How to get good at something useful: Part 11 When should you settle in your career?

Steve Jobs used to say you should “never settle.” But that’s not realistic advice. The real question is how to balance the costs of exploration against its benefits. Suppose you’re in a path that’s going OK. Should you stick with it, or should you switch in the hope of finding something even better?

Many successful people explored a lot early in their career. Former British prime minister Tony Blair was a rock music promoter before going into politics. Maya Angelou worked as a cable car conductor, a cook, and a calypso dancer before switching to writing and activism. Steve Jobs spent a year in India taking LSD and even considered moving to Japan to become a zen monk. That’s some serious exploration.

It’s often examples of people who specialised early, like Tiger Woods, that stand out. But that doesn’t mean it’s necessary to specialise early, and it’s probably not the norm, even among top performers. A large meta-analysis found sports people who tried several sports before settling on one tended to be more successful.¹ A 2021 study in Nature found that ‘hot streaks’ among scientists and creatives tended to follow periods of exploring several areas, probably because it allowed them to make new connections between fields.²

More importantly, if the differences in the impact between different career paths are bigger than people think, the case for exploring should be stronger than people think too, because there’s more to be gained from finding something better.

Today, it’s broadly accepted that people will work in several sectors and roles across their lifetime. The typical 25–34-year-old today changes jobs every three years, and changes are not uncommon later too. This is a good thing because it’s very hard to know what job you’ll be good at before you’ve even done any jobs.

But, even if you’re lucky enough to have the opportunity to explore, it will still be costly. Changing career paths can take years, and if you change jobs too often you can look flaky. Some paths can also be hard to re-enter once you’ve left them. And as we’ve seen, the world’s problems are urgent, so you don’t want to take unnecessary detours if you can avoid them.

Fortunately, there’s been a lot of research relevant to the question of how much to explore.³ We apply that here to recommend four strategies for exploring in your career.

Reading time: 10 minutes

1. A rational reason to shoot for the stars

Young people are often advised to “dream big,” “be more ambitious,” and “shoot for the stars,” but is this good advice?

Not always. When surveyed, more than 75% of Division I college basketball players thought they would go on to play professionally, but only 2% actually made it. Whether or not the players in the survey were making a good bet, they overestimated their chances of success by over 37 times. Telling people to aim high doesn’t make sense when they’re already so overconfident.

But, when you have a more accurate view, it can be good advice. Imagine you’re at a slot machine with multiple arms. You believe one arm will pay out $2, and another has a 50-50 chance of paying out either $0 or $3. Which should you pick?

This problem has been studied extensively in computer science. The answer is that, if you can pull only once, you should choose the first (because your expected payoff is $2, rather than $1.50). However, if you get to pull multiple times, you should choose the second. If it turns out that arm pays $3, you can crank it over and over. If it turns out to pay $0, you can go back to the first arm and still get $2.

This approach of starting with the arm that has the highest potential upside is called the ‘upper confidence interval’ algorithm, and it’s one of the best ways to approach these kinds of ‘explore-exploit‘ problems. When it comes to your career options, begin by ranking them according to which has the highest upside scenario, i.e. the option that would be best if things go especially well. Go after the top one. If it works out, keep going. If it doesn’t, move on to the next one on your list.

To give an example, suppose you’re comparing two options:

1. Earning to give in a safe corporate job
2. Trying to make it in AI policy

Imagine you think if you succeed in AI policy it’ll be twice as impactful, but your chances of success are only 25%. If you only get one shot to choose, you should earn to give because it has higher expected impact. But if you have the option to switch back later, then you should try policy first. If it works out, you’ll be in a much more impactful path for the rest of your career, and if not, you’ve only lost a few years of earning to give.

We could say that the policy path has higher ‘exploration value’ than the corporate one. Jobs have high exploration value when:

• They might be really, really good
• But you’re very, very uncertain about them
• And you can test them relatively quickly

In effect, exploration value is a reason to aim high: to pursue options that might be amazing but you’re unsure you can land. There’s often an asymmetry: by aiming high you might find a new amazing career path, while if it doesn’t work out you’re probably in a similar situation to before.

This doesn’t mean pursuing a career you only have a 2% chance of landing, like the college athletes. But it could mean pursuing the option that looks best in the top 10% or 20% of scenarios.

A clear finding of the research is that the longer your time horizon, the more you should explore, and so the higher you should aim. The reasoning is similar to career capital: the more time you have to make use of what you learn, the more useful that learning is.

Here’s an illustration with made-up numbers. Suppose you think that by exploring you have a 50% chance of finding a career that’s four times better than your current default. If your time horizon is the next 10 years, it would be worthwhile spending up to five years searching for the better option (because, in expectation, you’ll have twice the impact for the remaining five). If your time horizon is 30 years, it would be worth searching for up to 15.

