This time I will explain several important things about grad schools that don't appear in universities' handbooks. When most people think about education, they usually envision a curriculum, teachers, students, assignments, exams and marks. All these do not exist in graduate programs. It wouldn't be research if we could define beforehand how students need to graduate. The program really begins when the student realizes that everything he doesn't know is his fault.
First and foremost, research is a contribution to society. That means that researchers should not expect to get a cookie for every good thing they do. There are awards involved in research, but these are just artificial means to add excitement to the process. Very often the small contributions stay invisible to the public, but they accumulate together. What really matters is the production of the entire community, not specific individuals. Many people do not fully appreciate how many things around us today were shaped by research at universities many years ago. For example, consider the process that occurs with every keystroke on the keyboard. It involves the keyboard, computer, cpu, operating system, application program and so on, requiring knowledge in materials, physics, electrical engineering, mathematics, manufacturing etc. All these things didn't appear from nowhere.
Graduate school is also a privilege. Few people get the opportunity to live without an alarm clock, explore whatever they are interested in, where nobody is telling them what to do. But there is responsibility involved. I once heard a talk by a professor from a prestigious university on the use of punctuation marks such as commas and hyphens to improve text analysis. I have nothing against such projects outside universities, but there are way more important problems. In contrast, a mathematical-physics student once told me that in some models of the universe, there is a discrepancy of more than ten orders of magnitude between theory and observations. This is serious!
Often people tend to assume that people who earned the same degree studied the same things. However, every phd should be different by definition. When people can't compare things, the natural tendency is to compare what they can, like the page count, publication venues or citations which are irrelevant to appreciate the content. For example, this year two students from my office published papers at the same conference. One student worked on his paper for about three months. The other spent more than three years just to set up a system on which he can run his experiments. If we disregard the content, the papers are worth the same, but their content is completely incomparable. Reading such works is not enough for evaluating them. The tricky part is to know where the entire field was when they were published.
Many people ask why grad school takes so long. From the inside time flies very quickly, yet there are several reasons why it appears long from outside. First, people are more ready to accept ten people working on a project for a year than a single person working for ten years. There are professors who will sign their students on every publication coming out of their lab, while other professors will sign on their student's papers only if the student did everything on his own. During these years researchers are doing many things that never make it to the published documents. Secondly, current research requires a much higher technical starting point than it used to be, whereas the finish line is not well defined. In areas like computer vision, people are solving the same problems again and again, and yet the problems remain open. Third, research often takes longer for people with industrial experience. It requires adjusting to the mindset of searching for good open questions, whereas in industry people focus on making things to work based on well-established ideas. Another reason I can think of why it takes so long is that for a certain type of people, research can become very addictive.
