The Nonlinear Library

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 I'm still going out to bat for the EA movement post FTX, published by Gemma Paterson on September 30, 2023 on The Effective Altruism Forum.

From the looks of it, next week might be rough for people who care about Effective Altruism. As CEA acting CEO Ben West pointed out on the forum:

"Sam Bankman-Fried's trial is scheduled to start October 3, 2023, and Michael Lewis's book about FTX comes out the same day. My hope and expectation is that neither will be focused on EA .

Nonetheless, I think there's a decent chance that viewing the Forum, Twitter, or news media could become stressful for some people, and you may want to pre-emptively create a plan for engaging with that in a healthy way.

I really appreciated that comment since I didn't know that and I'm glad I had time to mentally prepare. As someone who does outward facing voluntary community building at my workplace and in London, I feel nervous. I've written this piece to manage that anxiety. I actually wrote a lot of it last year to process my feelings but now seems like a good time to share.

Thoughts on EA after 2022

During the spring/summer of 2022, I was studying for my final ICAS Chartered Accountancy case study exams. They usually involve:

Being given a scenario where you are the qualified accountant and have been asked to do some analysis for a group of companies.

As you need to meet a minimum standard of Public Trust and Ethics to join an accountancy body, there is usually a dodgy character or two involved at the client company or at the firm that employs you.

The dodgy characters are lax about controls and risk - asking you as the accountant to "look the other way" and the case is usually set up that there is an incentive for you as the accountant to do that.

There are often weak internal controls at the company which make fraud more likely. Some examples of weak internal controls:

Lack of segregation of duties - one employee has end-to-end control of a process without oversight

Inadequate access controls - sensitive systems and data are not restricted to authorized personnel

Circumvention of controls - existing controls are intentionally avoided or bypassed by employees

Unsigned approvals - documents requiring sign-off are not properly authorized

At the end, you submit your technical accounting, finance, tax, internal controls, and general business advice that was requested as part of the case study. You also need to submit an ethics memo which would outline your concerns about the dodgy people or controls.

You'll be glad to hear that after going through lots of practice papers, I passed!

Then three months later, the news on FTX dropped.

I'm not going to outline it here since I imagine the audience for this are already familiar - if someone has a good summary, please drop it in the comments. It wasn't just FTX, around that time, information was released on Wytham Abbey, Bostrom and the Time sexual harassment scandals.

For context, I'm mostly interested in expanding the reach of effective giving / EA impact methodologies and personally put a high discount rate on the future vs existing suffering, so, I didn't follow the FTX Future Fund that closely.

My primary feelings at the end of 2022 were anger and embarrassment.

Anger

I'm annoyed I must answer questions about the damned castle. I did not buy the castle. None of the charities I recommend you donate to bought the castle. Yet in my community building work, I have to answer questions about the castle. It was handled poorly and I'm angry about it.

When the Time article came out, I was running the London Women (and NBs) in EA group. Honestly, I incorrectly assumed it was just about EA in the Bay area so when it was revealed it was Owen Cotton-Barrett, that was a shock and tbh I was annoyed at a lot of the responses.

In hindsight, I wish I'd created a sp...

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 Winners of the 2023 Open Philanthropy AI Worldviews Contest, published by Jason Schukraft on September 30, 2023 on The Effective Altruism Forum.

Introduction

In March 2023, we launched the Open Philanthropy AI Worldviews Contest. The goal of the contest was to surface novel considerations that could affect our views on the timeline to transformative AI and the level of catastrophic risk that transformative AI systems could pose. We received 135 submissions. Today we are excited to share the winners of the contest.

But first: We continue to be interested in challenges to the worldview that informs our AI-related grantmaking. To that end, we are awarding a separate $75,000 prize to the Forecasting Research Institute (FRI) for their recently published writeup of the 2022 Existential Risk Persuasion Tournament (XPT). This award falls outside the confines of the AI Worldviews Contest, but the recognition is motivated by the same principles that motivated the contest. We believe that the results from the XPT constitute the best recent challenge to our AI worldview.

