Hi, Andy. As always, I love your reasoning and share it with those who promulgate misconceptions about the relative impact of AI versus other technologies on the environment, especially since I'm quite focused on the environment and climate change as a key policy area. My question is about your disclosures. Do you not have an AI company leader on your Board of Directors at Effective Altruism? That doesn't disqualify you, of course, from having great information and perspectives, but it does raise the bar on your citation of sources and disclosing if you have any conflicts of interest. Thank you.
Thanks Michael! Happy to disclose anything. This didn't come up as something I'd worry about for a few reasons:
1) The person on our board is at Anthropic. We had selected her to give us a perspective on what it's like working there and what's happening in AI lab world more broadly. She doesn't speak for Anthropic and we don't receive any direct money from Anthropic.
2) I personally don't feel any professional pressure at all to speak favorably about the AI labs. If anything, I've had some people in my role suggest I focus more on the negatives on them.
3) This blog is entirely in my personal capacity and I don't see it as an extension of my work at all. I don't receive any kind of compensation or incentive for it beyond what readers on Substack pay.
4) I feel like I'm making all my sources pretty accessible (haven't gone through this specific post and hyperlinked everything yet) so I didn't think I was failing to meet some standard of reporting.
5) More broadly, I try to be very clear that I'm the director of Effective Altruism DC in most places online. EA in general has been involved in different ways with AI labs, so I figured that was the more relevant disclosure.
6) It's pretty important to people on the ground in my organization that I not be deferential to the AI labs at all, and if there were a sense that I was motivated in some way to defend them it'd be a pretty serious problem for my job. Since I mainly think about that in the context of my work I hadn't thought about flagging for my personal blog in comparison, which seems much lower stakes.
Happy to answer specific questions and appreciate the flag.
Just a thank you to Andy for responding to my question - I am delighted that you took the time! Regards, Michael
Hi, Andy. As always, I love your reasoning and share it with those who promulgate misconceptions about the relative impact of AI versus other technologies on the environment, especially since I'm quite focused on the environment and climate change as a key policy area. My question is about your disclosures. Do you not have an AI company leader on your Board of Directors at Effective Altruism? That doesn't disqualify you, of course, from having great information and perspectives, but it does raise the bar on your citation of sources and disclosing if you have any conflicts of interest. Thank you.
Thanks Michael! Happy to disclose anything. This didn't come up as something I'd worry about for a few reasons:
1) The person on our board is at Anthropic. We had selected her to give us a perspective on what it's like working there and what's happening in AI lab world more broadly. She doesn't speak for Anthropic and we don't receive any direct money from Anthropic.
2) I personally don't feel any professional pressure at all to speak favorably about the AI labs. If anything, I've had some people in my role suggest I focus more on the negatives on them.
3) This blog is entirely in my personal capacity and I don't see it as an extension of my work at all. I don't receive any kind of compensation or incentive for it beyond what readers on Substack pay.
4) I feel like I'm making all my sources pretty accessible (haven't gone through this specific post and hyperlinked everything yet) so I didn't think I was failing to meet some standard of reporting.
5) More broadly, I try to be very clear that I'm the director of Effective Altruism DC in most places online. EA in general has been involved in different ways with AI labs, so I figured that was the more relevant disclosure.
6) It's pretty important to people on the ground in my organization that I not be deferential to the AI labs at all, and if there were a sense that I was motivated in some way to defend them it'd be a pretty serious problem for my job. Since I mainly think about that in the context of my work I hadn't thought about flagging for my personal blog in comparison, which seems much lower stakes.
Happy to answer specific questions and appreciate the flag.