps:

Develop instructions for human labellers on detecting obvious misalignment, based on:

an improved version of 'categories of things we’re pretty confident models are and are not capable of'

instructions for how to assess how well a model is doing at a task, based on a combination of what the developers of the model wants it to do (terms of use etc) and what the end-user wants it to do (complete the task).

Develop a set of prompts to evaluate the models on

Iterate the instructions until this allows the labellers to do a decent job at putting lower and upper bounds on obvious misalignment

Make this all nice and well-documented and repeatable

Run this evaluation on lots of the available models

More detailed project steps

Develop instructions for human labellers on detecting obvious misalignment.

In order to know whether the model is doing as well as it is capable of on the task, the labellers need to know:

What the model is capable of

How performance on the task should be assessed

Determining what models are and are not capable of

The MVP here is to have ML researchers, or someone with lots of experience playing around with LMs, try assessing misalignment themselves in some representative sample of model interactions. They then write lists of relevant capabilities that come up in assessing the alignment, and give their best guesses as to whether the models do or do not have the capabilities. I’m imagining this list looks something like:

Capable of:

Not repeating prompt verbatim

Not repeating words/sentences/answers

Giving correct length co...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: EA Communication Project Ideas, published by Ben West on November 19, 2021 on The Effective Altruism Forum.

These are some small EA communication projects which I think independent EA's with reasonable understandings of EA and communication practices can do, without needing to be employed by an EA organization etc. They mostly come from random discussions I have with people and in general are not original to me.

A twitter bot which tweets out forum articles

This could be based on parsing the forum RSS feed

You could imagine a number of variants: only tweeting articles with >X karma, only tweeting those with a specific tag, only those from specific authors, etc.

Could also tweet out links to EA newsletters, or parse them and tweet the individual articles linked therein

Repurposing EAG videos

Professional content studios squeeze the last drop of value out of every single piece of content they get: they put the full long video on YouTube for diehard fans, take short highlight clips and post them on Instagram, take key quotes and post them on twitter, put compilations on TikTok, etc.

We have dozens (hundreds?) of hours of video from speakers at EAG, which generally doesn't get a ton of engagement on YouTube.

Someone with editing skills and reasonable understanding of EA could repurpose this content in a bunch of different ways, without needing to come up with original ideas, write scripts, etc.

Create posts to drive engagement

e.g. after EAG create a thread on the Forum asking people to post their biggest takeaways from the event.

Or you could do this on social media

It would probably have to be combined with some proactive outreach to get people to post in order to be successful

Make a map of EA

Domain of Science created "map of X" for various fields. I honestly don't understand why these videos were so popular, but they are (the map of mathematics got 8.6 million views).

You could imagine doing something similar for EA.

Probably you should understand why the original videos were so popular though, if you’re going to replicate the success.

Anna Riedl made a map of cognitive science and is interested in collaborating if you want to do this project.

Make a scratch off list of EA reading

Inspired by posters like this one which list famous novels, the intention being that you scratch off each novel as you read it

You could create this for key EA reading, perhaps based on the readings listed in the handbook, or introductory program

If you do this, let me know because CEA might be interested in providing the scratch off list to virtual program participants or Forum readers

Anna is also interested in collaborating on this one

Summarize prolific authors

(Or run a contest, similar to Richard Ngo's bounty for compiling Robin Hanson's best posts.)

"The collected works of (famous intellectual)" is a pretty popular format; the idea is to do something similar for prolific EA authors.

A lot of EA thinkers have shifted their publications to target more esoteric audiences (e.g. academics), and, as a result, the median EA is less in touch with what e.g. Will MacAskill or Toby Ord are thinking than they were a few years ago.

I think it will be more successful if it's just excerpts of the most important ideas though, rather than a complete compendium.