Younger people not only have longer time horizons, they also know a lot less about their fit. Thankfully, society is structured to make it easier for young people to explore, with things like internships, gap years, or just the expectation that young people are more likely to switch jobs. So indeed it makes more sense to “shoot for the stars” when young. That might mean aiming for the option that looks best in the top 5% of scenarios. As you get older, you should gradually lower your bar, focusing on the top 10% of scenarios, then the top 20%, and eventually just doing what you expect to be best given what you’ve learned over your career.

Another reason to aim high is that you’re probably selling yourself a bit short. While you might have heard pop psychologists talking about how everyone is overconfident in their abilities, later research showed that this only applies to easy tasks. While more than 50% of people think they’re better drivers than average, when it comes to difficult skills like founding an organisation, becoming an expert in cybersecurity, or helping to tackle the world’s biggest problems, on average people underestimate their abilities relative to others.⁴

Read more about when to be more ambitious

2. Try out several paths (but with careful ordering)

You may have heard of the ‘secretary problem.’ A boss, presumably in an elegant Mad Men–style mid-century office, needs to pick a secretary. Candidates are brought in one-by-one. The boss must decide whether to hire or reject each candidate. The question is, how should they approach the decision in order to maximise their chance of picking the best secretary from the pool?

This problem turns out to have a mathematically optimal solution. They should interview 37% of the candidates in the pool, and then pick the next candidate that’s better than all of the candidates they’ve seen so far.

Popular presentations of this research often suggest the ‘37% rule‘ is the optimal way to make a hire, pick an apartment, career, or even romantic partner. Applied to your career, it suggests you should use about 37% of your total time horizon trying different paths, and then pick the next that seems better than everything you’ve tried so far (i.e. settle).

Of course, this model is a simplification. For instance, it assumes that if an option is rejected, you can never return to it. While this is (mostly) the case in dating, it’s usually not the case with careers, and the ability to return to a past option dramatically increases how long you should explore. On the other hand, it assumes you have no prior information about your options before you try them, which can dramatically reduce how much you should explore.

But even if we can’t say the optimal amount of time to explore is 37%, the basic idea seems correct: a reasonable exploration strategy is to allocate some of your time horizon to trying out different paths, then pick whatever option seems best at the end.

This strategy is an alternative to the first one of aiming high. It’s more costly to try out lots of career paths, so this strategy makes most sense when uncertainty is especially high, such as at the very start of your career.

If you’re adopting this approach, it’s important to do so strategically. First, put more reversible options first. For example, after finishing a PhD, you’re unlikely to land a permanent academic position unless you focus 100% on the academic path, so if you’re unsure about academia it’s worth exploring other paths before your PhD, rather than after. Similarly, it’s usually easier to go from a corporate job to a nonprofit job than vice versa, so if you’re unsure about the two, take the corporate position first.

Next, consider taking a job that lets you try out working in several industries, such as in freelance or consulting positions. This can help you figure out which kinds of cultures you most enjoy. Another option is to take a job that lets you use a variety of skills, as is often the case in smaller organisations, so you can test your fit with those skills.

If you’re already in a job, think about ways to try out a new option on the side. Could you do a short but relevant project in your spare time, or even as part of your existing job? If you’re at university, do as many internships and summer projects as you can.

Third, if you’ve recently graduated, consider making use of the ‘graduate school reset.’ You’re not expected to have your career figured out right away after you leave your undergraduate degree — generally, you’re given licence to try out something more unusual, like starting a business, living abroad, or working at a nonprofit. If it doesn’t go well, you can do a master’s, MBA, law degree etc., and from there return to a more standard path. Instead, many people rush straight into graduate school or other conventional options right after university, which means they miss one of their best opportunities to explore.

Jess — a case study in exploring

Remember Jess Whittlestone from the start of the guide? After graduating with a degree in maths and philosophy, she was drawn to philosophy of mind, but was worried it wouldn’t have much impact. Rather than jumping straight into a PhD, she used the next year to explore.

First, she tested finance. She didn’t expect to enjoy it, and she was right — but now she could cross it off her list with confidence. She also spent time at nonprofits and talked to dozens of people in fields she was curious about. These conversations led her to a PhD in psychology, focusing on decision making by policymakers.

“80,000 Hours has nothing short of revolutionised the way I think about my career.”

Read Jess's story

During her PhD, she kept exploring — interning at a public benefit company advancing evidence-based policy, the Behavioural Insights Team, and writing about psychology online to test the public outreach side of academia. At the end, she could have pursued any of the paths she’d explored so far. But more importantly, she had a much better idea of what would fit her best.

