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: Focus on the Hardest Part First, published by Johannes C. Mayer on September 11, 2023 on The AI Alignment Forum.

Here is some obvious advice.

I think a common failure mode when working on AI alignment[1] is to not focus on the hard parts of the problem first. This is a problem when generating a research agenda, as well as when working on any specific research agenda. Given a research agenda, there are normally many problems that you know how to make progress on. But blindly working on what seems tractable is not a good idea.

Let's say we are working on a research agenda about solving problems A, B, and C. We know that if we find solutions to A, B, and C we will solve alignment. However, if we can't solve even one subproblem, the agenda would be doomed. If C seems like a very hard problem, that you are not sure you can solve, it would be a bad idea to flinch away from C and work on problem A instead, when A seems so much more manageable.

If solving A takes a lot of time and effort, all of that time and effort would be wasted, if you can't solve C in the end. It's especially worrisome when A has tight fightback loops, such that you constantly feel like you are making progress. Or when it is just generally fun to work on A.

Of course, it can make sense to work on A first if you expect this to help you solve C, or at least give you more information on its tractability. The general version of this is illustrated by considering that you have a large list of problems that you need to solve. In this case, focusing on problems that will provide you with information that will be helpful for solving many of the other problems can be very useful. But even then you should not lose sight of the hard problems that might block you down the road.

The takeaway is that these two things are very different:

Solving A as an instrumental subgoal in order to make progress on C, when C is a potential blocker.

Avoiding C, because it seems hard, and instead working on A because it seems tractable.

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: Writing about my job: Growth lead at a startup in Kenya, published by Luke Eure on September 10, 2023 on The Effective Altruism Forum.

I think a lot of impact-minded people should seriously consider working at early-stage startups - particularly in low-middle income countries.

I'm an American currently working at an early-stage startup in Kenya, so thought I would write up:

How I got this job

How I think about the impact of my work

What my day to day job is like

My advice for people thinking of working at a startup

What would make me decide to stay or leave my company

I'm aiming to be concise, so sound off in the comments if there's anything you'd like me to expand on in more detail.

Context: I work as a growth lead at a company called Kapu Africa, a startup based in Kenya reducing the cost of living for Kenyans by offering cheap direct-to-consumer sales of every-day purchased items, with next day delivery

How I got this job

After undergrad I worked in consulting for 2.5 years (BCG) - first in Chicago and then in Nairobi Kenya

I decided I had learned enough from BCG, and that the best way for me to contribute to global development was to work at a startup. I've written here and here about that decision-making process

My coworkers at BCG Nairobi put me in touch with a bunch of people working in various startups in Nairobi as well as elsewhere in Africa. Over the course of 15 or so conversations, I got a sense for what kinds of companies and roles were interesting to me vs. which weren't

I was very lucky in that most of the companies I was talking to were looking for someone with my background. I never had any formal interviews. I was able to have an attitude of "I'm looking for a job where I can have a lot of impact, and if I find that job I will definitely get that job."

Kapu was actually in stealth mode when I joined it, so the only way I found out about it was because a former colleague worked there

I had like 3 conversations with different people from Kapu, decided it was the best shot at impact I had (see next section), and so went with it!

How I think about the impact of my work

I had decided on global poverty as what I wanted to work on in the near-term. Besides the obvious direct benefits, solving global poverty means also unleashing the potential of talented people (who happen to be very poor right now) who can help us solve humanity's other problems

Economic development seems like the best way to lift people out of extreme poverty

I think Kapu is particularly good in terms of impact on a few metrics:

Direct impact: Our core business is making goods cheaper for end consumers - targeting poor Kenyans. Our goal is to provide $1B in savings to customers. I view providing real savings on everyday purchases as quite similar to unconditional cash transfers in terms of impact

My personal potential to give a lot: If our company does super well and gets a high valuation, I'll make a lot of money I can then give away

Driving economic development via:

Low-medium skill job creation

High-skill job creation potentially setting up our existing team to go start great companies later

Increasing the visibility of Kenya as a destination for international investment capital (that is primarily profit seeking as opposed to impact-seeking capital of which Kenya has a lot now)

Contributing to building up the business ecosystem in Kenya by doing business with our partner companies (e.g., suppliers, logistics companies)

It will help me develop skills (directly driving impact, knowing how a business works from a 360 perspective, and focusing on moving fast) that will be valuable later in my career in the business or nonprofit world

So what is my job actually like?