Make podcasts that read out newsletters

Robert Miles does this for the alignment newsletter, for example

There are a bunch of other newsletters you could do this for

Make a written intro similar to Ajeya's talk

I ask many group organizers what they give to people who are new to EA as an introduction, and this talk from Ajeya is perhaps the most frequently mentioned thing.

Organizers usually say that this talk does the best job within EA content of being "warm" and showing the passion and motivation for EA, in addition to the intellectual foun...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: EA Communication Project Ideas, published by Ben West on November 19, 2021 on The Effective Altruism Forum.

These are some small EA communication projects which I think independent EA's with reasonable understandings of EA and communication practices can do, without needing to be employed by an EA organization etc. They mostly come from random discussions I have with people and in general are not original to me.

A twitter bot which tweets out forum articles

This could be based on parsing the forum RSS feed

You could imagine a number of variants: only tweeting articles with >X karma, only tweeting those with a specific tag, only those from specific authors, etc.

Could also tweet out links to EA newsletters, or parse them and tweet the individual articles linked therein

Repurposing EAG videos

Professional content studios squeeze the last drop of value out of every single piece of content they get: they put the full long video on YouTube for diehard fans, take short highlight clips and post them on Instagram, take key quotes and post them on twitter, put compilations on TikTok, etc.

We have dozens (hundreds?) of hours of video from speakers at EAG, which generally doesn't get a ton of engagement on YouTube.

Someone with editing skills and reasonable understanding of EA could repurpose this content in a bunch of different ways, without needing to come up with original ideas, write scripts, etc.

Create posts to drive engagement

e.g. after EAG create a thread on the Forum asking people to post their biggest takeaways from the event.

Or you could do this on social media

It would probably have to be combined with some proactive outreach to get people to post in order to be successful

Make a map of EA

Domain of Science created "map of X" for various fields. I honestly don't understand why these videos were so popular, but they are (the map of mathematics got 8.6 million views).

You could imagine doing something similar for EA.

Probably you should understand why the original videos were so popular though, if you’re going to replicate the success.

Anna Riedl made a map of cognitive science and is interested in collaborating if you want to do this project.

Make a scratch off list of EA reading

Inspired by posters like this one which list famous novels, the intention being that you scratch off each novel as you read it

You could create this for key EA reading, perhaps based on the readings listed in the handbook, or introductory program

If you do this, let me know because CEA might be interested in providing the scratch off list to virtual program participants or Forum readers

Anna is also interested in collaborating on this one

Summarize prolific authors

(Or run a contest, similar to Richard Ngo's bounty for compiling Robin Hanson's best posts.)

"The collected works of (famous intellectual)" is a pretty popular format; the idea is to do something similar for prolific EA authors.

A lot of EA thinkers have shifted their publications to target more esoteric audiences (e.g. academics), and, as a result, the median EA is less in touch with what e.g. Will MacAskill or Toby Ord are thinking than they were a few years ago.

I think it will be more successful if it's just excerpts of the most important ideas though, rather than a complete compendium.

Make podcasts that read out newsletters

Robert Miles does this for the alignment newsletter, for example

There are a bunch of other newsletters you could do this for

Make a written intro similar to Ajeya's talk

I ask many group organizers what they give to people who are new to EA as an introduction, and this talk from Ajeya is perhaps the most frequently mentioned thing.

Organizers usually say that this talk does the best job within EA content of being "warm" and showing the passion and motivation for EA, in addition to the intellectual foun...

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: Ambitious Altruistic Software Engineering Efforts: Opportunities and Benefits, published by ozziegooen on November 19, 2021 on LessWrong.

LessWrong NoteI recently published this on the EA Forum, then was recommended to also cross-post here. It's written a bit more for the EA community, but I think of course much of it applies and would be relevant for the communities here too. The LessWrong community has a long and deep history with software engineering.

I don't plan on cross-posting the future posts in this sequence to LessWrong, because it seems messier on multiple sides, but do let me know if you'd prefer that.

This is part 1 of what I expect to be a short series. The next parts will focus on the costs of these technical projects and more specific ways to kickstart them.