From there, she moved into applied AI ethics, later becoming director of AI policy at the Centre for Long-Term Resilience, a leading UK think tank focused on policy to reduce catastrophic risks. This role drew upon what she’d learned in many of the paths she’d explored over the years. While there, she worked on China-West AI diplomacy, helped major labs develop safety policies, and gave closing remarks at the first international AI safety summit. This work is what led to TIME naming her one of the 100 most influential people in AI.

3. Consider a wildcard

One problem with both of the strategies discussed so far is that you might not even have thought of the best option for you. If you’re not aware it exists, how can you try it out?

This is why many optimal approaches to tackling explore-exploit tradeoffs involve an element of randomness.⁵ Making a random move can help you avoid getting stuck in a ‘local optimum‘ when there’s an entire area you haven’t yet discovered. We wouldn’t recommend literally picking randomly, but we think there is wisdom in trying out options entirely outside your normal experience, things like living in a very different culture, participating in a new community, or trying different sectors from the ones you’ve already worked in.

I went to learn Chinese in China before I went to university. I didn’t have any specific ideas about how it would be useful, just a vague idea that China was important but poorly understood. Many years later, I was able to help run some conferences that helped to kick-start a community focused on coordination between China and Western countries around existential risks. There are now several organisations focused on this intersection, and I see this as some of the more impactful work I’ve done.

I also found it really fun. Chinese is entirely different to Western languages (with a lovely lack of conjugation). Completing daily tasks became like a puzzle, but one that lets you start to access the culture of a sixth of the world’s population. People in China are also extremely encouraging to those learning the language. And then there’s the food.

4. If in doubt, quit

Jim was an urban-planning consultant on the verge of retirement when his son started talking to him about AI risk. Many people would have dismissed the topic: Jim had already had a long career, and he could have easily concluded it was too late in the day to change course. Instead, he threw himself into a career transition: reading, networking, taking classes on AI safety, and applying for jobs. Jim went through an AI safety startup incubator programme and wound up working with Cadenza Labs, a research group working on developing a lie detector for large language models.

Many people who could make a career move like Jim’s instinctively reject the possibility. One reason is the sunk cost bias — the tendency to keep doing something that doesn’t make sense anymore because of how much has already been invested in it. On top of that is status quo bias, whereby people tend to keep on doing what they’re doing because it’s familiar, while switching requires energy and risk-taking.

These biases mean we should expect people to continue in their current path for too long, even when a change would serve them better. A large study found good evidence in support of this idea. Freakonomics author Steven Levitt recruited tens of thousands of participants who said they were deeply unsure about whether to make a big change in their life, such as leaving a job or relationship. After offering some advice on how to make hard choices, those who remained undecided were told to flip a coin to settle the issue — 22,500 did so.

Levitt followed up with these participants two and six months later to ask whether they had actually made the change, and how happy they were now on a scale of one to 10. It turned out that people who made a change in an important area of their lives gained 2.2 points of happiness out of 10 — a larger boost than the typical gain from getting married, or even coming out of a period of depression.⁶ The result suggests a simple rule of thumb: if you’re feeling on the fence between your current path and doing something else, it’s probably time to quit.

Try to imagine your career as if you were starting with a blank slate. If you were picking a job afresh, would you choose the same one? If not, there’s a good chance you’ve fallen prey to one of these biases. If you’re doing something relatively comfortable and prestigious you should be doubly wary, since social approval will likely bias you even more towards staying. The next step is to sketch out alternative options and make them as vivid as possible. Once that’s done, you can compare them side by side on a more level playing field with what you’re doing now.

The value of information

Finding the right career for you isn’t something you’ll figure out right away — it’s a process of improving your answers step by step over time. With each step you take, you’ll learn more about what fits you best.

This means it can be worth taking a job mainly for the information it’ll give you. Even if you try something and fail, you’ve still made progress by eliminating a path. It’s also a reason to aim higher, putting yourself in situations where you have the opportunity to surprise yourself. This means we need to add a fifth and final factor to the framework for comparing career options we’ve introduced: immediate impact, career capital, personal fit, supportive conditions, and finally exploration value.

We’ve also now finished our discussion on how to get good at something useful. In the fourth and final section of the book, we’ll explain how you can bring everything together to create a new career plan and put it into action. We’ll begin by explaining how you can use this career framework to pick your next job.

Read next:  Part 12: How to make career decisions

Continue →

Or see an overview of the whole guide.

Get the whole guide as a book

If you find this guide helpful, preordering the book (especially from a physical retailer!) is a great way to support us.

Preorder the book