His advice about research and published work applies to software as well. Just because someone published that shiny new framework before you and got lots of upvotes does not mean you should despair. You should look at and learn from it and ask yourself how you could have approached the problem differently. You will learn a lot and chances are pretty high there is still something unique in your work that can be factored out and either published on its own or integrated into the more popular framework.
Being wrong (or failing to arrive at an answer) is, when my ego doesn't get in way, amazing.
Until reading this article, I thought of it mainly as an opportunity to improve my knowledge base (which is why I also enjoy, "I don't know about that, please explain.")
But, the bigger win is when you improve your process of analyzing and solving. I suppose I have stumbled into that approach unconsciously but to make it explicit is far more powerful.
If you're lucky, you have an advisor that teaches you much more than science and research. My advisor taught a lot about how to work in a political environment (yeah, universities are highly political environments).
If you are very lucky, you have an advisor that asks you to teach. By teaching undergraduate student, you learn all the small details of your scientific area. I was lecturing in physical chemistry as a graduate student - and even supervising undergraduates - and after a couple of years I knew all the fundamentals of physical chemistry by heart. That knowledge helped me through writing my thesis.
How should we make it attractive for them [young people] to spend 5,6,7 years in our field, be satisfied, learn about excitement, but finally be qualified to find other possibilities?" -- H. Schopper
Given that cheap and disposable trainees — PhD students and postdocs — fuel the entire scientific research enterprise, it is not surprising that few inside the system seem interested in change. A system complicit in this sort of exploitation is at best indifferent and at worst cruel.
but finally be qualified to find other possibilities
I've always thought of legends like Manuel Blum and the institutions he's advised such as Shafi Goldwasser, Micali, Adleman -- and even their students -- as people who transcend this notion of the PhD as a mantle piece for their résumé. You can tell by the relevance and pointedness of their papers. These people are plugged in. They're solving really interesting problems in really elegant ways. You can tell a lot of love goes into their work.
It's not their job to market the field to people unlike themselves. I see inspiration in Dr. Blum's writing for people who truly want to philosophize like he does. It's your job to love thinking and working on the problems as much as them, and be good like them. When your contribution is cited and useful to others these notions of being "exploited" by "the system" must be secondary. If one feels that way one can go the Larry & Sergey route and change the world that way too.
I don't think there is a finally for any of the names I mentioned here when it comes to computer science.
As a scientist, it is fascinating to watch developed societies slowly but surely chomp down at the very hand that feeds them.
The U.S. is the world leader in scientific research, and what they do, others follow. The absolute gutting of the NIH and other institutes, along with the ruthless exploitation of extremely talented graduate students in a nefarious pyramid scheme, does not bode well for the future of scientific research.
Make no mistake, the scientific method is the only thing that separates homo sapiens from the plethora of majestic flora and fauna on this fragile planet. Only through arduous devotion of driven individuals carried out over hundreds of millennia have we reached our current, almost unfathomable state of knowledge which has conferred upon us unprecedented control over our environment.
A lot of times I can get by with just an intuition that the statement ought to be true and some knowledge of what the proof uses but without a full understanding of the proof. But that's not something to be done every time. Any advice on knowing when I can or can't skip a proof?
Proofs are not just about knowing whether something is true or not. Any good proof will reveal techniques that can be applied in other situations even when the assumptions don't quite hold. If you think those techniques might be applicable to your work then you should study the proof. No one else can make this distinction other than you.
Thanks for that reply but I think you've avoided answering the question. Maybe there's no good answer, but I was looking for heuristics to judge ahead of time how valuable the insight from reading a proof is. After all, reading one can be quite a bit of work so ideally I'd never read proofs that aren't insughtful, but which ones those are is hard to judg in advance for me. Maybe not so hard to judge for someon more experienced than me, though.
Look for clues to what other people thought of the proof. Did the paper appear at a prestigious conference, etc? Your opinion will probably be predicted by other peoples' opinions.
"CLAUDE SHANNON once told me..."
Name dropper. :)
His advice about research and published work applies to software as well. Just because someone published that shiny new framework before you and got lots of upvotes does not mean you should despair. You should look at and learn from it and ask yourself how you could have approached the problem differently. You will learn a lot and chances are pretty high there is still something unique in your work that can be factored out and either published on its own or integrated into the more popular framework.