At its core my responsibility is:

Have a strong hypothesis about what we need to grow our number of customers by 10x

Test t...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: About my job: "Plans Officer", published by Weaver on September 10, 2023 on The Effective Altruism Forum.

I've been a plans Officer for the United States Army Reserve since 2021. People have no clue what reservists do and much less what "Active Guard Reserve" Soldiers do, and I see it's Career Week, so here's my job. Also I'm shamelessly stealing Lizkas format, so thank her for me writing this.

Note that anything written here is my opinion and does not represent the department of defense, or any US Government agency.

My background & how I got here

I studied Computer Science in college and graduate into the great recession of 2008. After not getting the exact job I wanted (Marine Officer), I went to the Army recruiter and they gave me the option to be an Army Reserve Quartermaster Officer. I wanted something more active, but since Officer Candidate School was very competitive I wasn't in the top.

I served part time as a Platoon Leader, Company Commander and Company Executive Officer. I interviewed with a different unit to change my branch to Civil Affairs, transferred to the unit and then served there as a Battalion Logistics Officer (S4) until I got called to go to first deploy, then school. My small school team of four Captains placed in the top six(Commandants List) of our approximately 60 person class and I was a Civil Affairs(CA) Officer after about 8 years. I credit my time working alongside CA officers as a large part of how well I did.

At the same time I worked as a state department contractor full time. I also was going through Nursing school for most of it, though I didn't make it through.

During my deployment, I was always trying to quantify how much we were doing and I had problems doing so. I read books like Dead Aid and Toxic Charity, trying to get a sense of how to 'do international development' and 'foreign aid'. You might think that this is why I joined EA.

Nope.

My writing partner made me read Harry Potter and the Methods of Rationality and then, yes you can guess the rest. I got picked up for a 3 year tour of AGR and decided to put my civilian career on hold and see if I liked it.

What I actually work on

There is a very specific scope of things that only I do, mostly long range planning, facilitation, writing policy and working on strategy/prioritization for my organization. This includes everything that is mandated for the unit I work for to do to what is directed at a local level for us to do. Most of my work is office work, over various lovely software, and some of it is field work.

Here is what I do normally:

Field questions from people inside the unit

Answer Requests for Information (RFIs) from higher echelons or sister units

Write Operations Orders and iron out short term plans

Be the telephone between two organizations that need to talk.

I also run several meetings that are ad hoc or weekly dependent. I've had to work with a lot of people and I have met so many good people. It's not EA, but they all want to make a difference.

Reflecting on the Role

I opted into a very specialized career field in an already specialized group. I love the people and the unit culture is great, but they ask a lot of the part timers.

Some things I value:

I'm expected to work independently.

If the work is done, then it's done. I'm not waiting around for someone's input normally.

I get to work with top performers regularly.

I get to mentor a lot of junior Officers and Enlisted, which is a big draw.

Some things that can be hard:

I'm not in charge, I just work here. Even though I plan a lot in advance, sometimes the commanders will choose to take a different path, upending a lot of my work.

There's a lot of things that need to be on a regulatory timeline that I can not change. This is fine, except when people try to get around it.

Some Skills I develop in this job:

Lan...

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: Shrimp: The animals most commonly used and killed for food production, published by Rethink Priorities on September 10, 2023 on The Effective Altruism Forum.

Citation: Romero Waldhorn, D., & Autric, E. (2022, December 21). Shrimp: The animals most commonly used and killed for food production.

Summary

Decapod crustaceans or, for short, decapods (e.g., crabs, shrimp, or crayfish) represent a major food source for humans across the globe. If these animals are sentient, the growing decapod production industry likely poses serious welfare concerns for these animals.

Information about the number of decapods used for food is needed to better assess the scale of this problem and the expected value of helping these animals.

In this work we estimated the number of shrimp and prawns farmed and killed in a year, given that they seem to be the vast majority of decapods used in the food system.

We estimated that around:

440 billion (90% subjective confidence interval [SCI]: 300 billion - 620 billion) farmed shrimp are killed per year, which vastly exceeds the figure of the most numerous farmed vertebrates used for food production-namely, fishes and chickens.

230 billion (90% SCI: 150 billion - 370 billion) shrimp are alive on farms at any moment, which surpasses any farmed animal estimate known to date, including farmed insect numbers.

