We just released a new resource, which may become an important part of our site, and which might help reduce some cultural issues in the EA community.
As a little background, 80,000 Hours has always aimed to provide two types of advice:
- Information about specific problems and career paths and their impact
- General strategies & decision-making tips for high-impact careers, such as steps for how to compare two jobs, or whether to focus on career capital or immediate impact.
The key ideas page we released in April 2019 emphasised specific problems & paths over general advice, but it was always our intention to fill in our coverage of both (similarly to how the 2017 career guide covered both).
I think we’ve taken a big step towards filling this gap with our new, in-depth ‘career planning process’.
It’s a set of prompts that take you through how to make a career plan, checking you’ve asked yourself the most important questions, are aware of the best resources, and have taken the most useful steps to investigate.
It starts with high-level questions and then works from there to concrete next actions. Specifically, it covers:
- Clarifying your ultimate goals and what a high-impact career looks like
- Prompts for choosing a global problem to focus on
- Exploring ideas for longer-term career paths you might pursue
- Clarifying our career strategy, based on your answers to the above
- Ideas for your next career move based on the above
- Your alternatives and back-up options
- How to investigate your key uncertainties
- Putting your plan into action
To get started on the process, you can either dive into the article that discusses each step above, or start filling out the associated worksheet:
- An in-depth article with a section covering each step above
- A Google Doc worksheet you can copy and fill in
Later, we hope to release a ‘just the key messages’ version that aims to quickly communicate the key concepts, without as much detail on why or how to apply them. We realise the current article is very long – it’s not aimed at new readers but rather at people who might want to spend days or more making a career plan. Longer-term, I could imagine it becoming a book with chapters for each stage above, which contain advice, real-life examples and exercises. (Added: we'll also consider making a 'tool' version like we had in the 2017 career guide.)
I hope this process will be useful to anyone within (and outside) the community with the good fortune to be able to consider making a big career change, since the prompts don’t depend on which causes you support or career paths you’re able to work on. We’ve had good feedback from some local groups already, and EA Student Career Mentoring have adapted it for their advising process.
I also hope it clarifies a lot of our positions. In Key Ideas, we throw out a lot of options and concepts, but there’s a long way to go from those ideas to an individual's career plan, which takes account of their values, skills, personal constraints, and so on.
To oversimplify the problem, people often get the impression we think everyone should try to work on AI safety as quickly as possible.
This is wrong first because we’d like to see people work on many problems besides AI safety. But just as importantly, we also think there are many other considerations to take into account in career planning: we want people to think about their greatest strengths, how to build valuable skills over time, how to ‘work forward’ from idiosyncratic opportunities, and to consider exploration, personal growth, and many other rules of thumb. All this could easily mean someone's best option isn't pursuing one of our priority paths or working on one of our highest-priority problems.
Finally, I hope this career planning process will help the community reframe what it means to have an ‘effective altruist career'. Effective altruism is focused on outcomes, and for good reasons; but focusing a lot on outcomes can have some bad side effects.
For instance, some of our readers have felt bad that they don’t have the mathematical skills needed to become AI safety researchers. Because they think AI safety is one of the world's most pressing problems, they feel guilty about not being able to work on it and compare themselves negatively to others who can.
Instead, I’d like to make it easier for people to focus on finding the best option for them, which is what really matters in practice.
I hope this planning process will make that idea more concrete: if you work through these steps and think carefully about them, I think you’ll be close to planning your best career.
In my view, if we’re going to try to assess how ‘EA’ someone’s career is, then rather than assessing it based on whether they’re working in a priority path or not, or even whether they’ve had a lot of impact, we should ask whether they've thought carefully about their career plan and put it into practice.
Sketching a plan, trying to put it into practice, and updating as you learn more is, after all, all anyone can do.
The article and worksheet are still in the early stages. We’re keen to get feedback – we have a form here or leave a comment. We’re especially keen to hear thoughts about what you found personally most useful vs. confusing and hard to apply.
We’d also be keen to see filled in versions of the worksheet from readers of the forum -- so if you want to fill out the worksheet and get some comments, email it to firstname.lastname@example.org with the subject line "Worksheet from [Firstname_lastname]". As an incentive, we'll set aside time to give feedback on the first 15 we receive.
Update: We've turned the article into a weekly course and added a 3-page summary.
Hopefully this is one step to making it more digestible.
Here's the course. The link above goes to the summary.
Two suggestions for the list of "broad categories of longer-term roles that can offer a lot of leverage" under "Aim at top problems":
Similar changes could be made to the "Five key categories" in the article "List of high-impact careers".
Makes sense! We've neglected those categories in the last few years - would be great to make the advice there a bunch more specific at some point.
Exciting! I would be curious whether you could give more detail on how the 2020 career planning process differs from the general advice in the 2017 career guide?
