Folks, One thing you'll notice is that graduate students, and therefore faculty members, are very good at offering their opinion as if it were gold-plated fact. (Of course, I'm as guilty as anyone else.) After all, what's a thesis defense except the frantic attempt to convince four other people that you know what you are talking about? You may even find that most students will object violently to one or more of the pieces of advice herein. So, as you read through these, remember, these are personal opinions: they may or may not apply to you, and only you can decide. Good luck, and welcome! -brian noble 8/28/96 Here's the solicitation for advice: ==================================================================== I'm madly trying to put together the survival talk for this year. One of the things I thought might be interesting to do is to have a dose of "Things I wish I'd heard/listened to when I first sat in 5409." To that end, I'd like to collect bits of advice from all of you to relay on to the first years. I'll highlight any that seem particularly profound/funny in the talk, and I'll compile the responses into something to give to them all. This is going to be a written record. As such, I'll strip off all mail headers, but leave any signatures in the body of the mail. If you want your advice to be anonymous, don't include a signature. I reserve the right to exclude particular pieces of advice, but I promise not to reword anything unless I talk to you first. I anticipate excluding things only if they name/accuse particular individuals, or are obviously unhelpful. Examples of these are: "Prof. Bovik would steal research ideas from his own grandmother if he could figure out a way to channel her spirit. Avoid him at all costs." It's fine to point out to students that faculty don't have any more of a corner on the ethics market than anyone else. (In fact, I plan on it.) But, to prejudice the entire group of students against Harry before the've met him and formed their own opinions isn't really fair to either Harry or the students. Even if Harry *is* a certifiable jerk. Some people have reported working well with Harry, though I can't say I've met one personally. "Leave now, while you still can." I'd prefer to give them tools to survive and graduate, rather than give them reasons why the single grad school decision they've made so far, to come here, was stupid. "Make sure you start scheduling your thesis defense at least three months before you'd like to have it happen" This is probably pretty darn good advice, but it's not all that clear to me that, unless the first year in question is *truly* exceptional, they really need to hear this on the third day of the IC. (If any of you who are close to finishing/finished want to give a mid-life crisis talk, that would be a good place to put this.) Thanks for the help! -brian ====================================================== Now, on to the responses! ====================================================== > "Leave now, while you still can." > > I'd prefer to give them tools to survive and graduate, rather than > give them reasons why the single grad school decision they've made so > far, to come here, was stupid. Huh? You make it sound like they were stupid. (If there's one thing I've learned here at IBM, it's appreciation of the benefits of CMU.) My advice: Go to talks. Lots of talks. Blow your mind. Just do it. Develop research sense, meet people, learn, stimulate your thinking. Rob ====================================================== One point which you probably already have: The student-advisor relationship is a relationship between TWO people. Although there ARE advisors that are generally more helpful than others, and although there are advisors that will be good for large groups of people, finding an advisor that is going to work for YOU is important. There are some advisors that would NOT work for me that might work very well for someone else and vice-versa. An important part of choosing an advisor is to know yourself and what YOU need in an advisor. Choosing Harry Bovik as an advisor because others wax lyrical about him might not always be the best option. \x/ill :-} William Uther "He was a wise man who invented God." will@cs.cmu.edu - Plato (427?-348? B.C.) Dept. of Computer Science, Carnegie Mellon University http://www.cs.cmu.edu/~will/ ====================================================== Remember that the topic that you choose to work on will eventually play a part in the type of jobs you can look at when you graduate. While it's nice to be idealistic and work on something with little real-world applications, keep in mind that it may to be difficult to find non-academic positions once you graduate. -- Jennie ====================================================== 1. young faculty ==> energy / burnout? Young faculty often have more energy than senior faculty, and if you are lucky, you can "ride the wave", getting on lots of publications and doing lots of cool research. Consider though that maintaining high energy levels is really hard, and the faculty member might suffer burnout. Lots of such faculty have left CMU, leaving their students in a void. 2. old faculty ==> political experience / no time? Senior faculty have lots of political experience, but often have so many students and other obligations that they really don't have time for you. Typically their students MUST be more self-motivated than most. As a related point, count the total number of people your PA advises (grad students, undergrad students, staff, postdocs, etc), since they rarely count anyone outside grad students: but everyone else takes time too! 3. personality is 90% factor in whether you graduate Each year, at least one student picks their advisor based solely on research. No matter how many times you are told, someone doesn't listen. It bears repeating: if you can't get along with your advisor, you won't graduate. Comprise research, not personality. (Note that in most cases, your thesis won't be on the research you envisioned anyway.) 4. TAing always takes more time than you expect Try TAing both classes in one semester: that way only one semester is totally ruined. 