6 ms·

Thank you OP for your very insightful blog post. You have been better than any other academic at articulating exactly what happens when I (or one of my colleagu

Thank you OP for your very insightful blog post. You have been better than any other academic at articulating exactly what happens when I (or one of my colleagues) interact with the academic world and you made me realize that the academic world has a set of values that might be good but which I should not even attempt to incorporate in my own professional life. I have come to my personal conclusion that there is a difference between 'science' and 'the academic world' in much the same way as there is a difference between 'religion' and 'the church'. One can be very religious without going to the church and doing all the rites that are required by the church. In much the same way, a 'tinkerer' (or 'heathen') can be scientific without being academic. I believe that wiki summarizes it quite well for me: "Science is a systematic enterprise that builds and organizes knowledge in the form of testable explanations and predictions about the universe". There is no statement about giving proper credit (vanity) or not being allowed to reinvent the wheel without first making damn sure that you are indeed reinventing. I want to solve real world problems and the way I want to do that is in a structural and repeatable approach. Science helps me doing this. Of course I'm not living in a vacuum, so I will look up as much as I can about the subject at hand as is reasonable within the available amount of time. Interestingly enough, I almost never end up with a research paper, but almost always with blog posts, books, tutorials et cetera. Products don't sell themselves and neither do research papers apparently. I think there is something seriously wrong if that is not perceived as a crisis in the academic world.

