Bottleneck: At my age and with no formal education in CS and poor academic records, getting an admission is next to impossible by conventional wisdom.
But the challenge is what makes it interesting :-).
PS: If anyone has any data/anecdotes/advice on "impossible" PhD admissions, I'd be glad to hear it (either here or my email is in the profile).
If you are really serious about this, I would say that your best bet is to start making inroads into the field by first getting a masters at a lesser school. There you would be able to demonstrate that you are capable of doing well in grad level courses, which will be a serious question mark on your application given your poor grades. If you are lucky and work hard, you may also be able to publish & network. Especially at the top schools, academia is extremely insidery and knowing people counts for more than you might expect. At least, this has been my experience, and I am trying to do something very similar in a closely-related field (math).
Last, I see it all the time that people get hung up on gaining admission to the best schools. Why Stanford? Getting to rub shoulders with the leaders of the field would, admittedly, be awesome--but you should remember that there are many less selective departments which could afford you a solid and rewarding PhD experience. In my completely un-scientific estimation, 90% of your success or failure grad will be due intrinsic factors--intelligence, work ethic, communication skills, and so forth. The other 10% will come from your surroundings. Don't get so hung up on the latter that you lose sight of the former. Or, more simply: it's possible for you to be a successful PhD student and academic without going to Stanford.
"I see it all the time that people get hung up on gaining admission to the best schools. Why Stanford? "
I am not "hung up" on it in the sense that if I don't get in that is quite all right. I have zero ego attached to any academic credentials. As you rightly point out the odds are that I'll fail and that is OK.
This is an attempted "hack" somewhat akin to climbing Everest via the hardest route possible. If the goal is just to stand on top of Everest there are better ways. There are easier routes up the mountain (like getting the Masters first from a lesser school, then going to Stanford for the PhD etc etc), but the fun (for me, I am a crazy guy) is in doing the hardest thing possible. There are other things I could do like come over on an H1 visa first then attend some classes while working at some MegaCorp, ace those, get to know some professors etc etc, but I don't want to do it that way.
I don't mind if I never get a PhD. I am not doing it to increase my self worth or career opportunities. (I have plenty of both, and the time and effort put into a PhD can probably be used to build a half dozen startups instead if money were the goal). In Computer Science anyone with a laptop, Ubuntu, a network connection and an ACM/IEEE membership can do any kind of research he wants to, though that would be a very barebones approach.
This is just a(n artificially high) bar I am trying to jump over. Jumping over a bar is intrinsically pointless. The fun is in the attempt.
I understand if people think it is a crazy idea. I just put it down here because the OP asked for "interesting" goals. Crazy ideas are often interesting!
"At the risk of discouraging you, have a look at the applicant profiles people post up on some of the grad school forum sites. "
I am not discouraged at all. I personally know very "average" people who have PhDs from MIT/Stanford. The idea that every MIT/Stanford grad is some kind of technical/research superman is a myth. There is a bell curve there, and I am confident I am not on the lower end of that curve. I don't have the right "background" to get in but I've never let that stop me from trying anything so why start now?
So sure, by conventional thinking I have next to no chance. On the other hand, a scientist I work with (who supervises several PhD's at IISc and IITs) said to me "I've never seen anyone level up as fast as you do. You understand [his specialization in Machine Learning] more deeply and have more ideas than any of my students. If you ever want to do a PhD and want a recommendation, ask me".
I plan to. ;-)
Yes I would. There are some ultra awesome professors there I'd love to work with. And the atmsphere and facilities at Stanford would be much better (than my present situation in Bangalore) to do research. My focus wouold be to work with and learn from them, not to become a fulltime academic.
If I get my PhD I'd like to be the author of a few awesome research papers and have no intention of getting into the tenure track rat race and so on.
"It sounds like merely getting in would satisfy all of your goals."
My immediate goal is to get and admission. Why should meeting that goal satisfy all my goals?
Once you jump over one bar you reset it higher and try again.
If I were to get admitted, I'd set a goal of doing some stunning research, ace the classes and so on (which would be quite a challenge given the calibre of my fellow students).
I've worked against the odds all my life (people were always telling me I couldn't do this or that) so this is very encouraging!
Thank You :-).
http://3dpancakes.typepad.com/ernie/2005/03/re_phd_with_low....
My advice is think about what you really want to do, and why you want to do it. If you want to do good research then it makes sense to apply to highly ranked PhD programs because they will give you a support system to help you learn. If you just want to own a PhD from a specific school, then it will be harder to get in, harder to finish, and less meaningful when you're done.
My advice for your specific goal is to somehow get involved with a professor's research. Preferably someone who will be recognized by admission committees. Their letter will help a lot. This might involve getting an MS at the highest-ranked school you can get admitted to. Sadly, those programs are not usually funded (as opposed to American PhD programs).
Also since my undergrad was in a different field and I had to take pre req's I couldn't get direct funding for CS degree.
I transferred to Cornell after a year at less reputed university which gave me a bit of scholarship, I could have transferred to RPI where i had good chance of funding but I chose Cornell.
it would be easier if you had research experience in other fields and slaughtered the GRE general & subject...
but yeah stanford is not exactly hurting for applicants
Working on it (and on getting research exp) :-)
"slaughtered the GRE general & subject"
I have an almost perfect score on the GRE general (800/800 Verbal 790/800 Quant, 6/6 writing) . Taking the subject exams later this year. Not too worried about that either , but I heard these aren't of much value and research /recommendations are all that count. So I am working on those now.
Ah well you work with what you have :). Along with many brilliant dedicated people, I also see people with rather mediocre research and recommendations (but perfect academic scores etc) get in (and then not do anything particularly brilliant/drop out etc), so I suspect there might be a (small) chance. If every one who got in thoroughly outclassed me I would have no hope.
We'll see. :-)
I'd like to follow that path, but for now I'll be chasing $$$
Good Luck!
If I ever make millions I am going to establish a scholarship for "wildcard" admissions at the best schools. You can only use the scholarship to fund "imperfect" candidates!