This post is a summary of of the important points made in Brian Tomasik’s blog post Advice for Students on Earning to Give.
Majors
While there's no one-size-fits-all answer for what to study in college, I think a reasonably robust general recommendation is to major in computer science (CS) and minor in statistics, math, physics, engineering, finance, or business. If you want, you could make CS a minor and a quantitative subject your major, but it's important to at least get several core CS courses under your belt.
When I began college, I kept all options on the table for my major, including history and linguistics. However, a friend told me I should study a hard science because this would allow me to get any job later, and the friend was really right about this. Hard sciences are the master keys to careers—not just quantitative careers, but all careers. At a career fair in 2006, I talked with a representative of a consulting firm, and when I told her I was planning to major in math, she replied: "Oh, we love math majors, because they can think analytically." The "hardness" of a science correlates strongly with the average IQ of its majors, so studying physics or math signals your intelligence to outsiders.
College is not about studying what's most interesting; it's about signaling your intelligence and skills to future employers. You can (and should) learn philosophy and social sciences in your free time. But having formal qualifications in a hard science opens the most possible doors.
(emphasis mine)
Majoring in something more difficult, even if you struggle to pass a few classes, seems like a better option than following innate curiosity for a wide variety of goals. If you want to reduce suffering, then most of the many ways you could do that would benefit from a harder major. A slightly cynical perspective about the world is extremely useful. Though from my experience, the benefit of extra cynicism starts to hit a diminishing point of return quite fast, and even turns counter-productive at higher levels. In that case, you can even turn the cynicism on itself, and realize your internal explanation of your apathy – that people’s stated explanations are not truthful, and that they serve self-deceptive goals – is probably equally untruthful and self-deceptive.
Internships and Networking
When applying for internships, don't be discouraged if you send out 10 cover letters without any responses. On the other hand, in addition to dispersing applications to employers you don't know, also use your network to contact an existing employee of the company where you're applying because s/he may pass along your résumé more directly. This can be surprisingly easy if your school has an Alumni Directory that allows you to search alumni by their career fields. Alumni are often eager to talk with current students. Several alumni talked with me on the phone for 30-60 minutes each to give me advice. Many more provided insights by email and offered to pass along my application.
Breaking into a network is really easy as long as you're not shy about initiating contact.
Networking is easier than ever in the age of online forums, email, and social-networking sites. One reason the American Dream typically doesn't come true is that underprivileged students are often not mentored along the path toward achievement and so may have weaker academic credentials. They also may not know what questions to ask in order to begin learning about high-paying careers. Needing to work long hours to pay for college can also be a significant obstacle.
(emphasis mine)
Career Changes
One reason I favor careers that pay off right away is that you're very likely to change your mind about what career you want to pursue, perhaps many times during your life. This consideration favors doing things that are immediately valuable—including both earning money and working for an important charity.
You can always start over on a different career path if you abandon your current path. In this sense, I place less importance on "career capital" than 80,000 Hours does. I think it's not that hard to break into a different career later in life.
This may be a slight misinterpretation of Cal Newport’s original concept. Career capital is what lets you negotiate with employers (or, if you are an entrepreneur, the economy at large) to build the life you want to live. If you spend ten years doing one kind of hard intellectual work, you probably learn a great deal of softer skills and tacit knowledge. This experience can be used to negotiate terms in different contexts.
PhDs
If your goal is to earn money, there are very few cases where a PhD makes sense.
I love academia and research, so during college I sought career possibilities where a PhD would have high payoff. I prepared to apply to machine-learning PhD programs, with the intent of working for a startup after I graduated with my doctorate. Time and time again, people told me to skip the PhD and work directly. I had one friend who thought a PhD would pay off in the long run, but this minority view was overshadowed by at least ~5 experts who said the opposite, along with basically all blog posts on the subject. In summer 2008, I decided the weight of evidence was overwhelming and instead set out to apply directly to software companies for a job after graduation.
This proved undoubtedly to be the right decision. Within 3 years at Microsoft, I was promoted (at a normal promotion speed) to the same level at which PhD hires typically started. So by working in a company, I essentially got a fast-tracked PhD-equivalent. Plus, I avoided 5-6 years of lost income that I would have incurred had I gotten a PhD. A few years after I started working, a friend who did get a PhD began applying for jobs very similar to mine. He said: "Brian had the right idea in not getting a PhD."
A PhD is like marriage. It's generally best not to get married until at least your late 20s. When you're young, you might fall in love with a topic and think you want to get married to it, only to break up with it a few months or a few years later.