[Note, when reading this post, I advocate keeping this important comment from Seb in mind.]

As the CEA team reviews Pareto Fellowship applications, we are noticing a funny phenomenon. This phenomenon afflicts even some of our very top applicants.

These applicants will have an excellent goal G, like "Found a project that fills in some neglected area of the EA ecosystem" or "Have the following exceptional achievement: ________." However, they seem to have far too many indirect capacity-building steps between now and G. This is even the case when it's obvious to me that they already have many of the relevant skills, or access to the relevant resources to get started directly on G.

Here's an exaggerated example plan of the type:

1. Finish undergraduate studies to gain more knowledge and skills.

2. Finish graduate studies to gain more knowledge and skills and credentials.

3. Take a consulting job to gain more career capital.

4. Transition to finance to gain a treasure chest of funds.

5. Finally, go ahead with G.

Example from Rob Wiblin: 

"An example of this that set off red flags for me was someone who wanted to do public health advocacy and was doing a full medical degree first, after a relevant PhD. I thought that was too much training and too indirect as the person had a good shot at getting at public health immediately."

My guess is that these applicants are missing the following pieces:

  • Launching your own independent project often requires a very different set of skills than, e.g., gaining career capital, gaining abstract knowledge, gaining credentials, etc. It's probable that after 5-10 years of indirect preparation, you still won't have the relevant skills.
  • The best way to learn how to launch a successful project is by launching many projects. Many of these will be of not-great quality or they will fail. However, you will find that these costs are not as dire as you thought they might be.
  • Running a project and failing is often better for both your skills and career capital than never having run a project at all. In particular, note that even with many failed projects, you will have built up partially fungible networks and influence. This may be culture-specific, but I've seen many Bay Area entrepreneurs even bragging about how many failed startups they've launched. Personally, I would sooner hire someone who has launched a project and failed over applicants of similar quality.
  • Many successful EA projects (and organizations in general) have been founded or run by people who are very early in their careers or even still at university. See: Oliver Habryka, who is running EA Global this year at the age of 20. He can do this because he cut his teeth on launching/running many similar projects in the past, and is now probably the most qualified person for this project in all of EA.
  • Launching your own project is not as scary as you think. After it's up and running, you will be amazed that you were ever tentative about launching your own project.
  • In my experience (and in the experience of most project-founders I know), formal credentials are massively overrated. You can often get to G without formal credentials if you already have impressive achievements under your belt, or can demonstrate the relevant skills, or learn how to become good at networking, or learn how to become good at social skills. (Speaking from my own experience with myself: even socially-challenged nerds can learn how to become very good at these latter two.)

If you think the text above is relevant to you, I would advise a few exercises that have always helped me:

  • Imagine you try to tackle G and fail. How catastrophic would failure be? (E.g., Will no one ever hire you again? Will you die?) Would it still have been worth it to have tried on expected value grounds?
  • Map out all the steps you currently have planned before G. Now imagine that the earth is going to explode, and the only way to prevent this from happening is by cutting 80% of the steps. Which steps do you cut? Were these steps utterly critical - i.e., could you still get to G without them?
  • Is there a minimal version of G you can dip your toes into now without committing?
  • If your plan depends on getting formal credentials, ask yourself, "Might I get to G faster by racking up achievements, or by demonstrating the relevant skills that I already have, or by becoming good at networking, or by learning better social skills?"
Final notes and caveats:
  • Obviously, if everyone and their brother-in-law starts launching their own project, this will harm the signal-to-noise ratio inside EA. However, I'm fairly confident that we are far from saturation.
  • As Ryan Carey notes: "This kind of article is going to move a lot of people to be more rash, for better and for worse." So here's the caveat: launching your own project is still really freaking hard. E.g., most businesses fail and many more don't provide much value. Grit and operational planning and using tools like Guesstimate to create theories about your project's expected value are still going to be critical. But even provided such things, just as one falls of the bike when learning how to ride it, you should expect the equivalent when learning how to launch projects by launching projects. Start with training wheels (but also, push yourself down a hill). Anyway, importantly, this post isn't about how to create the best world-improvement projects, but rather that:
    • (a) Launching projects often doesn't require as much indirect capacity-building preparation as I've observed people doing in EA, and a lesser point that, 
    • (b) Launching projects is probably the best way to learn how to launch projects, rather than a whole bunch of indirect capacity-building steps. This is probably explained by something similar to context-dependent memory and the encoding specificity principle, but, in my grandma's words, "use your eyes." If deliberate, direct practice in X wasn't the best way to learn X, wouldn't that be surprising?


ps I recommend this post with a similar theme from 80,000 Hours: "What people miss about career capital: exceptional achievements."

pps Don't worry if you're one of these Pareto Applicants. If you're one of the best fits, you'll still make it in!