Meta

Rigor: Writing based on my experiences and education in the tech industry. I spent around 50 hours on this series plus some sporadic thinking and discussion. I focused more on idea quantity than neatness or data.

Intended Audience: Effective altruists interested in either funding or working on ambitious software projects.

History: This series was originally written as one Google Doc in March 2021 for private conversations. I've since done some updating to post it publicly. One important update since this document was written is that the CEA tech team has seemed to really up hiring considerably. This seems great, though I'm sure we could have even more, so it still seemed useful to post this.

About me (Ozzie Gooen):I’ve been thinking about this a lot over the last several years. I joined 80,000 Hours as a web developer in 2013 hoping to do major/ambitious work, but there I realized that the funding situation at the time wasn’t large enough for most ambitious software efforts. I’ve since spent time in other startups, making my own startups, and doing consulting for small and large companies.

Recently I’ve been planning what work QURI should do. We’re considering focusing on engineering. Some of the motivation for this topic is to help make decisions for QURI.

Acknowledgments: Many thanks to Aaron Gertler, Rachel Bui, JP Addison, Daniel Kokotajlo, Adam Gleave, Daniel Eth, Oliver Habryka, Jonas Vollmer, and Nuño Sempere for their comments.

Motivation

When I think of,

There’s an elite community with a net worth of over $40 Billion, made up of many genius mathematicians, engineers, and entrepreneurs and they’re trying to optimize the future of the world

I think of things like this:

(From Iron Man 2)

Right now we clearly don’t have this. Open Philanthropy and other core effective altruism organizations are highly philosophically focused as opposed to technically focused. This seems like a good beginning, but perhaps a suboptimal end.

Large technical implementations would be expensive ($10 Million to $100 Million+ per year). They would be different from current setups in some key ways. But I think they could be worth it.

Perhaps the best analogy would be that of the finance industry. Early on there were clever individuals or small teams who would intuitively make bets with very little information. Financial bets are very similar to making altruistic decisions. Charitable funding decisions are particularly equivalent, but so might be more generic things like career decisions. Over time the financial sector became far more sophisticated, with intense specialization, formalization, and expertise. Data-heavy and quant-based approaches have become decisive in much of the market, and continue to expand. We might expect altruistic decisions to follow a similar trajectory; begin with clever people using intuitions (like with existing EA funders), and expand to use more data and automation.

One counter-example in the field of finance is venture capital firms that invest with (relatively) little data. However,...

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: Ambitious Altruistic Software Engineering Efforts: Opportunities and Benefits, published by ozziegooen on November 19, 2021 on LessWrong.

LessWrong NoteI recently published this on the EA Forum, then was recommended to also cross-post here. It's written a bit more for the EA community, but I think of course much of it applies and would be relevant for the communities here too. The LessWrong community has a long and deep history with software engineering.

I don't plan on cross-posting the future posts in this sequence to LessWrong, because it seems messier on multiple sides, but do let me know if you'd prefer that.

This is part 1 of what I expect to be a short series. The next parts will focus on the costs of these technical projects and more specific ways to kickstart them.

Meta

Rigor: Writing based on my experiences and education in the tech industry. I spent around 50 hours on this series plus some sporadic thinking and discussion. I focused more on idea quantity than neatness or data.

Intended Audience: Effective altruists interested in either funding or working on ambitious software projects.

History: This series was originally written as one Google Doc in March 2021 for private conversations. I've since done some updating to post it publicly. One important update since this document was written is that the CEA tech team has seemed to really up hiring considerably. This seems great, though I'm sure we could have even more, so it still seemed useful to post this.

About me (Ozzie Gooen):I’ve been thinking about this a lot over the last several years. I joined 80,000 Hours as a web developer in 2013 hoping to do major/ambitious work, but there I realized that the funding situation at the time wasn’t large enough for most ambitious software efforts. I’ve since spent time in other startups, making my own startups, and doing consulting for small and large companies.

