As someone who did not attend, this is helpful, thank you.
1: The incentives in Econ academia are terrible. It’s a closed and very hierarchical system in which your career almost exclusively depends from references from top professors, which you will only get if you work on the “approved topics”. At least, this holds for macroeconomics, maybe other fields are more open. Not sure whether EA can do much about this besides offering enough money, and especially prestige, to convince economists to abandon their traditional career ladders.
(Not an economist but I used to work with many of them)
8: I get the same impression, why do you think this is? And how can we as a field reconcile this with the fact that it’s only going to get more common as fellowship spots grow at a much faster rate than full time positions? Related: What do people (or you) think someone should do after 2 fellowships, if they haven’t managed to get a full time research role?
I read this alongside "When digital minds demand freedom" (both great). In the older piece, you treat digital minds as potential strategic agents that might advocate for rights and shape public opinion. They could get to do this by virtue of being perceived as moral patients by some humans, regardless of whether they actually are.
Reading the strategic questions here, I wonder how that possibility slots in, especially in the Society section. Questions like "what memes should we spread?" or how coalitions will form seem to assume a discourse shaped primarily by human actors, but a meaningful share of AI welfare discourse may come to be generated or amplified by AIs themselves. We already see LLMs penning emails to professors; how long until they start contacting politicians too?
To be clear, I don't think this changes the questions themselves so much as it makes answering them harder, especially in the shorter-term when: a) uncertainty about moral patiency is still high, b) the digital minds field is young, c) the potential for AI participation in shaping the field is rising quickly as more agency is granted to current systems.
As someone who did not attend, this is helpful, thank you.
1: The incentives in Econ academia are terrible. It’s a closed and very hierarchical system in which your career almost exclusively depends from references from top professors, which you will only get if you work on the “approved topics”. At least, this holds for macroeconomics, maybe other fields are more open. Not sure whether EA can do much about this besides offering enough money, and especially prestige, to convince economists to abandon their traditional career ladders.
(Not an economist but I used to work with many of them)
8: I get the same impression, why do you think this is? And how can we as a field reconcile this with the fact that it’s only going to get more common as fellowship spots grow at a much faster rate than full time positions? Related: What do people (or you) think someone should do after 2 fellowships, if they haven’t managed to get a full time research role?
I read this alongside "When digital minds demand freedom" (both great). In the older piece, you treat digital minds as potential strategic agents that might advocate for rights and shape public opinion. They could get to do this by virtue of being perceived as moral patients by some humans, regardless of whether they actually are.
Reading the strategic questions here, I wonder how that possibility slots in, especially in the Society section. Questions like "what memes should we spread?" or how coalitions will form seem to assume a discourse shaped primarily by human actors, but a meaningful share of AI welfare discourse may come to be generated or amplified by AIs themselves. We already see LLMs penning emails to professors; how long until they start contacting politicians too?
To be clear, I don't think this changes the questions themselves so much as it makes answering them harder, especially in the shorter-term when: a) uncertainty about moral patiency is still high, b) the digital minds field is young, c) the potential for AI participation in shaping the field is rising quickly as more agency is granted to current systems.