iagovar 9y ago
Research papers are typically about very specific topics. It would be strange to end up reading one on the Internet unless you already know the field keywords and use google scholar or any other science-oriented browser.
- jacobolus 9y ago
  Indeed, google scholar and similar citation graphs are amazing. It makes tracking down vaguely relevant papers much easier than in the past. I can skim a few dozen papers one day, and then more closely inspect the handful that seem most relevant or interesting. I wish that google scholar showed bidirectional links though. I suspect they might have concluded that the list of citations included in a particular paper is part of the content of the paper (not just metadata) so that showing it might infringe copyright.
togelius 9y ago
"a systematic enterprise that builds and organizes knowledge" That is exactly my point. Doing your scholarship is doing your part to organize knowledge. Systematically building knowledge means building on others, and knowing what you build on. So the quote (which you don't attribute to anyone other than "wiki" - a citation here would be useful) really proves my point. If you're doing not doing your scholarship, you are not doing science, or research, you are tinkering. Again, there's nothing wrong with tinkering.
- oldandtired 9y ago
  Interestingly enough, when I have found research papers on the subject that I am looking at, most are so opaque that it was not worth reading in the first place. More helpful information has been presented by others who you would classify as tinkerers. They have been clearer in their explanations and much easier to build upon. In the sense of presenting the research in a manner that other people can digest, academia seems to be more of a ancient guild than an organisation for expanding knowledge.
  - randomsearch 9y ago
    Academic papers are written for an audience of experts. If you wish to understand them, you’ll need to be an expert in the field (you can sometimes get away with a bit less, depending on the topic). They are not intended for general dissemination.
    - kaybe 9y ago
      Yes, if you cannot follow review papers you need to read up some more. (Look for review papers if you're new to the area.)
      - traverseda 9y ago
        Sure, where do I get review papers without having an affiliation with an academic institution?
koonsolo 9y ago
> I almost never end up with a research paper, but almost always with blog posts, books, tutorials et cetera. That is probably related to your field of work. I once had to write a pathfinding algorithm for multiple robots in a tight space. With such problems, you will end up with research papers. I can imagine lots of other technologies (such as everything related to AI) where papers will help you the most. Blog posts etc are a solution to common problems. Sometimes you indeed find some gems on architectural level, or for game development there's lots of valuable content out there. But once you go to complex algorithms, that's the space when research becomes handy.
- metaobject 9y ago
  Similarly, a few years ago I swas developing a path-finding algorithm for time-varying 3D vector fields and I found myself knee-deep in various research papers. Some were better than others, and I did find a few web/blog resources helpful, but it was mainly research papers.
denzil_correa 9y ago
> Products don't sell themselves and neither do research papers apparently. I think there is something seriously wrong if that is not perceived as a crisis in the academic world. It's a case of expectation mismatch from your end. A scientific article is aimed to encapsulate all information which can help push the frontiers of human knowledge. Science is not meant to sell a product.
- cornelis 9y ago
  My point is that science is meant to 'sell' something. Not a product, but the knowledge that has been created. If the knowledge is never applied in the real world its missing out on a lot of potential. If 'tinkerers' are not able to digest the knowledge we do have an issue. I'd like to think that the tinkerers/hackers/programmers/makers in the end are the customers of computer science.
optimuspaul 9y ago
> difference between 'religion' and 'the church' I thought you were going to go the other way on that one... going to church doesn't make you religious. I might even say not all churches have religion. And the analogy still works, 'the academic world' is hardly a subset of 'science'
- cornelis 9y ago
  I don't understand your point. How is 'the academic world' not a subset of 'science'? Is it the same, a superset, or something else?
  - optimuspaul 8y ago
    not all academics is scientific, thought that was an obvious point.
Ultimatt 9y ago
> There is no statement about giving proper credit (vanity) or not being allowed to reinvent the wheel without first making damn sure that you are indeed reinventing. Crediting someone is not about vanity, at all. It's so in the future someone else can see where to go for ideas you've used, what the general thought and progress was that lead to an idea. See "scholarship" in the linked article. Crediting people facilitates scholarship and knowing the context of an idea. In theory you should be able to trace all human thought that lead to an idea, and what else those people were thinking at the time. Amazingly in modern science that's essentially possible. Most papers have an acknowledgements section where you even acknowledge conversations or private letters that lead to an idea. It's not vanity, even in the world of tinkering its worth crediting people with the input for your ideas and the source of your material as it makes it easier for someone else to riff off of your tinkering. It's not about avoiding reinventing the wheel either. It's about knowing all the criteria for wheelness, knowing what has been done before and what is widely accepted as a wheel rather than a roller or a track. Is there a body of related work that might apply in the abstract, if rollers are like wheels perhaps roller maths would be useful, perhaps a development in roller technology has not been applied in principal to wheels and could be a breakthrough. Both of these together are how you actually push things forwards. Giving credit especially is highly misunderstood. If I want to know if one paper Im using is a weird off shoot of thought of the author or not, what else were they working on, where did their inspiration come from, is there more I can get from their direction, or is there somewhere I can go they never dreamed of etc. That transparency is actually amazingly powerful for the creative process.
KKKKkkkk1 9y ago
I'm curious that you're taking issue specifically with taking credit and reinventing the wheel. Avoiding those things is the same as not stealing. I don't think it's academic at all. If you're using someone else's idea, and you just don't want to put some effort into finding the original owner of that idea, don't blame the author for it.
ksk 9y ago
You are choosing to not engage with the author and simply repeating your own opinion. Its obvious to me that your disagreement is actually not even relevant to the article or to what the author is saying. >There is no statement about giving proper credit (vanity) or not being allowed to reinvent the wheel without first making damn sure that you are indeed reinventing. The article isn't a discussion about what science means. Its a discussion about the principles of research that all of our current scientific progress is based upon. One of those principles is the organization and structuring of knowledge. Citations are one of the means of linking knowledge that allows easy navigation between topics and ideas for people both inside and outside the field. Because of the collective effort of people citing other work, it saves time and money of everyone involved since they can easily hop between topics and ideas without having to read a million unorganized pages of scientific data and figure it out every single time. > I think there is something seriously wrong if that is not perceived as a crisis in the academic world. Um, you don't really get to tell someone they're doing it wrong, when what they're doing has worked out brilliantly for the human race.
- cornelis 9y ago
  I might not have been clear enough about the point I was trying to make. OP's comment on the Super Mario Bros neural network was really insightful to me: "Seen as tinkering, that work and video is good work; seen as research, it is atrocious because of the complete lack of scholarship". This made me understand his, and most likely the academic, point of view. What I tried to convey is that I believe that there is value in embedding your knowledge in the network of previous knowledge but the academic world has perhaps gone too far in focusing on this aspect and too little on making sure that the created knowledge is applied. In regard to saying that something is seriously wrong; There is no doubt that the academic world has produced a lot of valuable knowledge during her existence. It's the current academic world that is in crisis. With the H-index as its ultimate false god. The H-index exists only since 2005. Why should I not be allowed to criticize that? If the system is still good, it will survive the criticism.
  - ksk 9y ago
    >but the academic world has perhaps gone too far in focusing on this aspect and too little on making sure that the created knowledge is applied. Okay, now I understand the point. But why is it necessarily the responsibility of a researcher to find industry applications? The industry has to figure out which portions of these ideas can be realistically turned into applications or products based on market criteria. Of course industry can and often does contribute papers to further our knowledge. And when they do invest in doing research they're probably going to be motivated to find applications. > It's the current academic world that is in crisis. With the H-index as its ultimate false god. The H-index exists only since 2005. Why should I not be allowed to criticize that? If the system is still good, it will survive the criticism. I was using 'don't get to' as a rhetorical device. In my view it was like walking up to Usain Bolt and telling him how to fix the problems with his form cause hes doing it all wrong. Sure, you can do that, but most people won't take you seriously. Or in other words, simply being able to identify surface level problems, doesn't mean much without a ton of analysis, explanations, etc, etc. But yeah, obviously you can opine on any subject you like, and people too have the freedom to respond.

