Ask HN: Any one else think it's a strange time to be a computer scientist?
10 comments
You might want to read about fully homomorphic encryption, the PCP theorem and many other mind-blowing results to have come out of computational complexity theory community in the past 15 years. These results use powerful mathematical tools like Fourier analysis on the boolean cube. Here are a few links to get you started:
http://www.scottaaronson.com/papers/philos.pdf
http://www.ccs.neu.edu/home/jmitesh/Mitesh_Jain_files/draft....
http://theoryofcomputing.org/articles/gs001/gs001.pdf
http://www.scottaaronson.com/papers/philos.pdf
http://www.ccs.neu.edu/home/jmitesh/Mitesh_Jain_files/draft....
http://theoryofcomputing.org/articles/gs001/gs001.pdf
DISCLAIMER: I'm not a computer scientist
I'm a self-taught programmer which means that I don't feel qualified to answer the cs questions, nevertheless, I think I can answer your question.
First, thinking that "the new stuff" is the same as the exciting stuff is wrong. The exciting stuff is the stuff that excites you and not necessarily the new stuff. For example, the principles behind the modern web apps aren't exactly new (client/server, CRUD), and yet, you had a love affair with rails :)
Second, there are two kinds of "the new stuff". The first kind is the stuff that's a new development in cs, the second kind is the stuff that's new to you. If you focus on the second kind you'll see that there's a whole new world out there waiting to be explored and conquered. Scratch that, there are worlds upon worlds of potentially exciting stuff waiting for your attention.
My advice is to keep looking.
I'm a self-taught programmer which means that I don't feel qualified to answer the cs questions, nevertheless, I think I can answer your question.
First, thinking that "the new stuff" is the same as the exciting stuff is wrong. The exciting stuff is the stuff that excites you and not necessarily the new stuff. For example, the principles behind the modern web apps aren't exactly new (client/server, CRUD), and yet, you had a love affair with rails :)
Second, there are two kinds of "the new stuff". The first kind is the stuff that's a new development in cs, the second kind is the stuff that's new to you. If you focus on the second kind you'll see that there's a whole new world out there waiting to be explored and conquered. Scratch that, there are worlds upon worlds of potentially exciting stuff waiting for your attention.
My advice is to keep looking.
I definitely felt a lot ennui about the state of things when I was about 4-5 years out of college (although to be fair I jumped out to join the dotcom boom).
I'm actually pretty stoked about the state of things right now though. Ubiquitous compute power through virtualization, big data (map reduce!), open cv, machine learning, SAT solvers, you name it. Now that we have cheap, plentiful and fast computational power all that stuff they dreamed about back in the 60s and 70s is now a reality.
CS as a discipline is nowhere near tapped out. Sure, we're probably not going to find too many more core algorithms in many of the various disciplines, but come on.. we don't even have true general purpose AI at this point!
I'm actually pretty stoked about the state of things right now though. Ubiquitous compute power through virtualization, big data (map reduce!), open cv, machine learning, SAT solvers, you name it. Now that we have cheap, plentiful and fast computational power all that stuff they dreamed about back in the 60s and 70s is now a reality.
CS as a discipline is nowhere near tapped out. Sure, we're probably not going to find too many more core algorithms in many of the various disciplines, but come on.. we don't even have true general purpose AI at this point!
"Some really exciting stuff I have come aware of: Timesharing, the internet, Smalltalk, Lisp, designing hardware for software, different methods of programming, tools to enhance our understanding of science."
On the timescale of human endeavor, those ideas, like the idea of computer science itself, are all new. But even though old on a relative scale, there is nothing wrong with pursuing an interest in them - lambda calculus, information theory, etc. don't rust as they age. Indeed, what makes many of those computer science ideas interesting is that they are still applicable and relevant today.
On the timescale of human endeavor, those ideas, like the idea of computer science itself, are all new. But even though old on a relative scale, there is nothing wrong with pursuing an interest in them - lambda calculus, information theory, etc. don't rust as they age. Indeed, what makes many of those computer science ideas interesting is that they are still applicable and relevant today.
Rails is Cobol. There is a lot of excitement in, for instance, AI, geoscale graphics or biocomputation.
- information theoretic optimal succinct/compact/implicit data structures
- cache oblivious algorithms, especially CO-B-Trees
- linear work suffix array construction
- new lower bounds on external store and search
- sure there must be plenty of results on distributed systems techniques too
Plenty of amazing results in the last few years.
Plenty of amazing results in the last few years.
If you had graduated in the 60s or 70s (or even 80s) chances are that you wouldn't have worked in Computer Science at all. Today you can choose from a vast field of possibilities. Keep looking and go to the edge of one area
Are you looking for intellectual stimulation? Or something else?
If the former, read Bram Cohen's paper on BitTorrent. A lot of academics were working on the peer to peer buzzword right after Napster came out. Bram's work shook that little corner of academia IMHO. Academics regained their footing with the horde of DHT papers that followed ala Chord, Pastry, etc. I think Cassandra uses these ideas.
If the former, read Bram Cohen's paper on BitTorrent. A lot of academics were working on the peer to peer buzzword right after Napster came out. Bram's work shook that little corner of academia IMHO. Academics regained their footing with the horde of DHT papers that followed ala Chord, Pastry, etc. I think Cassandra uses these ideas.
If you look at what we could have got versus what we actually got, two things will happen:
1. You'll drive yourself mad.
2. You'll start to appreciate the Right Thing.
1. You'll drive yourself mad.
2. You'll start to appreciate the Right Thing.
Pick your favorite topic and think about all the things computers can do for it that they don't right now. That's pretty exciting.
At first enthused about browser/server web applications (using rails framework) -- I find completely dull, and now feel to constrained by the browser and the server interaction. Working with squeak I find I just want my computer too be squeak and build off that.
(caveat: graduated 4 years ago)