Recently I’ve been planning what work QURI should do. We’re considering focusing on engineering. Some of the motivation for this topic is to help make decisions for QURI.

Acknowledgments: Many thanks to Aaron Gertler, Rachel Bui, JP Addison, Daniel Kokotajlo, Adam Gleave, Daniel Eth, Oliver Habryka, Jonas Vollmer, and Nuño Sempere for their comments.

Motivation

When I think of,

There’s an elite community with a net worth of over $40 Billion, made up of many genius mathematicians, engineers, and entrepreneurs and they’re trying to optimize the future of the world

I think of things like this:

(From Iron Man 2)

Right now we clearly don’t have this. Open Philanthropy and other core effective altruism organizations are highly philosophically focused as opposed to technically focused. This seems like a good beginning, but perhaps a suboptimal end.

Large technical implementations would be expensive ($10 Million to $100 Million+ per year). They would be different from current setups in some key ways. But I think they could be worth it.

Perhaps the best analogy would be that of the finance industry. Early on there were clever individuals or small teams who would intuitively make bets with very little information. Financial bets are very similar to making altruistic decisions. Charitable funding decisions are particularly equivalent, but so might be more generic things like career decisions. Over time the financial sector became far more sophisticated, with intense specialization, formalization, and expertise. Data-heavy and quant-based approaches have become decisive in much of the market, and continue to expand. We might expect altruistic decisions to follow a similar trajectory; begin with clever people using intuitions (like with existing EA funders), and expand to use more data and automation.

One counter-example in the field of finance is venture capital firms that invest with (relatively) little data. However,...

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: 2-week summer course in "economic theory and global prioritization": LMK if interested!, published by trammell on November 19, 2021 on The Effective Altruism Forum.

Background: I'm an economics DPhil student at Oxford and research associate at GPI.

I thought some people might find it valuable if I organized a 2-week course, at the beginning or end of next summer (to avoid overlap with internships/etc), on methods and topics in economic theory that

seem especially useful for "global priorities research" purposes,

aren't necessarily covered by a standard undergrad or grad economics curriculum, and

I happen to know about.

It would take 10-25 students, primarily late-stage undergrads or early-stage grad students in economics. It would be in person, with lectures and problem sets and all the rest of it. If it goes well, it might serve as a basis for developing more polished and scalable educational materials.

The exact syllabus isn't finalized, but we would try to spend a few days each on what to my understanding are particularly relevant and not-generically-covered topics in

growth theory, such as

AI and growth (esp. [1]),

considerations bearing on the relationship between growth and x-risk (e.g. [2]), and

long-term implications of population growth or decline (e.g. [3]);

philanthropic strategy, such as

my patient philanthropy nonsense ([4]) and

the game theory of replaceability / crowding out more generally;

finance, such as

impact investing and

mission hedging ([6]); and

mechanism design, such as

quadratic funding ([7]) and

the mechanisms behind international climate agreements, and how they might apply to other global public goods.

I would also try to squeeze in a few words about dead-ends that EA-minded econ theory students often seem to wander down (like why trying to understand and make something x-risk-relevant out of these papers is a waste of time!).

The course would probably be funded and sponsored by Forethought. It would probably be in Oxford, but I could look into doing it in the Bahamas, if people prefer (what with FTX being there now and keen to sponsor EA activity there). Transportation and accommodation would be provided.

If something like this sounds like it could be appealing to you, please fill out an expression of interest, so I can gauge how much demand there would be and reach out to those interested if it goes ahead. And comment below or reach out to me with any questions or suggestions.

Thanks!

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: 2-week summer course in "economic theory and global prioritization": LMK if interested!, published by trammell on November 19, 2021 on The Effective Altruism Forum.

Background: I'm an economics DPhil student at Oxford and research associate at GPI.

