The grad school choice was easy for me, I was 100% sure I wanted to be an academic. Also, grad school in the physical sciences is a little different than in CS

The grad school choice was easy for me, I was 100% sure I wanted to be an academic. Also, grad school in the physical sciences is a little different than in CS (i think.) Here are some things to consider: Are you currently doing research in a good research lab? Do you like it? Are there any professors/groups that you are ecstatic about working with? Do these profs/groups have funding for a grad student? If you answer yes to all of the above, then you have done enough due diligence to make an educated decision, go work for the groups you identified. So you know, I answered yes to all these questions, and I still hate grad school ;) Fundamentally, and we are all guilty of this, we think that the most important problems to solve are obvious from the get go. Worst, we think important problems must be personally interesting. Try another approach. Solve a problem that's interesting to you, better than anyone else, see where it takes you. The only reason a phd is fun is cause you get to do that.. It's not clear that these style-over-substance startups do not offer you the same opportunity.