5. make your advisor happy Some students have been booted because they hadn't proposed by the end of their fourth year. Others hadn't even seen the word "proposal" in their Black Friday letters. It all depends on what your advisor wants. And be sure you understand your advisor: students have been booted because of mis-communications about what the advisor wants. ====================================================== "Get to know the students and staff who keep the world running. Personal contact with such folks is very useful." "Zephyr *,* is evil but lots of fun and a good way to know what is going on in {world,industry,department}" ====================================================== "Pick an advisor that lets you do what you want to do." "Systems is where the funding is." "The real world cares about Intel machines running variants of Windows. Make sure your research matters to the real world." "Never TA 15-412 twice. Trust me on this." "Use Zephyr and the Zephyr Archives. It's better than the Oracle at Delphi" "Do not become addicted to Zephyr/the Net/News/etc." "Facilities is both very competent, and very overworked. Paranoia is a natural and healthy response to the situation." "Remember - It's supposed to be fun." Peter August Dinda pdinda@cs.cmu.edu Doctoral Student, School of Computer Science, Carnegie Mellon University http://www.cs.cmu.edu/afs/cs.cmu.edu/usr/pdinda/html/pdinda.html ====================================================== [[for more material, I think there was a list of advice in the 25th anniversary book (see Cathy Copetas).]] If you complete your third year without going to a refereed conference, something is wrong. If you don't go to a conference every two years thereafter, something is probably wrong. I feel strongly about this one. If, after two years here, you haven't refereed a paper, you should (you should read Alan Jay Smith's "The Task of the Referee"). This is both a duty incumbent on members of a scientific community and a really good opportunity to learn how to write before you try it yourself. If, after four years here, you haven't published a paper in a refereed journal or major refereed conference in your field, something might be wrong. Giving talks at seminars is very valuable. If you do it four times per year, you will absolutely get better at public speaking, even if you start out being afraid of and horrible at it (as most of us did). If you don't have something to talk about, somebody will gladly assign you an interesting paper to present. If you're afraid about something, or sure you can't do something, there are 10 people in the program who have faced and overcome that thing, and your officemate knows who those people are. Why not get expert help? You should probably start out by having a regularly scheduled meeting with your advisor once every week or two (even if only for five minutes). When you are just starting out, you may not feel like you have much "progress" to report, which is natural, so you can ask questions of your advisor instead: * What is our group's long-term plan? * Which are the best journals in our area? * What's the most interesting paper you've heard about recently? * What are the major open research areas in our area? If you find that regular meetings, or frequent meetings, don't work for you, your advisor will almost certainly be willing to space them out or give them up, but you should give them a try. ====================================================== My advice to new grads: `Find out what your advisor's outside commitments are. Find out whether they have a start-up somewhere. Find out whether they signed the Inventions Agreement or not.' You can strip my name from this, but it doesn't really matter: I'm sure people will guess. :-) Geoff. ====================================================== Advice to new students: 1. Don't lose sight of the big picture. Grad school is a means to an end. Having a PhD doesn't make you a better person, as such, not getting a PhD doesn't make you terrible, stupid, a loser, or anything else. Whatever you do, don't forget that the journey (in this case certainly) is probably as much or more important than the end. 2. Have a long term view of things. Think about who and what you want to be in 20 years, then think about how to get there. Sacrificing everything to do extremely well on a course may not be the optimal strategy. In particular, start thinking about what kind of job you'd like to get and what sorts of things you need to do to get it several *years* before you graduate. 3. Don't over focus on any one thing. One great thing about CMU is that we do research in the big, that means big groups with lots of students. It's easy to forget about other projects in your area, other areas of CS, and other areas related to what we do. Another hallmark of CMU is that it is world-class in so many areas relating to CS. Sample the plate. Go back for seconds. Peruse. Experience. You'll never get another chance. ====================================================== A random piece of advice for the first-years... [If non-academic advice is appropriate for the session you're running] Take up cross-country skiing. It makes Pittsburgh winters not just survivable but downright enjoyable. I wish I'd discovered it before my fourth year. -jab ====================================================== --Piece of advice #1-- I had a poor advisor breakup before switching advisors. My experience with this person was not universal enough to warn against them in any broad way; they have had successful students in the past. But, I do have some advice specific to this person on how to avoid this sort of breakup, and I'd be happy to share that advice with anyone who is considering my former advisor. The moral, which you will hear again and again: Talk to your potential advisor's students and former students before you choose. --Piece of advice #2-- As a grad student, you will have a computer on your desk---this is the most dangerous non-narcotic time sink ever invented. You know this already, but I will remind you again: Be aware of the following two traps: Reading zephyr all the time: It's hard to physically hang around with your friends *all* the time because any small set of them will eventually have to go away and work or eat or sleep or something. But on zephyr, you can rotate through groups of people without that natural "time to go do something else now" cue. This removes a built-in mechanism that limits your time spent hanging out. Be cautious. Hacking on fun non-research: There's lots of fun hacking to be done that will make your life or your colleague's lives happier. Just remember that, unless you've been really clever at picking a research topic, random hacks don't count towards your thesis. This doesn't mean you shouldn't do them, it just means you shouldn't do them all the time.