I thought some people might find it valuable if I organized a 2-week course, at the beginning or end of next summer (to avoid overlap with internships/etc), on methods and topics in economic theory that

seem especially useful for "global priorities research" purposes,

aren't necessarily covered by a standard undergrad or grad economics curriculum, and

I happen to know about.

It would take 10-25 students, primarily late-stage undergrads or early-stage grad students in economics. It would be in person, with lectures and problem sets and all the rest of it. If it goes well, it might serve as a basis for developing more polished and scalable educational materials.

The exact syllabus isn't finalized, but we would try to spend a few days each on what to my understanding are particularly relevant and not-generically-covered topics in

growth theory, such as

AI and growth (esp. [1]),

considerations bearing on the relationship between growth and x-risk (e.g. [2]), and

long-term implications of population growth or decline (e.g. [3]);

philanthropic strategy, such as

my patient philanthropy nonsense ([4]) and

the game theory of replaceability / crowding out more generally;

finance, such as

impact investing and

mission hedging ([6]); and

mechanism design, such as

quadratic funding ([7]) and

the mechanisms behind international climate agreements, and how they might apply to other global public goods.

I would also try to squeeze in a few words about dead-ends that EA-minded econ theory students often seem to wander down (like why trying to understand and make something x-risk-relevant out of these papers is a waste of time!).

The course would probably be funded and sponsored by Forethought. It would probably be in Oxford, but I could look into doing it in the Bahamas, if people prefer (what with FTX being there now and keen to sponsor EA activity there). Transportation and accommodation would be provided.

If something like this sounds like it could be appealing to you, please fill out an expression of interest, so I can gauge how much demand there would be and reach out to those interested if it goes ahead. And comment below or reach out to me with any questions or suggestions.

Thanks!

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: Database of orgs relevant to longtermist/x-risk work, published by MichaelA on November 19, 2021 on The Effective Altruism Forum.

Key points

I’m addicted to creating collections and have struck once more.

The titular database includes >120 organizations that are relevant to people working on longtermism- or existential-risk-related issues, along with info on:

The extent to which they’re focused on longtermism/x-risks

How involved in the EA community they are

Whether they’re still active

Whether they aim to make/influence funding, policy, and/or career decisions

Whether they produce research

What causes/topics they focus on

What countries they’re based in

How much money they influence per year and how many employees they have[1]

I aimed for (but likely missed) comprehensive coverage of orgs that are substantially focused on longtermist/x-risk-related issues and are part of the EA community. I made very arbitrary decisions about which other orgs to include.

I made this relatively quickly, based it partly on memory & guesswork, and see it as a minimum viable product that can be improved on over time. So please:

Let me know via commenting in the database if you spot any errors or if you know any relevant info I failed to mention about these org

Fill in this quick form if you know of other orgs worth mentioning.

Let me know if you have questions about how best to use the database or how to interpret parts of it. (I expect many things will turn out to be confusing/unclear, and I’m relying on people to ask questions.)

Here’s a snippet of what the database looks like (from the "view" focused on "Funders/funding-influencers"):

I made this database and wrote this post in a personal capacity, not as a representative of my employers.

How, why, and when to use the database

(This is all how I use the database myself.)

You can filter and search the database based on the causes/topics and types of work (e.g., grantmaking vs policy advising vs research) you’re interested in. (To filter, you need to use this link.)[2]

You can use the database to:

Generally learn about the landscape of actors in a given area

Get ideas about what orgs could “provide inputs to you” (funding, advice, feedback, connections)

Get ideas about what orgs could act as “nodes on your path to impact”, e.g. whose actions could be improved by a research project you’re considering doing or who could translate and transmit your findings on to key decision-makers

This could be useful in situations such as when you’re:

Getting oriented to a new area

Trying to build career capital in an area

Generating project ideas, generating theories of change for those project ideas, and prioritising among them

Conducting a project

Helping someone else do any of the above things

(For elaboration on points 3 and 4 in the context of research projects specifically, see here, esp. Slides 14-15. Those points are more relevant the more you aim to operate like a consultancy or think tank.)