More posts like this

Sorted by Click to highlight new comments since:

A lot of people start their project later than they should, and a lot start it earlier. So this kind of article is going to move a lot of people to be more rash, for better and for worse. The problem is that audiences probably self-select these articles based on fulfilling their existing preconceptions. On net, it's not obvious whether or not this ends up being useful, even granted that you think most people move too slowly on their projects (which could as easily be your mistake as theirs anyway).

I reckon it would be more useful to actually discuss some concrete situations (deidentified mashups of career stories) so that people could actually know how rash you think people were beforehand and where they're supposed to end up. This would help people to know how they're supposed to generalise your arguments to their own example, and would also help the community to figure out how it actually disagrees about the topic, rather than just pushing abstract arguments back and forth.

Strongly agree with this. I think there was some good material in the opening post, but I downvoted because I thought it was likely net-negative because of this dynamic of audiences self-selecting.

I think the advice ought to be: you should seriously consider just starting on a project to pursue G. But you shouldn't actually launch into G without getting a whole lot of outside views on whether it's a good idea, because it can be hard to tell from the inside, particularly if you don't already have deep expertise.

I agree with Owen's comments and the others. The basic message of my post, however, seems to be something like, "Make sure you compare your plans to reality" while emphasizing the failure mode I see more often in EA (that people overestimate the difficulty of launching their own project).

Would it be correct to say that your comments don't disagree with the underlying message, but rather believe that my framing will have net harmful effects because you predict that many people reading this forum will be incited to take unwise actions?

I agree that people should try to reason about the factors actually constraining the success of their project and try to unblock those.

But I don't agree that most people need to spend less time qualifying themselves to complete their projects. Sure, people should question whether their success is constrained by their knowledge and qualifications, but often it is.

I'd rather say:

  • Do you need a degree to emigrate?
  • Do you want to work for an existing policy organisation?
  • Do you want to work as an academic?
  • Do you want to do targeted movement-building with academics?
  • Do you want to network with many researchers? If so, your success will probably be limited if you don't get an undergraduate degree. You may need graduate education also.

I guess you'd agree. Most of what's done by leading EAs like Bostrom/Ord/MacAskill needs (or can be greatly helped) one or more degrees. Even for people like Musk or Yudkowsky who are seemingly completing their own projects, degrees can help in many ways.

People who are devoted enough to drop out of their degree at a young age based on reading complex online arguments are the same sorts of people whose influence in policy I'd want to preserve. These people need to be encouraged to make a 5+ year long plan, and to stay the course.

Sure, there are plenty of people who reach their thirties and forties and never quite manage to pull the trigger and start their own project, or move to that nonprofit job. And for that audience, you need to deliver your kind of message.

Sure, if you have extensive transferable skills, and want to perform activities like grassroots outreach and fundraising or starting a company of a less technical variety, then going in with few qualifications is appropriate. But that's a hard way and it's not the only way.

Another way of thinking about it is that there are intrapreneurs (people who pursue influence within an organisation) and entrepreneurs (people who go off and start their own thing). People who work within a system or start over outside of it. For the people who are insiders and intrapreneurs, I'd prefer they read a motivational post that had roughly opposite content to what's here.

I can only speak for myself, but I would say your message could have been delivered much more optimally and made a significant net positive impact by doing so.

Rather than telling people "go start stuff up," it could have said "go learn the skills and knowledge necessary to start stuff up, which are much less extensive than you think them to be, while also ensuring you build up the emotional and cognitive tools needed to deal with failure and updating your beliefs about your project."

My thoughts, in other words, are not a rejection of the fundamental message, simply a desire for a more optimal delivery of the message.

A lot of people start their project later than they should, and a lot start it earlier. So this kind of article is going to move a lot of people to be more rash, for better and for worse. The problem is that audiences probably self-select these articles based on fulfilling their existing preconceptions.

Good point. For this reason, advice that is customized to a particular person is in general to be preferred to generic advice, especially in tricky areas with lots of sign-changing factors. But of course customized advice is not always possible.

I reckon it would be more useful to actually discuss some concrete situations (deidentified mashups of career stories) so that people could actually know how rash you think people were beforehand and where they're supposed to end up.

I agree with this criticism, and would also suggest that it would be more useful to discuss what kind of knowledge about launching projects people should try to get before launching projects. There's plenty of literature about launching projects, and plenty of EAs can provide guidance and expertise about launching projects.

Often (in EA in particular) the largest cost to a failed started project isn't to you, but is a hard-to-see counterfactual impact.

Imagine I believe that building a synth bio safety field is incredibly important. Without a real background in synth bio, I go about building the field but because I lack context and subtle field knowledge, I screw it up having reached out to almost all the key players. They would now are be conditioned to think that synth bio safety is something that is pursued by naive outsiders who don't understand synth bio. This makes it harder for future efforts to proceed. It makes it harder for them to raise funds. It makes it harder for them to build a team.