Compared to the old planning process (https://80000hours.org/career-planning-tool/), the structure is similar, though this clearly separates out longer-term paths from next career moves. Otherwise, the main difference is that this has far more detailed, specific guidance.
More broadly, there are lots of clarifications to our general advice, or explanations with different emphasis. Some of the bigger newer bits include:
That sounds excellent, thank you so much for the detailed response!
This is still our most current summary of our key advice on career planning, and I think useful as a short summary.
If I was writing it again today, there are a few points where it could be better synced with our updated key ideas series, and further simplified (e.g. talking about 3 career stages rather than 4).
This is fantastic! I know several EAs who "feel guilty about not being able to work on [AI safety or one of the other top cause areas] and compare themselves negatively to others who can," and I think this could be a great resource for them.
Upon skimming the article and the Google Doc worksheet, I'm struck by how long/involved the process is. On one hand, this makes sense-- this is about career planning, and people who are serious about changing their careers should be willing to put in the effort. On the other hand, I wonder if there could be shorter/easier/lower-effort versions of some of these tools.
In its current form, I think the length and user interface of the tool will appeal to some highly dedicated EAs with a lot of spare time on their hands. The tool might be tremendously helpful for them, but I think there are some motivated-but-not-quite-as-dedicated EAs who would benefit from shorter, more streamlined versions.
Some specific ideas include:
I think the main objection to several of these suggestions is that it might lead to shallower reflection than the longer/more effortful version. I think this is fair-- some people who otherwise might have went with the more effortful version might instead go with the "lazier" version (and thus not benefit as much as they could have).
To mitigate this risk, I think you could recommend/nudge people toward the higher effort version, but still have complementary lower-effort versions for the (many) EAs who would be willing to do bite-sized versions of this but not the "18 page Google Doc with many open-ended questions & a complementary article" version. I also think there could be "foot-in-the-door" benefits-- if someone likes the shorter version, they might be more inclined to think seriously about devoting several hours or weeks to more in-depth reflection.
Nonetheless, I think is a fantastic tool and I will be recommending it to several friends :) thank you for making it!
Thank you for the thoughtful comments!
I agree that being too long and overwhelming is perhaps the main issue with it currently. Just checking you saw this paragraph, which might reassure you a bit:
Our top priority was 'just to get everything written down'. After we've had more feedback to check the stages / advice / structure is at least not obviously wrong, the next priority will be making it more digestible, engaging and easier to use. This may take some time, though, since I think we need to give the key ideas cover sheet and problem profiles some more attention next.
Ah, I completely missed that paragraph. Thank you for pointing it out, and best of luck as you create more digestible versions!
After reading the paragraph, I have a few additional thoughts:
I'm sure there are plenty of initiatives going on at 80k, and I have no idea where "creating new short modules/interactive tools for career planning" would rank on the list. Nonetheless, I think it'd be a valuable idea (potentially more valuable than long guides or "key points" materials that are more informational than applied), and I'd be excited to see/share them if you decide to pursue them.
I'm pretty tempted to break it up into standalone sections in the next version.
I agree the tools are worth doing at some point (and maybe breaking up into multiple tools). I guess you're also aware of our 'make a decision' tool that's in guided track?
I think I might be a bit more skeptical about tools though. They take a lot longer to make & edit, and some fraction of our core audience finds them a bit lame (though some love them). Personally, I'd prefer a google doc which I can easily customise, where I can see everything on one page, and easily share for feedback. And it seems like the youth might agree :p
From Ben's post:
"Later, we hope to release a ‘just the key messages’ version that aims to quickly communicate the key concepts, without as much detail on why or how to apply them. We realise the current article is very long – it’s not aimed at new readers but rather at people who might want to spend days or more making a career plan. "
[Edit: Ben said the same thing at the same time, but much more kindly!]
Really excited based on what I've read above. Some very hot takes before I go and read the detail...
It would be great to see some case studies in due course of people who applied this kind of thinking, what choices they made and what they learned; particularly to highlight other high impact careers which don't align with priority paths. And it's easier to make sense of how to use a technique - and how much relative effort to consider for each step - when you have other cases to refer to.
This is very welcomed. Drilling into another bad side effects of focusing on outcomes, I would be curious to see if this approach can help readers to make career decisions which are more compatible with a happy career / life. I suspect us EAs can be prone to a relentless focus on impact in the abstract, absolute sense to the detriment of thinking about what makes me personally impactful, and the latter is more likely to be where an individual will get results from placing their energy. And burnout is well worth avoiding because it's a bloody pox and not fun!
I completely agree. Adding more examples, both lots of quick ones as well as longer case studies for each section, is perhaps our top priority for further work (with making it shorter / less overwhelming as the other contender).