iagovar 9y ago

Research papers are typically about very specific topics. It would be strange to end up reading one on the Internet unless you already know the field keywords and use google scholar or any other science-oriented browser.

jacobolus 9y ago
Indeed, google scholar and similar citation graphs are amazing. It makes tracking down vaguely relevant papers much easier than in the past. I can skim a few dozen papers one day, and then more closely inspect the handful that seem most relevant or interesting. I wish that google scholar showed bidirectional links though. I suspect they might have concluded that the list of citations included in a particular paper is part of the content of the paper (not just metadata) so that showing it might infringe copyright.

togelius 9y ago

"a systematic enterprise that builds and organizes knowledge" That is exactly my point. Doing your scholarship is doing your part to organize knowledge. Systematically building knowledge means building on others, and knowing what you build on. So the quote (which you don't attribute to anyone other than "wiki" - a citation here would be useful) really proves my point. If you're doing not doing your scholarship, you are not doing science, or research, you are tinkering. Again, there's nothing wrong with tinkering.

oldandtired 9y ago
Interestingly enough, when I have found research papers on the subject that I am looking at, most are so opaque that it was not worth reading in the first place. More helpful information has been presented by others who you would classify as tinkerers. They have been clearer in their explanations and much easier to build upon. In the sense of presenting the research in a manner that other people can digest, academia seems to be more of a ancient guild than an organisation for expanding knowledge.
- randomsearch 9y ago
  Academic papers are written for an audience of experts. If you wish to understand them, you’ll need to be an expert in the field (you can sometimes get away with a bit less, depending on the topic). They are not intended for general dissemination.
  - kaybe 9y ago
    Yes, if you cannot follow review papers you need to read up some more. (Look for review papers if you're new to the area.)
    - traverseda 9y ago
      Sure, where do I get review papers without having an affiliation with an academic institution?

koonsolo 9y ago

> I almost never end up with a research paper, but almost always with blog posts, books, tutorials et cetera. That is probably related to your field of work. I once had to write a pathfinding algorithm for multiple robots in a tight space. With such problems, you will end up with research papers. I can imagine lots of other technologies (such as everything related to AI) where papers will help you the most. Blog posts etc are a solution to common problems. Sometimes you indeed find some gems on architectural level, or for game development there's lots of valuable content out there. But once you go to complex algorithms, that's the space when research becomes handy.

metaobject 9y ago
Similarly, a few years ago I swas developing a path-finding algorithm for time-varying 3D vector fields and I found myself knee-deep in various research papers. Some were better than others, and I did find a few web/blog resources helpful, but it was mainly research papers.

denzil_correa 9y ago

> Products don't sell themselves and neither do research papers apparently. I think there is something seriously wrong if that is not perceived as a crisis in the academic world. It's a case of expectation mismatch from your end. A scientific article is aimed to encapsulate all information which can help push the frontiers of human knowledge. Science is not meant to sell a product.

cornelis 9y ago
My point is that science is meant to 'sell' something. Not a product, but the knowledge that has been created. If the knowledge is never applied in the real world its missing out on a lot of potential. If 'tinkerers' are not able to digest the knowledge we do have an issue. I'd like to think that the tinkerers/hackers/programmers/makers in the end are the customers of computer science.

optimuspaul 9y ago

> difference between 'religion' and 'the church' I thought you were going to go the other way on that one... going to church doesn't make you religious. I might even say not all churches have religion. And the analogy still works, 'the academic world' is hardly a subset of 'science'

cornelis 9y ago
I don't understand your point. How is 'the academic world' not a subset of 'science'? Is it the same, a superset, or something else?
- optimuspaul 8y ago
  not all academics is scientific, thought that was an obvious point.

