| >> They're not looking for the most interesting work at that point, they're looking to keep their job and raise a family - it's a tradeoff. >> Isn't that a risk to move to a place without as many employment options? Yes. The reason for that risk is you can raise kids in a nice house for $500k rather than in a tiny, old, 2 bedroom apartment for 1.5M. You exactly summarized my thought process three years ago. To be fair, i'm very happy here. But it is a mixed bag. Housing is great compared to the bay area, but $500k is a stretch for DC/Virginia (check it out on zillow) though not impossible, there is a ton of inventory if you are willing to drive out a bit. Unlike FAANG folks out-buying houses from you, you have lobbyists and government/intelligence contractors outbidding houses from you. Def better than Bay area though. Except i'm in one of like five growth startups in the area and there are rarely any senior positions open, and the senior workforce appears larger than the pool of interesting senior jobs. Its a huge bet on your employer in addition to the already huge bet you're making w/ taking, say, illiquid early stage stock. So your career growth is much slower. OK, so what if you are OK w/o an interesting job? Then you have plenty of options. Also, plenty of options if you're willing to be a consultant and travel, but now you're making a 2nd tradeoff. All in, very happy but not a panacea and certainly multiple tradeoffs. |
In my specific case my SO grew up in Cupertino and her family is here, so that's another element that makes it hard to leave. I'm currently doing the live with roommates bit, but it's a hard tradeoff to make.
Also I saw where you work in your HN bio - do you partner with VRAD (or maybe are considering starting your own)? I've always though leveraging ML imaging software as part of a nighthawk radiology service would be really interesting. Basically hire radiologists to be part of the company (give them some equity) and leverage their readings to train the model. In a bit of a Tesla style self-driving play where you're providing the current capabilities via humans, but the ultimate goal is leveraging that to train and improve the software. My dad does neurorad so seemed like a natural fit for deep learning image recognition to me. If you're writing the reading software then you can tailor it for training too.
Unrelated to that, if you're looking for more options in DC the company I work for has a major office there (feel free to connect, my email is in my profile).