4 ms·

I have so much that I want to say about this - but beyond all the politics and moneyed/tenured interests holding back change, I can see that Jupyter notebooks a

I have so much that I want to say about this - but beyond all the politics and moneyed/tenured interests holding back change, I can see that Jupyter notebooks are the disruptive innovation that will force publishing to modernize. People I work with are producing notebooks that fuse text, formatted equations, code, and interactive figures in a way that looks and feels like a top-quality paper. Except with this paper, you can reveal the code that produced each figure, probe and alter the analysis methods directly, and download the raw data behind the figures (it's an attribute of the plot object! Imagine that!). This format will win. It's too useful and solves too many problems not to win. The only question is whether academicians have the foresight to build in a decentralized DARE I SAY blockchain/smart-contract-based peer review system to circumvent the rent seeking publishers.

edanm 8y ago
Err do you mind explaining what the blockchain has to do with this?
- majewsky 8y ago
  This is HN, so you have to include the word "blockchain" once every hundred words. /s
- laichzeit0 8y ago
  This is actually not a bad application, come to think of it.
- vishnu_ks 8y ago
  This is the perfect application for Blochcian * Nobody can manipulate the reviews as everything is stored in Blockhcian. * Nobody can start asking for money one day suddenly as the whole thing runs on a smart contract and once a smart contract is deployed there is no way to change the code. * The articles are not stored or controlled by a central server.
  - edanm 8y ago
    Thanks for the reply! I'm not sure I fully understand (or buy into) Blockchain, so forgive me for probing further on this, but: I'm still not sure how blockchain is helping here? Can't you get the same results by just publicly publishing the articles with an open license? That's kind of how things work right now in open-access journals (or with something like StackOverflow, for that matter) and it works fine, what is blockchain adding here? There are tons of copies of StackOverflow's data running around in mirrors, they are (or it's easy to do) hashed for a crypto signature, so you can't just change their text on people. No one can ask money for the answers because they are licenses CC. So what is Blockchain adding to this? As I understand it, what blockchain provides is most importantly an incentive mechanism, to make people actually donate the PC time to store things (iiuc), but it seems to work fine in popular cases without having the incentive mechanism (e.g. StackOverflow), and I'm guessing if it isn't popular, then just cause it's on a blockchain doesn't change anything and no one will bother storing it? But again, I'm not sure I understand what you're talking about, so I'd be more than happy to be corrected.
    - vishnu_ks 8y ago
      In the current industry the authors are required to publish papers in the most reputed journal rights for maximum rewards. The more reputed a platform becomes the more power does it have. So it's totally upto the people who runs the journal to change the rules one day if they decided to make more money. They can easily change the license of the new papers that are waiting for review and the authors would still have to publish in the journal because of it's reputation. They can also theoratically censor the articles or authors that they don't like. On the other hand this can't be done for an application for example running in Ethereum. Once a smart contracts is deployed there is no way to modify it. The other argument is modifying the text. I don't think dumps of data running around like that of StackOverFlow doesn't help much. Many sites have got heat for modifying the user comments or reviews in the past. The average person who is using a site is not going to go through the reviews and the dumps of data and see whether anything has been modified or not. StackOverflow makes money by storing the questions and answers. Thats the incentive mechanism. Also none of the users gets a portion of the money that they helped them to make. I am not sure whether this is the greatest example. One can easily write a smart contract that will facilitate the transfer of money from writers to reviewers without the need of any central server. The amount of money the reviewer gets would be a function of their reputation. So it would be in the best interest of the reviewer to produce high quality work and high quality reviews. A far as the storage is concerned one can make use of Ethereum or a general purpose blockchain for building this. You would have to pay for the storage(reviews, ratings etc) only once and the nodes in the network would make sure that its their for ever. The research papers can be stored in IPFS due to their large size. The files can be seeded by anyone. If the research paper is popular it would be always available. If it is not popular the author or the university can still seed it as its in their best interest. I don't think it matters much where the PDF file is stored. It can be in a central server run by University as well. So this is not a big issue. Does that makes any sense? Feel free to correct me as well :)

edanm 8y ago

Err do you mind explaining what the blockchain has to do with this?

majewsky 8y ago
This is HN, so you have to include the word "blockchain" once every hundred words. /s
laichzeit0 8y ago
This is actually not a bad application, come to think of it.
vishnu_ks 8y ago
This is the perfect application for Blochcian * Nobody can manipulate the reviews as everything is stored in Blockhcian. * Nobody can start asking for money one day suddenly as the whole thing runs on a smart contract and once a smart contract is deployed there is no way to change the code. * The articles are not stored or controlled by a central server.
- edanm 8y ago
  Thanks for the reply! I'm not sure I fully understand (or buy into) Blockchain, so forgive me for probing further on this, but: I'm still not sure how blockchain is helping here? Can't you get the same results by just publicly publishing the articles with an open license? That's kind of how things work right now in open-access journals (or with something like StackOverflow, for that matter) and it works fine, what is blockchain adding here? There are tons of copies of StackOverflow's data running around in mirrors, they are (or it's easy to do) hashed for a crypto signature, so you can't just change their text on people. No one can ask money for the answers because they are licenses CC. So what is Blockchain adding to this? As I understand it, what blockchain provides is most importantly an incentive mechanism, to make people actually donate the PC time to store things (iiuc), but it seems to work fine in popular cases without having the incentive mechanism (e.g. StackOverflow), and I'm guessing if it isn't popular, then just cause it's on a blockchain doesn't change anything and no one will bother storing it? But again, I'm not sure I understand what you're talking about, so I'd be more than happy to be corrected.
  - vishnu_ks 8y ago
    In the current industry the authors are required to publish papers in the most reputed journal rights for maximum rewards. The more reputed a platform becomes the more power does it have. So it's totally upto the people who runs the journal to change the rules one day if they decided to make more money. They can easily change the license of the new papers that are waiting for review and the authors would still have to publish in the journal because of it's reputation. They can also theoratically censor the articles or authors that they don't like. On the other hand this can't be done for an application for example running in Ethereum. Once a smart contracts is deployed there is no way to modify it. The other argument is modifying the text. I don't think dumps of data running around like that of StackOverFlow doesn't help much. Many sites have got heat for modifying the user comments or reviews in the past. The average person who is using a site is not going to go through the reviews and the dumps of data and see whether anything has been modified or not. StackOverflow makes money by storing the questions and answers. Thats the incentive mechanism. Also none of the users gets a portion of the money that they helped them to make. I am not sure whether this is the greatest example. One can easily write a smart contract that will facilitate the transfer of money from writers to reviewers without the need of any central server. The amount of money the reviewer gets would be a function of their reputation. So it would be in the best interest of the reviewer to produce high quality work and high quality reviews. A far as the storage is concerned one can make use of Ethereum or a general purpose blockchain for building this. You would have to pay for the storage(reviews, ratings etc) only once and the nodes in the network would make sure that its their for ever. The research papers can be stored in IPFS due to their large size. The files can be seeded by anyone. If the research paper is popular it would be always available. If it is not popular the author or the university can still seed it as its in their best interest. I don't think it matters much where the PDF file is stored. It can be in a central server run by University as well. So this is not a big issue. Does that makes any sense? Feel free to correct me as well :)