[lazyjeff]

I'm an assistant professor in computer science and have tried a few things to tune down the pressure in my lab. It's always hard to tell what works and what doesn't due to the very small N (N = number of phd students). I also run a small lab (N = 3 right now) so don't have the scaling and funding issues as larger labs.

The thing that seems to be most successful is having every PhD student work with a couple of undergraduate students. Every project has undergraduate student collaborators, so it's never a solo endeavor. There's momentum from others in the project making progress so you're not pushing the boulder yourself. That way they can commiserate together when the results are bad or I'm being sucky.

The other thing I'm starting to try is before I give any negative feedback in person, I ask them how they think they're doing. If they think they're doing great, then it's clearly a mismatch of expectations. I need to figure out how to make the expectations converge before giving negative feedback. I can't think of anything more demoralizing than thinking you're performing well, and then your advisor unexplicably saying you're not productive enough.

Anyways, I think the environment varies substantially between groups. There's certainly some systematic problems (some department culture, some due to the advisor, and some because of being academia in general). But there's probably as many (or more) research labs as startups. So making blanket statements about advisors/advisees is like saying something like "all startups are under extreme VC pressure to monetize" or that "there's a toxic 24/7 work culture at disruptive startups" is similarly overgeneralizing (but probably true for some large number of them).

[musicale]

Nice ideas. Isolation is definitely an issue, and having them work with undergrads will also remind them that they're not at the absolute bottom of the social status hierarchy. They might get jealous of people who actually have a life though.