25 trillion (90% SCI: 6.5 trillion - 66 trillion) wild shrimp are directly slaughtered annually, a figure that represents the vast majority of all animals directly killed by humans out of which food is produced.

At this moment, the problem of shrimp production is greater in scale-i.e., number of individuals affected-than the problem of insect farming, fish captures, or the farming of any vertebrate for human consumption. Thus, while the case for shrimp sentience is weaker than that for vertebrates and other decapods, the expected value of helping shrimp and prawns might be higher than the expected value of helping other animals.

Introduction

Recently, Birch et al. (2021) and Crump et al. (2022a) reviewed the evidence of sentience in decapod crustaceans, with a focus on pain experience. Similar to findings previously reported by Waldhorn (2019; see also Waldhorn et al., 2020), these studies concluded that there is substantial, although limited, evidence that decapods might be sentient. Notably, the low strength of current evidence likely corresponds to the little scientific attention various decapod taxa have received-which is particularly the case for shrimp, especially for those belonging to the Penaeidae family (see also Comstock, 2022).

Decapods like crabs, lobsters, and shrimp serve as a major source of human food worldwide, having partly driven the global growth of the aquaculture sector in recent years (de Jong, 2018). Decapod production also represents the fastest-growing major fishery activity worldwide (Boenish et al., 2022). If these animals are sentient, current commercial practices pose serious welfare risks both when decapods are farmed and/or handled during capture, transport and sale, and when they are slaughtered (see Birch et al., 2021 and Crump et al., 2022b). These welfare issues might be particularly pressing given the high numbers of farmed decapods (see Mood & Brooke, 2019a), plus other uncounted individuals captured from the wild.

Furthermore, increasing human population, changes in consumer preferences, technological advancements, and income growth suggest that decapod production stands to increase in the future (Boenish et al., 2022; FAO, 2022f), which may, in turn, augment the scale of the problem of decapod welfare.

Currently, only partial data exists about the scope of this issue. Mood & Brooke (2019a) calculated that between 255 billion and 605 billion decapods are farmed every year, the great majority of such individual...

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 Meta Coordination Forum 2023, published by MaxDalton on September 10, 2023 on The Effective Altruism Forum.

Continuing our efforts to be more transparent about the events we organize, we want to share that we're running the Meta Coordination Forum 2023 (an invite-only event scheduled for late September in the Bay) and provide community members with the opportunity to give input.

Highlights

Event Goal: Help key people working in meta-EA make better plans over the next two years to help set EA and related communities on a better trajectory.

Agenda:

Updates from subject-matter experts from key cause areas.

Discussions on significant strategic questions.

Clarification of project ownership and forming of actionable plans.

Attendees: A group of key people focused on meta / community-building work. Not focused on other key figures beyond the meta space.

Organizing Team: The Partner Events team at CEA (Sophie Thomson, Michel Justen, and Elinor Camlin) is organizing this event, with Max Dalton and other senior figures in the meta space advising on strategy.

Community Engagement: We'd like to hear your perspectives via a survey by 11:59 PM PDT on Sunday, 17 September. The survey asks about the future of EA, the interrelation between EA and AI safety, and potential reforms in meta-EA projects.

Post-Event: A summary of survey responses from the event will be made public to encourage wider discussion and reflection in the community.

The event is a successor to past "Leaders Forums" and "Coordination Forums" but is also different in some important ways. For further details, please see the post below.

Why we're running this event

Now seems like a pivotal time for EA and related communities

The FTX crisis has eroded trust within and outside the EA community and highlighted some important issues. Also, AI discourse has taken off, changing the feasibility of various policy and talent projects. This means that now is an especially important time for various programs to reconsider their strategy.

We think that more coordination and cooperation could help people doing meta work make better plans

We think it could be useful for attendees to share updates and priorities, discuss group resource allocation, and identify ways to boost or flag any concerns they have with each other's efforts. (But we don't expect to reach total agreement or a single unified strategy.)

What the event is and isn't

We think it's important to try to give an accurate sense of how important this event is. We think it's easy for community members to overestimate its importance, but we're also aware that it might be in our interests to downplay its importance (thus inviting less scrutiny).

The event will potentially shape the future trajectory and strategies of EA and related communities

First, some ways in which the event is fairly important:

It will bring together many important decision-makers in the meta-EA space (more on attendees below).