FRI Prize ($75k)

Existential Risk Persuasion Tournament by the Forecasting Research Institute

AI Worldviews Contest Winners

First Prizes ($50k)

AGI and the EMH: markets are not expecting aligned or unaligned AI in the next 30 years by Basil Halperin, Zachary Mazlish, and Trevor Chow

Evolution provides no evidence for the sharp left turn by Quintin Pope (see the LessWrong version to view comments)

Second Prizes ($37.5k)

Deceptive Alignment is <1% Likely by Default by David Wheaton (see the LessWrong version to view comments)

AGI Catastrophe and Takeover: Some Reference Class-Based Priors by Zach Freitas-Groff

Third Prizes ($25k)

Imitation Learning is Probably Existentially Safe by Michael Cohen

'Dissolving' AI Risk - Parameter Uncertainty in AI Future Forecasting by Alex Bates

Caveats on the Winning Entries

The judges do not endorse every argument and conclusion in the winning entries. Most of the winning entries argue for multiple claims, and in many instances the judges found some of the arguments much more compelling than others. In some cases, the judges liked that an entry crisply argued for a conclusion the judges did not agree with - the clear articulation of an argument makes it easier for others to engage. One does not need to find a piece wholly persuasive to believe that it usefully contributes to the collective debate about AI timelines or the threat that advanced AI systems might pose.

Submissions were many and varied. We can easily imagine a different panel of judges reasonably selecting a different set of winners. There are many different types of research that are valuable, and the winning entries should not be interpreted to represent Open Philanthropy's settled institutional tastes on what research directions are most promising (i.e., we don't want other researchers to overanchor on these pieces as the best topics to explore further).

We did not provide any funding specifically for the XPT, which ran from June 2022 through October 2022. In December 2022, we recommended two grants totaling $6.3M over three years to support FRI's future research.

The link above goes to the version Michael submitted; he's also written an updated version with coauthor Marcus Hutter.

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

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: What Has the CEA Uni Groups Team Been Up To? - Our Summer 2023 Review, published by Joris P on September 30, 2023 on The Effective Altruism Forum.

TLDR

CEA's University Groups Team worked on a wide range of projects this summer

We wrapped up a round of UGAP and OSP (our programs supporting university group organizers with digital resources, mentorship, training, and for UGAP a stipend) and launched a new one. Both rounds had over 100 people participating across both programs

We ran a summit for relatively experienced group organizers, had four interns join us (who worked on projects like updating the Groups Resource Centre and setting up residencies), and continued to give out (relatively small) group support grants, to cover operational costs for university groups

We're excited about potential future projects, like supporting more 'top' universities and/or seeding more support for AIS groups, but continue to reflect on what projects would be most valuable

Intro

CEA's University Groups Team is working on supporting university groups focused on EA ideas, and their organizers. We wrote more about our mission and vision last year here - note it is slightly outdated. Our team currently consists of Jessica McCurdy (team lead), Jake McKinnon, and Joris Pijpers.

In this post we share what we've been working on this summer, including launching new rounds of UGAP & OSP, a summit, an internship program, and more. We won't touch much upon what we did in spring '23 (but have a retrospective of UGAP fall '22 here), and have also chosen to not expand in detail on why we are focusing on these programs and not others, in favor of getting the post out sooner. Additionally, we are still in the process of deeper impact analysis. We hope to publish more details on these in the future.

Overall, we are pretty happy with what we were able to achieve this summer and are excited about future directions.

Why university groups?

We think university groups have the potential to be especially promising places to introduce people to EA ideas, and then help them learn more about and act on them:

University group members (and organizers in particular) have a strong track record of making significant contributions to EA priorities, both in and beyond meta work

Students are usually in the process of thinking through their priorities for their careers and lives. In this process, they are often open to exploring new ideas and communities, and have time to do so

Universities are places where people build communities and social networks. These in-person social ties are important for people being more likely to take significant action on EA ideas

Universities can have high concentrations of pre-screened talent

Universities are a very large source of new EAs each year, which means that student groups might have an outsized influence on the development of EA culture and movement priorities. Setting norms of good epistemics through advice and content aspirationally may improve that of EA

A note on which universities we focus on

We think top universities, such as those previously listed as focus universities, have a high concentration of top talent. That means that we extend higher touch support to top universities and have some programming focused on them specifically (like our summit and residencies). However, the majority of talented, altruistic and greatly influential people do not come from top universities. Therefore we think scalable support (like UGAP & OSP) which includes both top and non-top universities is also worthwhile. We have many examples of people who are doing work that we think is really important, who did not come from top universities, and might have not gotten involved with EA ideas if it wasn't for their university group. We also think people that we help from these universities tend 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: Open Philanthropy is hiring for multiple roles across our Global Catastrophic Risks teams, published by Open Philanthropy on September 30, 2023 on The Effective Altruism Forum.

