Something that this MATS resource expanded on and that Matt B. developed beautifully is the timeline breakdown across 3-6 months: https://docs.google.com/document/d/1tYj50GtJ7JS78w2WugoLaUF-b7XyE_VmY7XDRZs-l_s/edit?usp=drivesdk It helps those interested transform their plans into granular actions across those weeks that has been really helpful in context building. Thank you for your writing!
Wondering what the best person for context week would be? If that has been figured out I would love to know what the thinking is? I could see a case for many different individual types.
This post gave great context about how to give people context. Thanks for the context!
lovely week! thanks harry and josh <333
Something that this MATS resource expanded on and that Matt B. developed beautifully is the timeline breakdown across 3-6 months: https://docs.google.com/document/d/1tYj50GtJ7JS78w2WugoLaUF-b7XyE_VmY7XDRZs-l_s/edit?usp=drivesdk It helps those interested transform their plans into granular actions across those weeks that has been really helpful in context building. Thank you for your writing!
Wondering what the best person for context week would be? If that has been figured out I would love to know what the thinking is? I could see a case for many different individual types.
For this iteration:
- someone who is committed, but needs time to sit down and sort out their threat models / uncertainties before gaining confidence in an intervention
- someone on the periphery of AI safety who needs the "shock" of thinking about this issue seriously with others to act
For BlueDot's future context upskilling programs I expect there to be a slightly different angle.
I think the findings from this would be extremely useful for university organizing.