I have MSc degrees in Physics and Energy with a very broad background outside of this - I consider myself pretty generally-knowledgeable about AI Safety, and participated in John Wentworth's SERI MATS track. I am working on one of the research sprints for this track now.
However I have no published academic papers, and very little "legible" work online, such as blog posts or GitHub. I may try and create some.
I've currently applied for positions at Redwood (it seems for a position which is no longer hiring) and CAIS so far. I am hesitant - for now - to apply for more well-known labs like Anthropic/OpenAI, since these often have many questions about, e.g. "past machine-learning projects" which I would have to leave blank or write very underwhelmingly.
Question resolves YES if I am offered a person-to-person (possibly online, but not like an automated test) interview for any of the AI safety positions I am applying for. While I do not at present have plans to apply for AI Safety PhDs, such interviews would count if, as a PhD I would be considered an employee of the relevant university. Similarly for positions which are not nominally about AI safety but which I applied to primarily with AI Safety in mind (e.g. Neuralink).
Technicalities resolve at my discretion :)