It's been another busy year at Open Philanthropy; after nearly doubling the size of our team in 2022, we've added over 30 new team members so far in 2023. Now we're launching a number of open applications for roles in all of our Global Catastrophic Risks (GCR) cause area teams (AI Governance and Policy, Technical AI Safety, Biosecurity & Pandemic Preparedness, GCR Cause Prioritization, and GCR Capacity Building).

The application, job descriptions, and general team information are available here. Notably, you can apply to as many of these positions as you'd like with a single application form!

We're hiring because our GCR teams feel pinched and really need more capacity. Program Officers in GCR areas think that growing their teams will lead them to make significantly more grants at or above our current bar. We've had to turn down potentially promising opportunities because we didn't have enough time to investigate them; on the flip side, we're likely currently allocating tens of millions of dollars suboptimally in ways that more hours could reveal and correct.

On the research side, we've had to triage important projects that underpin our grantmaking and inform others' work, such as work on the value of Open Phil's last dollar and deep dives into various technical alignment agendas. And on the operational side, maintaining flexibility in grantmaking at our scale requires significant creative logistical work. Both last year's reduction in capital available for GCR projects (in the near term) and the uptick in opportunities following the global boom of interest in AI risk make our grantmaking look relatively more important; compared to last year, we're now looking at more opportunities in a space with less total funding.

GCR roles we're now hiring for include:

Program associates to make grants in technical AI governance mechanisms, US AI policy advocacy, general AI governance, technical AI safety, biosecurity & pandemic preparedness, EA community building, AI safety field building, and EA university groups.

Researchers to identify and evaluate new areas for GCR grantmaking, conduct research on catastrophic risks beyond our current grantmaking areas, and oversee a range of research efforts in biosecurity. We're also interested in researchers to analyze issues in technical AI safety and (separately) the natural sciences.

Operations roles embedded within our GCR grantmaking teams: the Biosecurity & Pandemic Preparedness team is looking for an infosec specialist, an ops generalist, and an executive assistant (who may also support some other teams); the GCR Capacity Building team is looking for an ops generalist.

Most of these hires have multiple possible seniority levels; whether you're just starting in your field or have advanced expertise, we encourage you to apply.

If you know someone who would be great for one of these roles, please refer them to us. We welcome external referrals and have found them extremely helpful in the past. We also offer a $5,000 referral bonus; more information here.

How we're approaching these hires

You only need to apply once to opt into consideration for as many of these roles as you're interested in. A checkbox on the application form will ask which roles you'd like to be considered for. We've also made efforts to streamline work tests and use the same tests for multiple roles where possible; however, some roles do use different work tests, so it's possible you'll still have to take different work tests for different roles, especially if you're interested in roles across a wide array of skillsets (e.g., both research and operations). You may also have interviews with mu...

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 Lighthaven Campus is open for bookings, published by habryka on September 29, 2023 on LessWrong.

Lightcone Infrastructure (the organization that grew from and houses the LessWrong team) has just finished renovating a 7-building physical campus that we hope to use to make the future of humanity go better than it would otherwise.

We're hereby announcing that it is generally available for bookings. We offer preferential pricing for projects we think are good for the world, but to cover operating costs, we're willing to rent out space to a wide variety of people/projects.

How do I express interest?

Chat with us (you can use LessWrong intercom in the bottom right)

Submit an inquiry via this form

Comment on this post

What kinds of things can I run here?

Team Retreats (15-80 people)

We offer cozy nooks with firepits, discussion rooms with endless whiteboards, plus lodging. The space is very modular, and we can put up walls and dividers that will separate out your own private section of the campus.

Parties & Events (10 - 500 people)

From private dinners to 500 person parties. Sound setup, music, use of private kitchens, snacks/catering.

