Rules of Career Making

A series of tips provided by Jan Peters for postdocs about to make their next career move.

1. Every career move should be a step up.

2. Choose the right duration
Productivity increases with time. Novelty follows the opposite trend. The optimal job switching time is between 3 and 5 years.

3. Change continent for bigger gains.

4. What's stepping up?
PhD to postdoc, depends on the respective quality of the university/group of your PhD and postdoc. As a postdoc, doing another postdoc is usually considered a step down but you can bounce back if the group of your second postdoc is uniquely good or you have done amazing work.

5. Look at the market

  • 1990: The Russians clog the US prof. market.
  • 1997: AI winter: Stefan Schaal was last roboticist to be hired until 2003
  • 2000: Internet bubble: computer systems scientists all got jobs
  • 2017: AI/ML hype: make other people think you are good then Deep Mind, Uber, etc. pays any price.

While the last two entries constitute examples of favorable markets, keep in mind that a current bubble predates a future direly bad market.

6. First create opportunities, then select!
Big opportunities: Dec-Feb are the deadlines in the US for faculty positions. Lecturer positions (UK) pop up when the REF (in the UK, REF decides on the state funding of the school in future based on the faculty members employed by the university) is closer.
Apply whenever you can to see your market value and find out whether the shoe fits

7. Never reject an opportunity based on just your prior!

8. Create your own hype!

  • Reasons to post to the mailing lists (workshops, special issues, ...)
  • Research successes

9. Give sufficiently many REAL job talks (not just test talks)

10. Start high, go low only if you don't succeed

Minor things

  • If you go to Asia, rather stay there (as postdoc) or go for an internship.
  • Going to the US is often considered a step up (if you go to a top 5 school at least).
  • For the UK: higher than postdoc position is a step up. If you go just as a postdoc, go for the top 5 (Cambridge, Oxford, UCL, Edinburgh, ICL)