These people will be discussing topics that are important for EA's future, and discussions at the event might shape their actions.

The event aims to improve plans, and hopes that this leads to a better trajectory for EA and related communities.

The event may facilitate further trust and collaboration between this set of people, possibly further entrenching their roles (though we're also trying to be more careful about how much we limit this; see below).

The event will not foster unanimous decisions or a single grand strategy

Some ways in which the event is less important:

It is not a collective decision-making body: all attendees will make their own decisions about what they do. We expect that attendees will come in with lots of disagreements and will leave with lots of disagreements (but hopefully with better-informed and coordinated plans). This is how previous ...

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: Thoughts on EA, post-FTX, published by RyanCarey on September 10, 2023 on The Effective Altruism Forum.

Introduction

This is a draft that I largely wrote back in Feb 2023, how the future of EA should look, after the implosion of FTX.

I. What scandals have happened and why?

1.

There are several reasons that EA might be expected to do harm and have scandals:

a) Bad actors. Some EAs will take harmful actions with a callous disregard for others. (Some EAs have psychopathic tendencies, and it is worth noting that that utilitarian intuitions correlate with psychopathy.)

b) Naivete. Many EAs are not socially competent, or streetsmart, enough to be able to detect the bad actors. Rates of autism are high in EA, but there are also more general hypotheses. Social movements may in-general by susceptible to misbehaviour, due to young members having an elevated sense of importance, or being generally impressionable. See David Chapman on "geeks, mops and sociopaths" for other hypotheses.

c) Ideological aspects. Some ideas held by many EAs - whether right or wrong, and implied by EA philosophy or not - encourage risky behaviour. We could call these ideas risky beneficentrism (RB), and they include:

i. High risk appetite.

ii. Scope sensitivity

iii. Unilateralism

iv. Permission to violate societal norms. Violating or reshaping an inherited morality or other "received wisdom" for the greater good.

iv. Other naive consequentialism. Disregard of other second-order effects

There are also hypotheses that mix or augment these categories: EAs might be more vulnerable to generally psychopathic behaviour due to that kind of decision-making appearing superficially similar to consequentialist decision-making

2.

All of (a-c) featured in the FTX saga. SBF was psychopathic, and his behaviour included all five of these dimensions of risky beneficentrism. The FTX founders weren't correctly following the values of the EA community, but much of what would be warning signs to others (gambling-adjacency, the Bahamas, lax governance) weren't to us pursuing a risky, scope-sensitive, convention-breaking, altruistic endeavour. And we outside EAs, perhaps due to ambition and naivete, supported these activities.

3.

Other EA scandals, similarly, often involve multiple of these elements:

[Person #1]: past sexual harassment issues, later reputation management including Wiki warring and misleading histories. (norm-violation, naive conseq.)

[Person #2]: sexual harassment (norm-violation? naive conseq?)

[Person #3] [Person #4] [Person #5]: three more instances of crypto crimes (scope sensitivity? Norm-violation, naive conseq.? naivete?)

Intentional Insights: aggressive PR campaigns (norm-violation, naive conseq., naivete?)

Leverage Research, including partial takeover of CEA (risk appetite, norm-violation, naive conseq, unilateralism, naivete)

(We've seen major examples of sexual misbehaviour and crypto crimes in the rationalist community too.)

EA documents have tried to discourage RB, but this now seems harder than we thought. Maybe promoting EA inevitably leads to significant amounts of harmful RB.

4.

People have a variety of reasons to be less excited about growing EA:

EA contributed to a vast financial fraud, through its:

People. SBF was the best-known EA, and one of the earliest 1%. FTX's leadership was mostly EAs. FTXFF was overwhelmingly run by EAs, including EA's main leader, and another intellectual leader of EA.

Resources. FTX had some EA staff and was funded by EA investors.

PR. SBF's EA-oriented philosophy on giving, and purported frugality served as cover for his unethical nature.

Ideology. SBF apparently had an RB ideology, as a risk-neutral act-utilitarian, who argued a decade ago why stealing was not in-principle wrong, on Felicifia. In my view, his ideology, at least as he professed it, could best b...

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: Panel discussion on AI consciousness with Rob Long and Jeff Sebo, published by Aaron Bergman on September 10, 2023 on The Effective Altruism Forum.