Great read, even for this grad school drop out.
Being wrong (or failing to arrive at an answer) is, when my ego doesn't get in way, amazing.
Until reading this article, I thought of it mainly as an opportunity to improve my knowledge base (which is why I also enjoy, "I don't know about that, please explain.")
But, the bigger win is when you improve your process of analyzing and solving. I suppose I have stumbled into that approach unconsciously but to make it explicit is far more powerful.
Now, where else can I go apply that...
My best advice to graduate students: enjoy!
If you're lucky, you have an advisor that teaches you much more than science and research. My advisor taught a lot about how to work in a political environment (yeah, universities are highly political environments).
If you are very lucky, you have an advisor that asks you to teach. By teaching undergraduate student, you learn all the small details of your scientific area. I was lecturing in physical chemistry as a graduate student - and even supervising undergraduates - and after a couple of years I knew all the fundamentals of physical chemistry by heart. That knowledge helped me through writing my thesis.
The best advice to beginning graduate students is: Don't do it. Graduate school is the modern incarnation of The Emperor's New Clothes.
I think you're getting downvoted because you didn't substantiate. So, why shouldn't one do it? What do you mean "it's the emperor new clothes"?
http://en.m.wikipedia.org/wiki/The_Emperor's_New_Clothes
I knew what the expression meant, just wondering why he made that claim.
How should we make it attractive for them [young people] to spend 5,6,7 years in our field, be satisfied, learn about excitement, but finally be qualified to find other possibilities?" -- H. Schopper
Given that cheap and disposable trainees — PhD students and postdocs — fuel the entire scientific research enterprise, it is not surprising that few inside the system seem interested in change. A system complicit in this sort of exploitation is at best indifferent and at worst cruel.
https://news.ycombinator.com/item?id=7766377
but finally be qualified to find other possibilities
I've always thought of legends like Manuel Blum and the institutions he's advised such as Shafi Goldwasser, Micali, Adleman -- and even their students -- as people who transcend this notion of the PhD as a mantle piece for their résumé. You can tell by the relevance and pointedness of their papers. These people are plugged in. They're solving really interesting problems in really elegant ways. You can tell a lot of love goes into their work.
It's not their job to market the field to people unlike themselves. I see inspiration in Dr. Blum's writing for people who truly want to philosophize like he does. It's your job to love thinking and working on the problems as much as them, and be good like them. When your contribution is cited and useful to others these notions of being "exploited" by "the system" must be secondary. If one feels that way one can go the Larry & Sergey route and change the world that way too.
I don't think there is a finally for any of the names I mentioned here when it comes to computer science.
As a scientist, it is fascinating to watch developed societies slowly but surely chomp down at the very hand that feeds them.
The U.S. is the world leader in scientific research, and what they do, others follow. The absolute gutting of the NIH and other institutes, along with the ruthless exploitation of extremely talented graduate students in a nefarious pyramid scheme, does not bode well for the future of scientific research.
Make no mistake, the scientific method is the only thing that separates homo sapiens from the plethora of majestic flora and fauna on this fragile planet. Only through arduous devotion of driven individuals carried out over hundreds of millennia have we reached our current, almost unfathomable state of knowledge which has conferred upon us unprecedented control over our environment.
Those who who don't know history...
Does anyone know how to answer this question:
How do I know when a proof is worth learning?
A lot of times I can get by with just an intuition that the statement ought to be true and some knowledge of what the proof uses but without a full understanding of the proof. But that's not something to be done every time. Any advice on knowing when I can or can't skip a proof?
Proofs are not just about knowing whether something is true or not. Any good proof will reveal techniques that can be applied in other situations even when the assumptions don't quite hold. If you think those techniques might be applicable to your work then you should study the proof. No one else can make this distinction other than you.
Thanks for that reply but I think you've avoided answering the question. Maybe there's no good answer, but I was looking for heuristics to judge ahead of time how valuable the insight from reading a proof is. After all, reading one can be quite a bit of work so ideally I'd never read proofs that aren't insughtful, but which ones those are is hard to judg in advance for me. Maybe not so hard to judge for someon more experienced than me, though.
Look for clues to what other people thought of the proof. Did the paper appear at a prestigious conference, etc? Your opinion will probably be predicted by other peoples' opinions.