The worst case is that you start a project, fail, but don't quit. This can block the space, and stop better projects from entering it.

These can be worked around, but it seems that many of your assumptions are conditional on not having these sorts of large negative counterfactual impacts. While that may work out, it seems overconfident to assume a 0% chance of this, especially if the career capital building steps are actually relevant domain knowledge building.

Agreed. This updates my view.

Would asking people on the street if they'd be willing to donate money to effective charities such as AMF (or similar marketing efforts to try to raise money quickly rather than focus on high quality movement building) have this negative counterfactual impact?

What is a good way to evaluate this risk of a negative counterfactual impact for candidate projects one is considering launching in general?

I think the classic 'drop out of college and start your own thing' mentality makes most sense if your own thing 1) is in the realm of software development, where the job market is still rather forgiving and job opportunities in case of failure abound 2) would generate large monetary profit in the case of success

Perhaps many Pareto fellowship applications do not meet these criteria and therefore applicants want to be more on the safe side regarding personal career risk?

While it may be implied, I don't get the feeling from reading this article that it supports the idea of simultaneously working towards career goals the traditional way and launching projects at the same time. If it does, please disregard the following paragraph. Instead, it feels like the article is recommending young people to "try to tackle G and [take on the high risk of failure, even repeated failure]."

To counter the idea of dedicating one's early life entirely to project launching, I would like to note that this strategy may not suit the vast majority of people, especially not in the age group that the Pareto Fellowship is targeting. Students usually have no career capital and no funding. That leaves them with no specialized skills or startup funds to launch and follow through with a project, and if the project fails or does not generate enough revenue for financial stability, the young person's strategy also results in no fallback plan besides potentially taking a low-paid starting position and living at home if a project fails and they have lost their degree by dropping out to pursue a project full-time. Even older individuals frequently cannot generate the funds necessary to pursue a project full time, let alone sustain it and recover if the project does not succeed, especially with obligations like childcare and saving for retirement. I agree that I will not be killed off if a project fails, but the risks and uncertainty are simply too high to devote 100% of my resources to pursue something so uncertain. I'll note the oft-cited statistic that nine in ten startups fail. Probably because of the high risk of failure with the project and in achieving future career goals, 80000 Hours does not seem to recommend serial entrepreneurship to most young people looking at potential future careers.

As one of the Pareto hopefuls, and one who's "G" may have been alluded to in the article, I'd like to point out that I am currently following the dual-strategy of launching projects (including my G) and simultaneously working towards generating income and building career capital. I believe that this strategy is far superior to dedicating 100% of resources and time towards early-life project launching because it gives people the necessary background and benefits to launching and managing projects while also overcoming the inherent risk of losing everything. Also, Oliver Habryka and many other EA's I've met appear to be implementing my "dual strategy" as well. I suggest that the article suggest this dual strategy as an alternative, or the preferred alternative, to 100% early-life project launching, or clarify that this is the idea the article meant to convey.

one in ten startups fail

You probably meant nine in ten startups fail.

Thanks for the catch! Post edited.

Fascinating - this ranks as both my most downvoted and most shared post of all time.

= controversy sells.

(b) Launching projects is probably the best way to learn how to launch projects, rather than a whole bunch of indirect capacity-building steps... If deliberate, direct practice in X wasn't the best way to learn X, wouldn't that be surprising?

I have a somewhat different perspective. I think it's highly beneficial first to learn about the fundamental aspects of launching projects before launching one. There's plenty of literature available on the vast majority of projects a person might want to launch. Likewise, there are experience project launchers one can talk to about launching projects. Just like it's better to learn to play piano using a textbook and a piano teacher, having background knowledge and knowledgeable personal guidance is better than just trying it. After that, you can do the deliberate practice that will take you to mastery.

Personally, I would sooner hire someone who has launched a project and failed over applicants of similar quality.

I would only do so if the applicant was wise about launching the project. If the applicant launched the project in a foolhardy way, this indicates to me someone who is way overconfident, and I wouldn't want this person working for me.

When I learned about x-risks in 2007 I decided to do something about it and the only thing I could do at the moment was to translate main articles by Bostrom and EY into Russian. I spent half a year doing it and it had two results. I improved my English and I learned a lot about the topic so I started to get my own ideas about it which resulted into a book in Russian. So I think it is right thing to do - directly go to your goal.

Props for doing this. I was recently reflecting that it would be great to have a bunch of the LW Sequences or other works describing AI value-alignment problems translated into Chinese. If anyone who knows Chinese sees this and it seems like their kind of thing, I'd say go for it!

Thank you for this inspiring post Tyler! I was not accepted into this year's Pareto Fellowship, but this does not mean I do not want to make a difference in the EA Movement, and this is a timely reminder that I can still make a difference as an MSW student.

Curated and popular this week
Relevant opportunities