Conferences (50-600 people)

20+ session spaces and lodging for up to 80 attendees.

Lodging (10 - 80 people usually)

We have 45 bedrooms sleeping up to 80 people that can be booked together with or independently from events.

Other

You could host all kind of things here, like choir rehearsals, dinners, LARPing, etc.

What does it cost?

We determine pricing on a case-by-case basis, but a good approximation is $100 - $250 per person per day for retreats conferences and $25 - $75 per person for parties.

We offer some groups large discounts (including free), if we think what you are doing makes the world better. Do really reach out to us if price is an issue. We will often be able to work something out.

You can use your own caterer, or we can provide catering. Our default caterer offers meals from $20/meal to $50/meal that most visitors have found decently satisfying.

We engage in some price discrimination. If we think your best alternative to our venue would be much more expensive, we may charge more than our listed price. We try to find a fair price that splits the difference between our own costs and the value you get out of the space.

We also provide more bespoke services, for an additional charge (see below). Overall we try to be reasonable about pricing, and don't expect to make much profit on the space.

What services do you provide?

For an additional charge we can provide (among many other things):

Catering services, snacks & drinks

General ops-support for your event

Event branding and marketing

Our team has a lot of experience running all kinds of different events, from small 15-20 person dinners to 1000+ person conferences, to 2-month fellowships and research programs.

We can't guarantee we can provide everything you need, but if you are facing some kind of problem or obstacle in the course of preparing or running your event, we can probably help you.

What are the various buildings and areas?

The AtriumUpstairs has a common space and 7 bedrooms. Downstairs is great for parties and weekend conferences (though not available during weekday business hours).

The Bayes House20 bedrooms, it can host sessions of up to 60 people. Includes the surrounding gardens (200+ capacity).

The CottageStorage and laundry. You can store things here during events.

The DenFront of building has 7 offices/event-spaces. Back of building has 4 of our nicest and most secluded bedrooms.

The Extension4 small session spaces and/or bedrooms, a gym, and a 60 person session space.

The FarmhouseCommon space surrounded by beautiful gardens and a nice outdoor bar.

The GuesthouseAn additional house a block away, with 10 additional bedrooms.

What is your goal with Lighthaven?...

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 FAR Labs, an AI safety coworking space, published by bgold on September 29, 2023 on LessWrong.

FAR Labs is a coworking hub in downtown Berkeley for organizations and individuals working on AI safety and related issues. Since opening the space in March 2023, we have grown to host approximately 30 members. Our members are primarily drawn from four anchor organizations, but we also host a number of independent researchers and research teams.

Now that our initial setup is complete, we are pleased to announce an open call for applications from individuals or organizations.

Our initial aims for FAR Labs:

First and foremost it should be a place to do great work. Our members are working on challenging problems, and we want to improve their effectiveness, reduce distractions, and provide a professional environment for them to work in. That includes providing a variety of workspaces (private offices, dedicated desks and hot-desks, general areas), catering, and other office amenities such as a gym.

A warm, intellectually generative culture. Having interesting and fun conversations is one of the best parts of working in a shared environment, and championing a culture that enables those interactions is incredibly important to us.

Supporting collaborations between members, other alignment organizations, and outside collaborators (e.g. academics, or industry researchers). While membership is tied to actively working on AI safety (technical or governance) or related areas (e.g. field building, advocacy, fundraising), we also want to make a space that's welcoming to many viewpoints, which we expect to benefit both members and visitors.

FAR AI's broader mission is to support research and initiatives that promote trustworthy and safe AI systems. FAR Labs is an investment in operations and coordination. By creating research environments and good operational scaffolding, we can accelerate safety research and x-risk reduction across projects and orgs.

For the past six months that's looked like setting up the space and getting the basics in place (office, food, equipment). Moving into 2024 the Labs team will begin offering programs for members - as well as others in the AI safety ecosystem - for developing relevant skills for research and operational excellence. We're particularly excited about identifying best practices and providing training to help members in building and scaling high performing teams.

FAR Labs runs at cost/a slight loss; we're aiming for a fully member supported office and community space.