Intro

Recent 80k guest and philosopher specializing in AI consciousness Rob Long (@rgb) recently participated in a panel discussion on his paper "Consciousness in Artificial Intelligence: Insights from the Science of Consciousness" (pdf) with co-authors Patrick Butlin, Yoshua Bengio, and Grace Lindsay and moderator Jeff Sebo (@jeffsebo).

You can watch it on Youtube (below), watch/listen as a podcast on Spotify , or read the transcript below.

Paper abstract

Whether current or near-term AI systems could be conscious is a topic of scientific interest and increasing public concern. This report argues for, and exemplifies, a rigorous and empirically grounded approach to AI consciousness: assessing existing AI systems in detail, in light of our best-supported neuroscientific theories of consciousness. We survey several prominent scientific theories of consciousness, including recurrent processing theory, global workspace theory, higher-order theories, predictive processing, and attention schema theory. From these theories we derive "indicator properties" of consciousness, elucidated in computational terms that allow us to assess AI systems for these properties. We use these indicator properties to assess several recent AI systems, and we discuss how future systems might implement them.

Our analysis suggests that no current AI systems are conscious, but also suggests that there are no obvious technical barriers to building AI systems which satisfy these indicators.

Youtube description

This event took place on Tuesday September 5, 2023 and was hosted by the NYU Mind, Ethics, and Policy Program.

About the event

This panel discussion featured four authors from the recently released and widely discussed AI consciousness report. This report argues for, and exemplifies, a rigorous and empirically grounded approach to AI consciousness: assessing existing AI systems in detail, in light of the best-supported neuroscientific theories of consciousness. The paper surveys several prominent scientific theories of consciousness, including recurrent processing theory, global workspace theory, higher-order theories, predictive processing, and attention schema theory. From these theories the authors derive "indicator properties" of consciousness, elucidated in computational terms that allow them to assess AI systems for these properties. They use these indicator properties to assess several recent AI systems, and discuss how future systems might implement them.

In this event, the authors summarized the report, offered perspectives from philosophy, cognitive science, and computer science, and responded to questions and comments.

About the panelists

Patrick Butlin is a philosopher of mind and cognitive science and a Research Fellow at the Future of Humanity Institute at the University of Oxford. His current research is on consciousness, agency and other mental capacities and attributes in AI.

Robert Long is a Research Affiliate at the Center for AI Safety. He recently completed his PhD in philosophy at New York University, during which he also worked as a Research Fellow at the Future of Humanity Institute. He works on issues related to possible AI consciousness and sentience.

Yoshua Bengio is recognized worldwide as one of the leading experts in artificial intelligence, known for his conceptual and engineering breakthroughs in artificial neural networks and deep learning. He is a Full Professor in the Department of Computer Science and Operations Research at Université de Montréal and the Founder and Scientific Director of Mila - Quebec AI Institute, one of the world's largest academic institutes in deep learning. He is also the Scientific Direc...

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 Love Letter to EA, published by utilistrutil on September 10, 2023 on The Effective Altruism Forum.

Dear EA,

I love you. In hindsight, my life before we met looks like it was designed to prepare me for you, like Fate left her fingerprints on so many formative books and conversations. Yet, for all that preparation, when my friend finally introduced us, I almost missed you! "Earning to give-me-a-break, am I right?" I didn't recognize you at first as the one I'd been looking for - I didn't even know I'd been looking. But that encounter made an indelible impression on me, one that eventually demanded a Proper Introduction, this time performed in your own words. At second sight, I could make out the form of the one I am now privileged to know so well.

We survived the usual stages: talking, seeing each other casually, managing a bit of long (inferential) distance. It has been said that love is about changing and being changed - Lord, how I've changed. And you have too! I loved you when CEA was a basement, I loved you in your crypto era, and I'll love you through whatever changes the coming years may bring.

I will admit, however, that thinking about the future scares me. I don't mean the far future (though I worry about that, too, of course); I mean our future. We're both so young, and there's so much that could go wrong. What if we make mistakes? What if I can't find an impactful job? What if your health (epistemic or otherwise) fails? In this light, our relationship appears quite fragile. All I can do is remind myself that we have weathered many crises together, and every day grants us greater understanding of the challenges that lie ahead and deeper humility with which to meet them.

