I feel like I keep saying this, but I’m very glad you keep posting stuff like this. There’s a time and place for pie in the sky doom forecasting, but even if you’re a true believer, someone has to hammer out the in-between.
And if you’re not a true believer (like myself), this is arguably more important than the long-term alignment research or pouring a million man-hours into mech interpretation or whatever. I feel like EAs often overlook the governance side of things (they say it’s “core priority” but in practice I see far more engineers than public servants) so I really appreciate the work you’re doing!
I feel like I keep saying this, but I’m very glad you keep posting stuff like this. There’s a time and place for pie in the sky doom forecasting, but even if you’re a true believer, someone has to hammer out the in-between.
And if you’re not a true believer (like myself), this is arguably more important than the long-term alignment research or pouring a million man-hours into mech interpretation or whatever. I feel like EAs often overlook the governance side of things (they say it’s “core priority” but in practice I see far more engineers than public servants) so I really appreciate the work you’re doing!
Thanks so much for this Harjas! This means a lot. Getting feedback like this is very validating and tells me I'm not alone.