Task Y seems analogous to GiveDirectly—less effective that the top option but with the ability to take on many more resources.
Some ideas for task Y. I’m mainly spitballing, so don’t take them too seriously:
Trying to convince friends and families of good policy ideas—EAs could be encouraged to learn and spread good policy ideas. I agree that politics is a polarising, but policy need not be zero sum.
Forecasting effective causes. I think there should be a board for forecasting impact of different causes. People can vote and then whenever GiveWell gets round to checking, you could test accuracy. Low cost and the accuracy might be better than random.
Testing decision-making interventions at work. If there is an effective decision making strategy then EAs could test it at their workplaces. This would lead to lots of case studies and adoption if effective.
Task Y seems analogous to GiveDirectly—less effective that the top option but with the ability to take on many more resources.
Some ideas for task Y. I’m mainly spitballing, so don’t take them too seriously:
Trying to convince friends and families of good policy ideas—EAs could be encouraged to learn and spread good policy ideas. I agree that politics is a polarising, but policy need not be zero sum.
Forecasting effective causes. I think there should be a board for forecasting impact of different causes. People can vote and then whenever GiveWell gets round to checking, you could test accuracy. Low cost and the accuracy might be better than random.
Testing decision-making interventions at work. If there is an effective decision making strategy then EAs could test it at their workplaces. This would lead to lots of case studies and adoption if effective.