We are opening for new membership applications. Currently we hope to onboard one to three alignment oriented organizations, and perhaps a handful of independent members, aiming for a total membership of 40-50 people. If you're interested in working from FAR Labs, or would like to learn more, please reach out.

Programs, external visitors, and workshops will be grant funded, while our ongoing day to day office costs are covered by member dues.

While we host several independent researchers we do prioritize organizations.

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: Bids To Defer On Value Judgements, published by johnswentworth on September 29, 2023 on LessWrong.

Consider two claims:

"broccoli is good for you"

"broccoli decreases cholesterol"

Even though the former might be considered a lossy summary of the latter, the two feel very different; they pull very different levers in my brain. "Broccoli decreases cholesterol" pulls levers like:

Is the claim even true? Does broccoli really decrease cholesterol? Would I expect to hear people claim this even in worlds where it is false?

How much does broccoli decrease cholesterol? Is it a tiny effect size? Also how much broccoli?

Where did this information come from? Was it perhaps among the endless stream of bullshit nutrition studies?

Relative to what baselines? Is broccoli substituted for something, or added? What's the population?

Do I want lower cholesterol? Do I want it more than I want to eat food tastier than broccoli?

(Probably other people will not have these exact same levers, but I expect most people instinctively respond to "eating broccoli decreases cholesterol" with some kind of guess about where that information came from and how trustworthy it is.)

The other version, "Eating broccoli is good for you", not only doesn't pull those levers, it feels like. the sentence is making a bid to actively suppress those levers? Like, those levers are all part of my value-judgement machinery, and the sentence "broccoli is good for you" is making a bid to circumvent that machinery entirely and just write a result into my value-cache.

This is a "bid to defer on a value judgement": the sentence is a bid to directly write a value-judgement into cache, without going through my own internal value-judgement machinery. If I accept that bid, then I'm effectively deferring to the speaker's value-judgement.

The Memetic Parasite Model

If broccoli is good for you (and presumably for most other humans, in general), then sharing that information is a friendly, helpful, prosocial action.

More generally: if a value judgement is correct, then passing it along is typically a friendly, helpful, prosocial action. After all, it will help other people to make more "good" decisions if they have more correct information cached about what's "good"/"bad".

But this gives rise to a potential parasitic meme dynamic:

Alice, at some point, hears that broccoli is good for you. She caches that value judgement.

When talking to Bob, Alice notices that it would be helpful and prosocial for her to tell Bob that broccoli is good for you. After all, according to her cached value judgement, broccoli is in fact good for you, so it would be prosocial to pass that information along.

Now Bob hears from Alice that broccoli is good for you and, unless he actively disbelieves what he's hearing, caches that value judgement.

. and that memetic loop can run just fine regardless of what benefits broccoli does or does not have.

Note one difference from more general information cascades: information has to be salient for some reason to be passed along. Value judgements tend to be inherently salient; they lend themselves directly to use, since they directly say what would be good or bad.

Another difference from more general information cascades: value judgements naturally lend themselves to black-boxing. They don't need to interact much with gears, because they circumvent the gearsy machinery of value judgement.

Now, at first glance this model seems rather maxentropic; one could claim that anything at all is "good" or "bad", and the same dynamic will propagate it, so at first glance there aren't predictions about which value judgements we will/won't see propagating memetically. But now we can note that there are factors which favor memeticity of some such claims over others.

Value judgements producing outcomes which are actually "good" by ...

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: Should 80,000 hours have more near-termist career content?, published by NickLaing on September 29, 2023 on The Effective Altruism Forum.

TDLR: Assuming that Long Termist causes are by far the most important, 80,000 hours might still be better off devoting more of their space to near-termist causes, to grow the EA community, increase the number working in long term causes and improve EA optics.

Background

80,000 hours is a front page to EA, presenting the movement/idea/question to the world. For many, their website is people's introduction to effective altruism (link). and might be the most important internet-based introduction to Effective altrurism, - even if this is not their main purpose.

As I see it, 80,000 hours functions practically as a long-termist career platform, or as their staff put it, to "help people have careers with a lot of social impact". I'm happy that they have updated their site to make this long-termist focus a little more clear, but I think there is an argument that perhaps 20%-30% of their content should be near-term focused, and that this content should be more visible. These arguments I present have probably been discussed ad-nauseam within 80,000 hours leadership and EA leadership in general, and nothing I present is novel. Despite that, I haven't seen these arguments clearly laid out in one place (I might have missed it), so here goes.

