Giving people advice on how to get a job is a minefield, because the way that I get jobs is probably rather different from the way that you get jobs, and so “generic” advice (or even advice from my personal experience) isn’t likely to be that effective. That being said, this may make for interesting reading that helps someone, even if it isn’t perfectly tailored to help you.
For context, my experience on work is this:
After graduating from my master’s degree (quantitative political science), I worked as a Research Manager at Social Change Lab, looking into the effectiveness of protest movements.
In late 2022, I got an Emergent Ventures grant to pursue independent work, which funded this blog and a few other things. I travelled and made connections during this time, as well as writing and doing lots of thinking about what I wanted to work on.
In January 2024, I moved to the Forecasting Research Institute, where I worked under Phil Tetlock on the social science of prediction.
In September 2025, I started working on a new project (soon to be announced properly), and got funding from several sources to direct that.
1) In general, applying for jobs you think you have no hope of getting seems like a good (if slightly demoralising) practice to get into, as does raising your level of ambition. I’ve applied for many, many jobs I thought I had very little hope of getting. For most of them, I was completely right (alas!).
For a few, I got very far in the application process only to lose to a more experienced candidate. This piece of advice would probably be more convincing if I’d actually got a job that I thought I had no hope of getting, but coming very far in the application process is some evidence that people can often be poorly calibrated when it comes to their job prospects.
2) AI should plausibly change your priorities in the sort of job you’re looking for. Jobs that allow you to develop a personal brand and a network are rising in importance, whereas jobs that develop your research skills (or coding, or something else that AIs are likely to be able to do in the next decade) are falling. Take AI seriously!
3) Many people should try to work remotely for American organisations. I did this from London while I worked for the Forecasting Research Institute, and earned more money than I would have in the UK, without any clear sacrifices (other than having to take meetings later than I did when working for a UK-based organisation).
4) It’s important to have colleagues who are collaborative and want the best for you. Working for both of my previous employers, I very much felt that they were interested in what was good for me, as well as what was good for the organisation. I think prioritising collaborative and thoughtful colleagues is important, and you can probably get a sense of these things by arranging Zoom calls with existing employees at an organisation.
5) My view of the things that matter less than people think are:
Pay, probably. Although I do think a reasonable income is important for wellbeing, I think many people tend to overrate it.
Whether you’re extremely interested in the org’s work before you start. This goes contrary to my intuitions, but AFAICT the evidence for this is reasonably good - you will generally start to take an interest in the work you do, even if you weren’t that interested initially.
Whether there’s an obvious path to you to take on more and more responsibility seems slightly overrated. I think often you can just do something for a temporary period and enjoy yourself and get good experience, and then shift to something more traditional (with room for promotion) later on. Let’s see if this will backfire in my personal career!
6) If you haven’t heard of it already, you should look at the 80,000 Hours job board to find jobs that are good for the world and will be very interesting to work on. Applying to have a call with their advisory team is also very valuable, and was useful for me when I was first looking for work. Probably good is another good job board less focused on x-risk.
7) Most career trajectories are boring. You can make a very large effort to pursue something more interesting, and many people should do this. The main trade-off (among others) is that you might have to accept lower pay, although you won’t always have to do this. My view is that many people overestimate the benefits of high pay, and are particularly likely to overestimate it when they’re sacrificing status for higher pay, which you may be doing if you choose a less interesting job. You have a lot of agency in deciding the shape of your career.
8) I also have some experience in hiring and screening candidates. One thing I would note is that the proportion of very bad candidates in a hiring round is quite high. These are mostly people who seem to have found the role through LinkedIn or some other non-specialised job board and written the application as quickly as possible (often using LLMs) as a punt. If you hear there are thousands of applications to some job you want, don’t let that depress you - many of those applications will be laughably bad!
9) I’ve found more success in applying for very few jobs with a very high-effort application than I have from the pray and spray method that sites like LinkedIn encourage. When I got my job at FRI, I think that was the only job I had applied to that month. I spent some time making Anki decks for the interview, and mostly did that rather than applying to other jobs. Frankly, I thought the interviews actually went very badly, but evidently not badly enough to stop me from getting the job. YMMV.
10) Connections seem important, and if you apply to a job where they know who you are already (for whatever reason, not just from an intro), I would imagine that this enormously boosts your chance of success.
For the first time, I am going to paywall some parts of these posts, just to earn a little bit of extra money (given the pay cut I recently took to pursue my own project). If you’ve enjoyed any posts from Samstack and want to contribute, or if you simply want to read locked posts, please consider subscribing. I intend to post much more often in the future, so hopefully you should get some meaningful value out of subscribing.
(I think at some point in the past I promised a free subscription to people who made large donations to effective charities - at one point, I had a spreadsheet of who these people were, but have since lost it. If this is you, please feel free to email me and I will comp you.)
Keep reading with a 7-day free trial
Subscribe to Samstack to keep reading this post and get 7 days of free access to the full post archives.