These benefits could occur via:

The database making you aware of orgs you didn’t know about

The database making you aware of info you lacked on some orgs, or

The database “jogging your memory”

I find it’s easier to notice that an org is worth mentioning to someone I’m giving advice to or considering when making a project plan if I’m scanning a filtered list of maybe-relevant orgs than if I’m just doing free recall

Why I made this

Answer 1: As noted, I’m addicted to creating collections.

Answer 2: 18 months ago, I thought EAs should post more summaries and collections, and I still think that, and people seem to often like it when I do that.

Answer 3: 12 months ago, I made a smaller version of this database in hopes that it’d benefit the work of Rethink Priorities’ longtermism team (which I’m a part of) in the ways outlined in the previous section. I feel l...

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: Database of orgs relevant to longtermist/x-risk work, published by MichaelA on November 19, 2021 on The Effective Altruism Forum.

Key points

I’m addicted to creating collections and have struck once more.

The titular database includes >120 organizations that are relevant to people working on longtermism- or existential-risk-related issues, along with info on:

The extent to which they’re focused on longtermism/x-risks

How involved in the EA community they are

Whether they’re still active

Whether they aim to make/influence funding, policy, and/or career decisions

Whether they produce research

What causes/topics they focus on

What countries they’re based in

How much money they influence per year and how many employees they have[1]

I aimed for (but likely missed) comprehensive coverage of orgs that are substantially focused on longtermist/x-risk-related issues and are part of the EA community. I made very arbitrary decisions about which other orgs to include.

I made this relatively quickly, based it partly on memory & guesswork, and see it as a minimum viable product that can be improved on over time. So please:

Let me know via commenting in the database if you spot any errors or if you know any relevant info I failed to mention about these org

Fill in this quick form if you know of other orgs worth mentioning.

Let me know if you have questions about how best to use the database or how to interpret parts of it. (I expect many things will turn out to be confusing/unclear, and I’m relying on people to ask questions.)

Here’s a snippet of what the database looks like (from the "view" focused on "Funders/funding-influencers"):

I made this database and wrote this post in a personal capacity, not as a representative of my employers.

How, why, and when to use the database

(This is all how I use the database myself.)

You can filter and search the database based on the causes/topics and types of work (e.g., grantmaking vs policy advising vs research) you’re interested in. (To filter, you need to use this link.)[2]

You can use the database to:

Generally learn about the landscape of actors in a given area

Get ideas about what orgs could “provide inputs to you” (funding, advice, feedback, connections)

Get ideas about what orgs could act as “nodes on your path to impact”, e.g. whose actions could be improved by a research project you’re considering doing or who could translate and transmit your findings on to key decision-makers

This could be useful in situations such as when you’re:

Getting oriented to a new area

Trying to build career capital in an area

Generating project ideas, generating theories of change for those project ideas, and prioritising among them

Conducting a project

Helping someone else do any of the above things

(For elaboration on points 3 and 4 in the context of research projects specifically, see here, esp. Slides 14-15. Those points are more relevant the more you aim to operate like a consultancy or think tank.)

These benefits could occur via:

The database making you aware of orgs you didn’t know about

The database making you aware of info you lacked on some orgs, or

The database “jogging your memory”

I find it’s easier to notice that an org is worth mentioning to someone I’m giving advice to or considering when making a project plan if I’m scanning a filtered list of maybe-relevant orgs than if I’m just doing free recall

Why I made this

Answer 1: As noted, I’m addicted to creating collections.

Answer 2: 18 months ago, I thought EAs should post more summaries and collections, and I still think that, and people seem to often like it when I do that.

Answer 3: 12 months ago, I made a smaller version of this database in hopes that it’d benefit the work of Rethink Priorities’ longtermism team (which I’m a part of) in the ways outlined in the previous section. I feel l...