Assumptions

For the purposes of this discussion, I'm going to make these 3 assumptions. They aren't necessarily true and I don't necessarily agree with them, but they make my job easier and argument cleaner.

Long termist causes are by far the most important, and where the vast majority of EAs should focus their work.

Most of the general public are more attracted to near-termist EA than long-termist EA

80,000 hours is an important "frontpage" and "gateway" to EA

Why 80,000 hours might want to focus more on near-term causes

1. Near termist focus as a pathway to more future long termist workersMore focus on near termist careers might paradoxically lead to producing more long termist workers in the long run. Many people are drawn to the clear and more palatable idea that we should devote our lives to doing the most good to humans and animals alive right now (assumption 2). However after further thought and engagement they may change their focus to longtermism. I'm not sure we have stats on this, but I've encountered many forum members who seem to have followed this life pattern. By failing to appeal to the many who are attracted to near-termist EA now, we could miss out on significant numbers of long termist workers in a few years time.

2. Near termist causes are a more attractive EA "front door" 80,000 hours is a "front door" to Effective Altruism, then it is to make sure as many people enter the door as possible. Although 80,000 hours make it clear this isn't one of their main intentions, there is huge benefit to maximizing community growth.

3. Some people will only ever want/be able to work on near-term causes.There are not-unreasonable people who might struggle with the tenets/tractability of long-termism, or be too far along the road-of-life to move from medicine to machine learning, or just don't want to engage with long termism for whatever reason. They may miss out a counterfactual fruitful living-your-best-EA-life, because when they clicked on the 80,000 hours website they didn't manage to scroll down 5 pages to cause area no. 19 and 20 - the first 2 near-term cause areas "Factory farming" and "Easily preventable or treatable illness". These are listed below well established and clearly tractable cause areas such as "risks from atomically precise manufacturing" and "space governance"

There may be many people who never want to/be able decide to change careers to the highest impact long-term causes, but ...

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: "Diamondoid bacteria" nanobots: deadly threat or dead-end? A nanotech investigation, published by titotal on September 29, 2023 on The Effective Altruism Forum.

Confidence level: I'm a computational physicist working on nanoscale simulations, so I have some understanding of most of the things discussed here, but I am not specifically an expert on the topics covered, so I can't promise perfect accuracy.

I want to give a huge thanks to Professor Phillip Moriarty of the university of Nottingham for answering my questions about the experimental side of mechanosynthesis research.

Introduction:

A lot of people are highly concerned that a malevolent AI or insane human will, in the near future, set out to destroy humanity. If such an entity wanted to be absolutely sure they would succeed, what method would they use? Nuclear war? Pandemics?

According to some in the x-risk community, the answer is this: The AI will invent molecular nanotechnology, and then kill us all with diamondoid bacteria nanobots.

This is the "lower bound" scenario posited by Yudkowsky in his post AGI ruin:

The nanomachinery builds diamondoid bacteria, that replicate with solar power and atmospheric CHON, maybe aggregate into some miniature rockets or jets so they can ride the jetstream to spread across the Earth's atmosphere, get into human bloodstreams and hide, strike on a timer.

The phrase "diamondoid bacteria" really struck out at me, and I'm not the only one. In this post by Carlsmith (which I found very interesting), Carlsmith refers to diamondoid bacteria as an example of future tech that feels unreal, but may still happen:

Whirling knives? Diamondoid bacteria? Relentless references to paper-clips, or "tiny molecular squiggles"? I've written, elsewhere, about the "unreality" of futurism. AI risk had a lot of that for me.

Meanwhile, the controversial anti-EA crusader Emille Torres cites the term "diamondoid bacteria" as a reason to dismiss AI risk, calling it "patently ridiculous".

I was interested to know more. What is diamondoid bacteria? How far along is molecular nanotech research? What are the challenges that we (or an AI) will need to overcome to create this technology?

If you want, you can stop here and try and guess the answers to these questions.

