This is short because I’m mainly trying to gather feedback on how much the idea might be worth pursuing, but how much value does it seem the existence of an open access AI safety journal would provide? Some reasons I can think of in favor:
Have a journal focused on AI safety, giving the field better visibility.
Peer review from researchers active in AI safety research.
Open access venue for AI safety research to avoid ethical concerns with publishing in closed journals.
A venue where AI safety ideas too far outside the mainstream but are academically rigorous can be published.
Give AI safety researchers a respectable venue for publishing to satisfy their career needs while reducing the effort they have to expend to get their work published, allowing more AI safety research to happen.
Some reasons against:
There are already plenty of journals.
Pre-prints are good enough.
Journals are a waste of time to signal quality that is generally easy to assess from reading articles themselves.
Most of the value could be had by regularly publishing a reviewed list of pre-prints and otherwise published articles in AI safety.
Drag on time of folks who could be doing research instead of reviewing and editing articles.
It appears that starting an open access journal is a little tedious but not especially hard, sc. this step-by-step guide, so it could probably be done by a team of 1-3 volunteers, but hopefully could be professionalized if it takes off.
Idea: Open Access AI Safety Journal
This is short because I’m mainly trying to gather feedback on how much the idea might be worth pursuing, but how much value does it seem the existence of an open access AI safety journal would provide? Some reasons I can think of in favor:
Have a journal focused on AI safety, giving the field better visibility.
Peer review from researchers active in AI safety research.
Open access venue for AI safety research to avoid ethical concerns with publishing in closed journals.
A venue where AI safety ideas too far outside the mainstream but are academically rigorous can be published.
Give AI safety researchers a respectable venue for publishing to satisfy their career needs while reducing the effort they have to expend to get their work published, allowing more AI safety research to happen.
Some reasons against:
There are already plenty of journals.
Pre-prints are good enough.
Journals are a waste of time to signal quality that is generally easy to assess from reading articles themselves.
Most of the value could be had by regularly publishing a reviewed list of pre-prints and otherwise published articles in AI safety.
Drag on time of folks who could be doing research instead of reviewing and editing articles.
It appears that starting an open access journal is a little tedious but not especially hard, sc. this step-by-step guide, so it could probably be done by a team of 1-3 volunteers, but hopefully could be professionalized if it takes off.
Thoughts?