TL;DR: EA Rwanda is piloting a three-week career advising workshop with Rwanda-based students and professionals. In Week 1, rather than asking participants what job they wanted, we asked them to find evidence of where they have already been useful, ask two other people what they seem unusually good at, and identify patterns.
This is anecdotal evidence from a small pilot rather than an evaluation, but one thing already seems worth sharing.
Several participants rated themselves as fairly clear about their career direction. The average was 8/10. Yet when we asked what they would probably do over the next 90 days if the workshop did not exist, the answers often sounded like:
"I will apply for jobs and internships, learn more skills, continue research, attend events, network or take opportunities as they appear."
All the above are sensible things, but often without a strong mechanism for deciding which of those activities deserved disproportionate effort.
One participant put it better:
“I would likely be making progress, but mostly by trying different opportunities rather than following a clearly defined career strategy.”
That made me wonder whether some early-career people are not suffering primarily from a motivation problem or even an information problem. They may have a prioritisation problem.
Instead of beginning Week 1 with “What career do you want?”, we asked participants to reconstruct evidence. They each identified one past experience where they had felt particularly useful. Then they asked two people: when have you seen me doing some of my best work? What do people seem to rely on me for?
Finally, they had to complete:
“The evidence so far suggests I may be particularly good at ______.”
One participant described repeatedly turning messy field data into information other people could actually use for decisions. Another had helped turn youth-unemployment data into a practical solution during a data-science hackathon. Another coordinated teenagers, volunteers and leaders through a demanding camp and discovered that coordination and delegation seemed to come unusually naturally.
Another participant had helped mobilise hundreds of people for a charity activity supporting vulnerable students. One had developed an AI model aimed at identifying malaria parasites from low-cost microscope images. Someone else had analysed epidemiological and environmental data around dengue in Africa.
Four of the six strongest patterns centred around research/data/problem-solving; two were more clearly around people leadership and coordination.
A CV is usually an inventory of where somebody has been. Career advising needs to help reveal what produced value while they were there.
I realised that “Data analyst”, “student”, “volunteer” and “researcher” are weak descriptions for career planning.
More portable hypotheses look more like “Turns unstructured information into decisions, moves a group from ambiguity to execution, persists on technically difficult problems when the application matters, explains difficult things so other people can act.”
It seems easier for these hypotheses to be tested against many different roles and cause areas.
There are already much more sophisticated career resources in the EA ecosystem. I don't think a local group should try to replicate them, but a local group may have a useful complementary role before and between those resources.
These can include helping people make their own evidence more legible; helping them investigate locally accessible problems and pathways; creating accountability for small career experiments; making introductions; and noticing promising people who would benefit from more specialised advising.
A whole lot. The 6 responses, from a cohort of 13 people, tell us very little.
The clarity scores are self-reported, and we have no counterfactual group. Also, identifying a promising strength is several causal steps away from actually having more impact.
The outcome I care about, as the Lead Facilitator, is not whether participants finish three workshops saying they feel clearer. It is whether the process changes consequential behaviour:
That's what we want to follow, during and after the workshop has ended.
For Week 1, we looked backwards for evidence. The next step is to look outward.
Participants will use their emerging evidence to construct multiple career hypotheses, investigate important problems, identify their biggest uncertainties, and design cheap ways to test those uncertainties.
I would particularly like to learn from people who have run local or regional EA career advising:
I'll share what we learn as the cohort progresses.
