Pointing out the landmine in his career path was my most rewarding moment as a teacher in the last two years. I'm relieved that he won't waste seven years getting a Ph.D. like I did.
Doing what you love helps with success, but it's not a panacea - you're still subject to the same constraints as everybody else.
Besides, I don't know what things are like in your neck of the woods, but around here, I don't know too many starving PhDs.
Hollywood is still largely an exploitative, winner-take-all system but there are some built-in protections in the system to make it more sustainable. Academia seems to be on the verge of coming undone if they don't do something to keep the exploited class from revolting or just quitting en-masse.
Do what you love that could realistically pay the bills. Otherwise the rest may well not sort itself out. It's okay to try something with little chance of success for a short time, though, especially during youth.
Data scientists are in huge demand, anthropologists not so much. Although there are few jobs in physics, physicists are some of the most brilliant general problem solvers that I have met.
Students with a specific and informed plan can ignore articles like this. But few students have such a plan. The biggest problem within the education bubble is the complete absence of guidance about what any given degree program really can do for you.
China tracks the employment success rate of each degree program. Successful degree programs get more funding, and failing degree programs get less.
Trevor Hastie at Stanford told me that Google revolutionized the outlook for stats PhDs. Prior to Google, statistics was an important field but jobs were scarce, and at places like research labs. But since Google taught the world where the magic lies, Stanford stats grads are inundated with offers from Google, Facebook, Microsoft, Yahoo - and most importantly - countless startups.
Many of the best data scientists I have met studied physics. After the dry complex problems they have faced, commercial applications are human, exciting, straightforward, and a way to make an impact. (yes all our basic advances came from physicists doing physics. Im just saying Data Science could be a great alternative for some people).
I know that I'm biased, so take this with a grain of salt, but tech startups can be an exciting place for a data scientist, where you can make a major impact on the financial success of the business and share in that success.
Happy to give anyone specific suggestions via email.
Yes that's a terrible fate
My "field" = nuclear physics. Since the Russians turned into a bunch of cowardly wusses the opportunities in the 'converting Moscow to a glowing crater' field have diminished somewhat.
Fortunately I have since managed to use those 4years of experimental physics experience to design big bits of the Airbus A380, laser eye surgery machines, automated remote mining operations and now 3D medical imaging systems.
My colleagues who stayed in academia - at least those that missed out on being the 10,000th author on an LHC experiment paper - are grinding out a series of temp posts doing feasibility studies on what experimental result a new series of detectors would produce IF they were ever funded. Which they wont be in their lifetime.
Many of my other ex-students had to demean themselves to using their physics/chemistry PhD to earn a fortune in the city as quants or analysts.
I know of many classmates who got degrees in EE but went to work for a petroleum company, or people with degrees in Mech. Engineering who graduated and started writing code day 1.
I know of none of my classmates who got out and went into counseling or marketing (although I'm sure some exist.)
Stuff like this: "Although the overall unemployment rate of chemists and other scientists is much lower than the national average, those figures mask an open secret: Many scientists work outside their chosen field" should demonstrate why the smartest or most economically aware people who are interested in science might want to think about ancillary fields (like hacking).
I've also co-written a longish guide for undergrads interested in science: http://jseliger.wordpress.com/2012/04/17/how-to-think-about-... , since many of them don't fully understand how science really works in an institutional setting.