Epistemic status: I’m just a guy, but I’ve been thinking about this for a while.

Forget everything you heard from 80,000 hours. Forget the cause areas, forget everything.

The only thing you need to keep is:

• The understanding and significance of long-tails [0], and how to apply them in real life. And how to apply them even more granularly. Even there, yes.
• Some understanding of aptitudes [1] [2] and how to build them. Some conception of what would a plan be like to try them out. Some experience of experiencing them can be useful.
• A dash of agency (the more the merrier).
• A spoonful of creativity (and willingness, or ability to break down problems into their underlying components and find solutions bottom-up).

Now, let’s start.

The best thing to do, in your twenties, is throw shit into a wall and see what sticks. Try things out.

There was a framework of cheap tests and key uncertainties (or cruxes) that you can use. Part 8 of the 80,000 Hours career guide is pretty solid for this. Here is where agency and creativity comes in hand. Here’s a few examples:

• Find partners for independent research projects at conferences. When I wanted to try out AI policy research, I attended an EA conference, filtered through everyone in the attendee list that mentioned wanting to do research projects in this field, and I pinged them with that request. Talked to 10–15 people in the end and vibed the most with 1–2. Followed up later on and wrote a forum post together. It was shit, but it got the foot in the door.
• Find research topics from senior researchers. How to choose a topic to do (independent) research on. It’s difficult (very difficult) to come up with a brilliant problem to research. It takes years of context and learning to have an intuition on what’s most important right now. However, there are people like that, and they have way more ideas than time to act them. Ask them.
  • Ask them to elaborate if necessary. If the ideas they give you are too vague and you don’t know where to start, ask them to elaborate on a 15-minute call.
  • Think of them as your ‘client’. Think of yourself more like a programmer satisfying a client than a researcher. Your goal is to figure out what this [senior researcher] wants to know, and offer it to them. Boom, after you finish your research you can be confident that (a) at least one person wants to read it, (b) it has a much-much higher chance of being relevant for others, (c) it’s easier for others to follow-up on you and for you to follow-up in on the same topic.
  • It’s an easy transition into a ‘serious’ researcher. If you get really good at it, it’s an easy transition to be a research assistant for that [senior researcher], or to do other actual paid jobs in that particular research field. Bada-bim bada-boom.
• Cold email small organisations or startups. Startups and startup-y organisations (small, agile; within the AI policy field, activist-oriented, filled with young people) are the easiest to just get into. Since they might not have time, capacity or desire to recruit people, they might be very happy if you cold-email them saying you wanna help. If you’re half-decent at thinking, you show up and do the work, they might even keep you! And you save them hours upon hours on recruiting (which they might not have even done).

Do that until you have a better understanding of what skills you are best suited to train. Then train them until you’re 40.

Now, you might wanna have an impact earlier (like me). Or you might be concerned with value drift. Well, best way to stay engaged is to never not be engaged. So you might wanna choose a cause... Right.

Regarding cause prioritization

• Cause prio is difficult. Know that true cause prioritization is very complicated, based on unresolved philosophical issues and on data that nobody has. However, it can still be worth a try. Here’s a course to follow.
• Your cause is valid. Don’t worry if your top cause doesn’t end up one that’s listed on a particular website. But do do the work of comparing and reasoning and finding at least some data. And figuring out why that particular website lists those topics as the most important.
• No really, your cause is valid. If you’re hell-bent on a particular cause, or otherwise choose one that’s not talked about in the effective altruism community at this moment, fear not. Here’s where your skills as a rational thinker come in. You can still use the concepts in effective altruism (like prioritizing interventions based on data), to choose the most effective interventions and approaches within that field.
  • It is still vital to compare interventions. This is where the appreciation of long-tails comes in – taking even half a day to compare your options and choose the most effective one, can mean the difference of having ~no impact or significant impact. This is the effective altruist way. Charity Entreprenurship’s research process is a good north star to aim for.
• Remember what you are maximizing. Remember that if you’re optimizing for skill-building right now, you are operating under the assumption that the bulk of your impact will come when you’re maxxed out on your skills (in your 40s usually). So, the cause you follow while skillbuilding, might not be that important (depends on the particular skill you build though).
• Oh, I guess also check your timelines. If you think AI is gonna kill us in 2 years, or climate change will get us in 5, that’s probably gonna change the way you think about your career :)

What’s most important

Throughout all this is that you take your own feelings and motivations as the first guidance throughout this journey. If you choose a path that you don’t enjoy, there’s not much use. You’ll be miserable and not very effective. You deserve better.

Be prepared that your inner guidance can lead you to unexpected places. That’s also okay.

If you don’t know what you want, take up meditation or yoga or whatever embodiment practice that helps you train that. Do a 10-day course if you feel masochistic. Exercises like Internal Family Systems, Core Transformation, and even CFAR techniques like Internal Double Crux and Focusing help you do that. Group exercises, like Circling might be similarly helpful. The list of options is endless. When in doubt, just spend 3 months in Berlin :D