Maybe unpopular advice, but I do find that companies over-index on your current role. (FWIW I have hired lots of new and experienced PhDs in ML but am not hirin

Maybe unpopular advice, but I do find that companies over-index on your current role. (FWIW I have hired lots of new and experienced PhDs in ML but am not hiring now.) So if you take a "placeholder" job, I suggest trying to find one that sounds at least as prestigious as your current position. So, changing your resume from "Fresh MIT PhD grad on the market" to "Researcher in boring lab of dying old industrial company" will be a bad move. If your passion is AI, it's a good idea to go for a placeholder job to be in AI/ML as well. Startups can be good unless they are obviously ridiculous or terrible ones - managers are used to seeing unknown startup companies and fresh grads, and you'll probably be forced to do lots of productive coding. But a bad research lab can be really bad - trailing edge research, bad code, not enough resources to do something meaningful, publishing in low-tier venues that nobody reads, ugh....