It is my hope that by trying to answer these questions, I can give you a taste of what nanoscale research actually looks like. It ended up being the tale of a group of scientists who had a dream of revolutionary nanotechnology, and tried to answer the difficult question: How do I actually build that?

What is "diamondoid bacteria"?

The literal phrase "diamondoid bacteria" appears to have been invented by Eliezer Yudkowsky about two years ago. If you search the exact phrase in google scholar there are no matches:

If you search the phrase in regular google, you will get a very small number of matches, all of which are from Yudkowsky or directly/indirectly quoting Yudkowsky. The very first use of the phrase on the internet appears to be this twitter post from September 15 2021. (I suppose there's a chance someone else used the phrase in person).

I speculate here that Eliezer invented the term as a poetic licence way of making nanobots seem more viscerally real. It does not seem likely that the hypothetical nanobots would fit the scientific definition of bacteria, unless you really stretched the definition of terms like "single-celled" and "binary fission". Although bacteria are very impressive micro-machines, so I wouldn't be surprised if future nanotech bore at least some resemblance.

Frankly, I think inventing new terms is an extremely unwise move (I think that Eliezer has stopped using the term since I started writing this, but others still are). "diamondoid bacteria" sounds science-ey enough that a lot of people would assume it was already a scien...

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: List of how people have become more hard-working, published by Chi on September 29, 2023 on The Effective Altruism Forum.

I've recently asked how people have become more hard-working. I compiled the answers across the EA Forum and LessWrong (and some private messages) in a list for myself to make it easier for me to experiment with the suggestions. I thought I'd share the list here in case it's useful for anyone else. I also list the things that people said didn't work and a couple of other things.

This wasn't done to be "proper", so the list is sloppy in many ways: I liberally paraphrased what people said; often I could have easily counted something two people said as the same or two different things, which would change the way I counted how often something was mentioned; I very roughly grouped the things that were said into categories but easily could have categorised many things differently.

Notable points

Individual points that were mentioned the most:

(Soft) accountability (deadlines, beeminder, accountability buddy, posting about your goals, boss as a service, promising friends) (9)

Working on interesting problems/enjoyable work (and in an enjoyable work environment) (8)

Focusmate/Coworking (often poms) (7)

Some things that weren't mentioned a lot but that I found interesting:

Identifying (or being thought of) as hard-working (3)

Categorising work as "not work" and instead as something enjoyable, adjusting work environment accordingly (1)

Other things

Age at the time of the shift in hard-workingnesswas usually not mentioned, but when it was mentioned, it was between 20-30

Some people managed to become permanently more hard-working after experiencing one period of working hard, even when they switched to less enjoyable or just very different work. That initial period would either be induced by external pressure or by working hard on something they didn't consider work. (3)

Full list of what made people more hard-working

Here is the full list, ordered by how often things in a category were named. (Note that often the same person would list multiple things in the category, so the sums aren't summing over people)

Thing that workedHow many people mentionedSocial Focusmate/Coworking (often poms)8(Regular) contact with other people to talk about work, debug, check-in etc.4Identifying (or being thought of) as hard-working3Surrounding yourself with ~hard-working people in life in general3Supportive work environment2Having a manager1 21What kind of work Working on interesting problems/enjoyable work (and in an enjoyable work environment)8Feeling like you're good at what you're doing, getting positive feedback4Working on things you consider important3More clear tasks, feedback, endpoints etc.2Less pressuring work1Autonomy1Making work more fun1At least one (work) thing you like per day1 21External pressure (Soft) accountability (deadlines, beeminder, accountability buddy, posting about your goals, boss as a service, promising friends)9Children/poverty: External motivation to do work2Almost being fired1 12Learning more about yourself and your goals Figure out which work hours are most useful, schedule

different kinds of work for different times to work more efficiently3Thinking about what you want to do with life and what (work) motivates you3Experimenting with what actually makes you (less) productive e.g. via tracking and realising that productivity advice is very personal2Repeated experience of joy from achieving big things1Deciding how many hours you endorse working1 10Misc. specific techniques Productivity books2Productivity systems2Having policies for ways of making time productive when there are trade-offs e.g. with money1Physical Kanban boards1Walking meetings with yourself1Leverage momentum: Start the day with a small experience of success and let that spiral1Work ...