Thank you for this. I love Star Trek Discovery too.
Something I'm a bit puzzled by - what would we do to make benevolent AI over and above what we already do? We already train LLMs to do things we (for a given value of "we") find useful and not useless. But a plan for a benevolent AI must require more than that. If you had to sketch out how to teach an AI benevolence, how would you do it?
I love this! Reminds me of Vavilov guarding the seed bank.
One of the worries I've heard shared about longtermism is that it's hard to find interventions that keep on being cost effective when there's a lot of funding to throw at them - like where the last dollar is as valuable as the first dollar? How do you think that would play out here? To what extent are there diminishing returns?
Thank you, this is fascinating. Is there an option to monitor wastewater just from airports (as well as generally for a whole city)? Then anything brought in on international flights might be less diluted and you might be able to detect it sooner, idk?
I realise that the world is a little bit different than in 1918, but given that the Spanish Flu was spread by troop movements, I wonder what the various militaries are doing and if they see themselves as having a role in pandemic prevention?
Thanks for doing this! I find your blog really inspiring and I'm enjoying reading this q&a.
Maybe you've already answered this elsewhere but I'm just curious about the switch from donating to global health charities to doing direct anti biorisk work. Was this motivated by theoretical long termist considerations at all? If you could have switched to doing direct malaria prevention work would you have done that instead? (If you hadn't been able to switch to direct work would you have switched your EtG donations to a long termist cause?) Do you worry much about philosophical arguments for long termism Vs helping people right now, or do you focus on more practical considerations when deciding how to help?
Thank you for this. I love Star Trek Discovery too.
Something I'm a bit puzzled by - what would we do to make benevolent AI over and above what we already do? We already train LLMs to do things we (for a given value of "we") find useful and not useless. But a plan for a benevolent AI must require more than that. If you had to sketch out how to teach an AI benevolence, how would you do it?
I love this! Reminds me of Vavilov guarding the seed bank.
One of the worries I've heard shared about longtermism is that it's hard to find interventions that keep on being cost effective when there's a lot of funding to throw at them - like where the last dollar is as valuable as the first dollar? How do you think that would play out here? To what extent are there diminishing returns?
Thank you, this is fascinating. Is there an option to monitor wastewater just from airports (as well as generally for a whole city)? Then anything brought in on international flights might be less diluted and you might be able to detect it sooner, idk?
I realise that the world is a little bit different than in 1918, but given that the Spanish Flu was spread by troop movements, I wonder what the various militaries are doing and if they see themselves as having a role in pandemic prevention?
Thanks for doing this! I find your blog really inspiring and I'm enjoying reading this q&a.
Maybe you've already answered this elsewhere but I'm just curious about the switch from donating to global health charities to doing direct anti biorisk work. Was this motivated by theoretical long termist considerations at all? If you could have switched to doing direct malaria prevention work would you have done that instead? (If you hadn't been able to switch to direct work would you have switched your EtG donations to a long termist cause?) Do you worry much about philosophical arguments for long termism Vs helping people right now, or do you focus on more practical considerations when deciding how to help?