Downsides and risks should also be considered. You write:
and even contribute to safe AI governance (e.g. by securing the US AI lead over China)
but it could also accelerate AI capabilities progress, which would leave less time for AI safety work.
There’s also the meat-eater problem, i.e. increasing animal product consumption and factory farming, if we help move people to countries where they’ll consume more animal products.
Regarding the AI Safety point, I want to think through this more, but I note that the alignment approach of OpenAI is very capabilities-driven, requiring talent and compute to align AI using AI. I think one’s belief of the sign of immigration on x-risk here might depend on how much you think top labs like OpenAI actually take the safety risks seriously. If they do, more immigration can help them make safe AI.
Regarding the meat-eater problem, I think the possibility of an animal Kuznets curve is relevant. If such a relationship exists (and there is some evidence it might), speeding up economic growth through immigration to higher-income countries might reduce animal suffering in the long-run.
Or perhaps we’re just in a state of cluelessness here...
FWIW, I believe not every problem has to be centered around “cool” cause areas, and in this case I’d argue both animal welfare and AI Safety should not be significantly affected.
Downsides and risks should also be considered. You write:
but it could also accelerate AI capabilities progress, which would leave less time for AI safety work.
There’s also the meat-eater problem, i.e. increasing animal product consumption and factory farming, if we help move people to countries where they’ll consume more animal products.
Thank you, good to flag these points.
Regarding the AI Safety point, I want to think through this more, but I note that the alignment approach of OpenAI is very capabilities-driven, requiring talent and compute to align AI using AI. I think one’s belief of the sign of immigration on x-risk here might depend on how much you think top labs like OpenAI actually take the safety risks seriously. If they do, more immigration can help them make safe AI.
Regarding the meat-eater problem, I think the possibility of an animal Kuznets curve is relevant. If such a relationship exists (and there is some evidence it might), speeding up economic growth through immigration to higher-income countries might reduce animal suffering in the long-run.
Or perhaps we’re just in a state of cluelessness here...
FWIW, I believe not every problem has to be centered around “cool” cause areas, and in this case I’d argue both animal welfare and AI Safety should not be significantly affected.