I also understand you better with each passing year. And that's a relief because, let's face it, loving you is complicated! There's so much to understand. Sometimes I feel like I'll never fully Get You, and then I feel a little jealous toward people who have known you longer or seem to know a different side of you. When I get to thinking this way, I am tempted to pronounce that I am only "adjacent" to you, but we both know that this would be true only in the sense that a wave is adjacent to the ocean. And you? You make me feel seen in a way I never thought was possible.

We tend to talk the most when some Issue requires resolution, but I hope you know that for every day we argue, there are 99 days when I think of you with nothing but fondness. 99 days when I relish your companionship and delight in my memories with you: traveling the world, reading in the park, raising the next generation, talking late into the night, bouncing a spikeball off Toby Ord's window . . .

I love your tweets, even when they make me cringe, I trust your judgement, even after you buy a castle, and I cherish a meal with you, even when it's a lukewarm bottle of Huel shared in an Oxford train station. When we disagree, I love how you challenge me to refine my map of the world. I am constantly impressed by your boundless empathy, and I am so thankful for everything you've taught me. I love your eccentric friends and rationalist neighbors. My parents like you too, by the way, even if they don't really get what I see in you.

I love you x 99.

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: US presidents discuss AI alignment agendas, published by TurnTrout on September 9, 2023 on LessWrong.

None of the presidents fully represent my (TurnTrout's) views.

TurnTrout wrote the script. Garrett Baker helped produce the video after the audio was complete. Thanks to David Udell, Ulisse Mini, Noemi Chulo, and especially Rio Popper for feedback and assistance in writing the script.

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: Writing about your job is (still) great - consider doing it, published by Lizka on September 9, 2023 on The Effective Altruism Forum.

Many of us know very little about what people in different roles actually do or how they got there. There's a good chance that you have experience that other Forum users would be interested in; so consider writing about your job! (Especially since Career Conversations Week is starting.)

Here are some existing "job profile" posts that you can explore.

(This post is somewhat redundant with Aaron's, but I hope it's a useful reminder.)

Why write about your job

I can present my experience as evidence for this claim. When I was finishing college, I was pretty clueless about my post-college options and kind of defaulted to graduate school (which was the main thing I knew). I had realized my lack of knowledge was a problem, so I asked friends for advice. Someone told me to go on LinkedIn and check people's backgrounds - and to reach out to people and ask for meetings. I did the former and found it helpful, but it could only do so much; I still didn't understand the day-to-day of what an "analyst" did and whether it would be a good fit. And I was too worried about wasting people's time to reach out for calls (except in my more limited personal network, which was heavily skewed towards academic mathematicians and adjacent crowds). I still had a picture-book understanding of jobs. The thing that really helped was talking to people at conferences and informal events later - asking about their backgrounds, what they liked, etc.

The fact that jobs are mysterious to you (especially jobs that are somewhat plausible given your values, strengths, etc.) hampers your career.

You probably won't apply to jobs you know little about, won't know how to build relevant skills, might end up in jobs you dislike, etc. And I expect jobs that EA Forum users have are often more likely a good fit for another Forum user than a random job would be (if only because many Forum users share some values, like wanting to use their careers or resources in significant part to have a positive impact on the world), so hearing about other Forum users' careers can be particularly helpful.

Some specific ways an "about my job" post can help:

Improve someone's longer-term career plan

You might write about a job that's a great fit for one of your readers (or that they hadn't realized was very useful), but that they hadn't even considered as an option.

You might share that you think certain skills are particularly useful for your type of career, and interested readers can try to test those skills.

You can share what you like and dislike about your job, and how much you care about those things, which can inform someone who's not sure if they'd enjoy a certain type of work.

Etc.

Make someone feel a lot better

When someone can't see themselves in the jobs that they're familiar with, they can feel like they will never "fit" a job, or like they'll never have an impact.

For instance, before I encountered EA, I was probably on an academic path, but pretty sad about this fact. My skills were much more generalist than those of some of the mathematicians I knew, and I knew few real generalists. I thought that my lack of highly specialized interests/skills (or my generalist quality) was almost entirely a disadvantage (until I worked at Rethink Priorities and talked to people there).

Impostor syndrome and fear of failure also grow because we rarely talk about the lower points in our past; "about my job" posts can provide useful and encouraging context (e.g. by showing that "successful" people have felt like they weren't succeeding in the past).

Connect with someone

For instance, a reader could be interested in collaborating with you and might reach out after you post.

Or people in your network will ...