Note: This was written long ago when I was quitting my faculty job. I still feel it’s relevant, and so I’m just posting it.

For a few brief decades in the mid-20th century, science was finally not a vanity project of the wealthy, but a career. Private and public enterprises invested in basic research. Bell labs developed the transistor, the C programming language, and the LASER. Government funding agencies sowed the seeds of silicon valley in Northern Californian military aerospace startups. The vast majority of our present day economy was born at that time, typed lovingly behind thick horned rimmed glasses into impractical computers so large humans would step inside of them. 

My humble childhood dream was to become one of these people. High school teachers, PBS, college professors and doctoral advisors all nodded approvingly that my humble plan was solidly within my capability. Even today as faculty I have a growing list of citations, I work with excitement on projects I genuinely feel have transformative social potential. I train students who have gone on to good six-figure jobs using skills I taught them. My work attracts collaborations from Google and cult-like offers of mentorship from people who seem strangely concerned that I quit. Yet regrettably my dream is dead. I am leaving academia of my own free will because I’m certain I can do more science elsewhere.

You can write grants for a living, or travel and self-promote to an audience of old men for a living, make college students feel like they learned, or evangelize science to children for a living, but you can’t really just quietly do honest science. It’s not an option - at least not an option with strong enough odds that any sane person should bet on doing it. One could find far more time to focus on science working in a patent office than in academia…

Anyone still in science will disagree with that statement. Half are lucky enough to have it be untrue for them. The other half are just deluded. Any department chair reading this now would disagree, any recent large grant recipient would disagree. Why should he/she keep reading? My perspective is not some bitter fringe opinion. It’s why our postdocs are measurably depressed. That’s why my friends who got prestigious postdoc fellowships work for Google now, not a university. Even if you think academia is working for you, you should be concerned about how poorly its working for most people.

Why more quit-lit? Well, I have a few simple suggestions: things I think could/should be easily fixed. I don’t mean any offense to my dear friends who are still doing science -  many of whom are doing great honest stuff, nor to NSF program managers. Everyone means well. There’s a banality to the complex evil afflicting academia. Academia is a million earnest people, all making the future worse despite good intentions, because of a lack of information, coordination and sober honesty. It’s an “inadequate equilibrium”.

Here are my modest proposals:

  • NSF/DOD/DOE retain 1% ownership of all grant products. The money is re-invested into grants. 
  • Direct funding of students. Students choose who to work with. They bring their money to that advisor.
  • An immediate end to luxurious travel and workshops. Google Hangouts/webcasts work just fine. 
  • NSF/DOD/DOE must stop promoting a STEM shortage which doesn’t exist, and instead must release honest statistics about career outcomes. All funding requirements to evangelize must end. 
  • Dis-attribution of faculty who contribute nothing to a manuscript. 
  • Require universities receiving public funds to report the fraction that is actually spent on scientific personnel and equipment. 
  • Every type of peer review needs to measure its social biases, and have a plan to reduce them. If peer review wants to be considered legitimate, its very real gender, age, institution, familiarity etc. biases need to be reported alongside citation metrics. (average female score, average score with non-western name, average score with person who has attended your conference etc. and reported alongside anonymized controls) 
  • To whatever extent peer review cannot be unbiased, it should be replaced with a lottery and a search engine. (Github does great without any anonymized peer review!)
  • Identification and breakup of citation circles.
  • No grant opportunities should be written in a way which makes them only accessible to certain institutions. 
  • No grant search or application should take more than a week of time. 
  • Tenure decisions decoupled from how much overhead $$ a PI generates.