You can pick up a Lenovo P1 Gen 2 with decent specs for about 1000 to 1500 on ebay. Max ram is 64MB and usually comes with an i7. The battery life isnt great if it has a built in Nvidia quadro card, I get about 4 hours. I bought it to be able to do ML with it. My only regret is getting one with the nvidia card built in. I believe it gets about 8 hours without it. I ended up buying a RTX 3080 Aorus Gaming Box to do ML instead via the laptops thunderbolt port. No matter what you get, if you want to do ML on it. I highly suggest just getting a laptop with at least a thunderbolt 3 port, and just skip on having an nvidia card built in.
I did the opposite, collected a bunch of salaries for my experience level on levels.fyi, showed them to my boss, and asked for a raise. Got a 50K bump without having to do much. I'm only in the mid 200s TC mind you, but not bad for just looking up what I could be worth.
idk, for me dask was more of a pain to set up than a single node spark cluster, and now that databricks has added a fully compatible pandas api to pyspark, there is even less reason for me to use dask locally.