Ultimatt 9y ago

> There is no statement about giving proper credit (vanity) or not being allowed to reinvent the wheel without first making damn sure that you are indeed reinventing. Crediting someone is not about vanity, at all. It's so in the future someone else can see where to go for ideas you've used, what the general thought and progress was that lead to an idea. See "scholarship" in the linked article. Crediting people facilitates scholarship and knowing the context of an idea. In theory you should be able to trace all human thought that lead to an idea, and what else those people were thinking at the time. Amazingly in modern science that's essentially possible. Most papers have an acknowledgements section where you even acknowledge conversations or private letters that lead to an idea. It's not vanity, even in the world of tinkering its worth crediting people with the input for your ideas and the source of your material as it makes it easier for someone else to riff off of your tinkering. It's not about avoiding reinventing the wheel either. It's about knowing all the criteria for wheelness, knowing what has been done before and what is widely accepted as a wheel rather than a roller or a track. Is there a body of related work that might apply in the abstract, if rollers are like wheels perhaps roller maths would be useful, perhaps a development in roller technology has not been applied in principal to wheels and could be a breakthrough. Both of these together are how you actually push things forwards. Giving credit especially is highly misunderstood. If I want to know if one paper Im using is a weird off shoot of thought of the author or not, what else were they working on, where did their inspiration come from, is there more I can get from their direction, or is there somewhere I can go they never dreamed of etc. That transparency is actually amazingly powerful for the creative process.

KKKKkkkk1 9y ago

I'm curious that you're taking issue specifically with taking credit and reinventing the wheel. Avoiding those things is the same as not stealing. I don't think it's academic at all. If you're using someone else's idea, and you just don't want to put some effort into finding the original owner of that idea, don't blame the author for it.

ksk 9y ago

You are choosing to not engage with the author and simply repeating your own opinion. Its obvious to me that your disagreement is actually not even relevant to the article or to what the author is saying. >There is no statement about giving proper credit (vanity) or not being allowed to reinvent the wheel without first making damn sure that you are indeed reinventing. The article isn't a discussion about what science means. Its a discussion about the principles of research that all of our current scientific progress is based upon. One of those principles is the organization and structuring of knowledge. Citations are one of the means of linking knowledge that allows easy navigation between topics and ideas for people both inside and outside the field. Because of the collective effort of people citing other work, it saves time and money of everyone involved since they can easily hop between topics and ideas without having to read a million unorganized pages of scientific data and figure it out every single time. > I think there is something seriously wrong if that is not perceived as a crisis in the academic world. Um, you don't really get to tell someone they're doing it wrong, when what they're doing has worked out brilliantly for the human race.

cornelis 9y ago
I might not have been clear enough about the point I was trying to make. OP's comment on the Super Mario Bros neural network was really insightful to me: "Seen as tinkering, that work and video is good work; seen as research, it is atrocious because of the complete lack of scholarship". This made me understand his, and most likely the academic, point of view. What I tried to convey is that I believe that there is value in embedding your knowledge in the network of previous knowledge but the academic world has perhaps gone too far in focusing on this aspect and too little on making sure that the created knowledge is applied. In regard to saying that something is seriously wrong; There is no doubt that the academic world has produced a lot of valuable knowledge during her existence. It's the current academic world that is in crisis. With the H-index as its ultimate false god. The H-index exists only since 2005. Why should I not be allowed to criticize that? If the system is still good, it will survive the criticism.
- ksk 9y ago
  >but the academic world has perhaps gone too far in focusing on this aspect and too little on making sure that the created knowledge is applied. Okay, now I understand the point. But why is it necessarily the responsibility of a researcher to find industry applications? The industry has to figure out which portions of these ideas can be realistically turned into applications or products based on market criteria. Of course industry can and often does contribute papers to further our knowledge. And when they do invest in doing research they're probably going to be motivated to find applications. > It's the current academic world that is in crisis. With the H-index as its ultimate false god. The H-index exists only since 2005. Why should I not be allowed to criticize that? If the system is still good, it will survive the criticism. I was using 'don't get to' as a rhetorical device. In my view it was like walking up to Usain Bolt and telling him how to fix the problems with his form cause hes doing it all wrong. Sure, you can do that, but most people won't take you seriously. Or in other words, simply being able to identify surface level problems, doesn't mean much without a ton of analysis, explanations, etc, etc. But yeah, obviously you can opine on any subject you like, and people too have the freedom to respond.