Blockchain in name only - why you should not use blockchain in scholarly publishing

Transcript

1. Blockchain
   In
   Name
   Only
   How to think about blockchain in STM publishing
   @IanMulvany - Head of Transformation - SAGE Publishing
   

   View Slide

2. While discussing interoperable metadata and
   interchangeable standards is GOOD,
   blockchain is inappropriate for our industry
   because it is too complex, requires lack of
   trust to be valuable and does not offer as much
   potential ROI as other initiatives.
   

   View Slide

3. Why are people
   talking about this?
   

   View Slide

4. I’m going to propose that a possible solution to the
   growing problem of authentication and accounting in
   scholarly publishing may be found in the technology
   behind Bitcoin
   the blockchain.
   Phil Davis writing on The Scholarly Kitchen 2016
   Third-party companies have made huge profits from the
   work of scientists for many years, selling journals and
   reprints to a variety of institutions. In blockchain-
   powered scientific publishing, researchers would
   control whether or not institutions are required to pay to
   use their work, while making it freely available to the
   public and researchers alike, if they so wish.
   Manuel Martin is CEO and co-founder of Orvium.
   

   View Slide

5. I’m going to propose that a possible solution to the
   growing problem of authentication and accounting in
   scholarly publishing may be found in the technology
   behind Bitcoin
   the blockchain.
   Phil Davis writing on The Scholarly Kitchen 2016
   Third-party companies have made huge profits from the
   work of scientists for many years, selling journals and
   reprints to a variety of institutions. In blockchain-
   powered scientific publishing, researchers would
   control whether or not institutions are required to pay to
   use their work, while making it freely available to the
   public and researchers alike, if they so wish.
   Manuel Martin is CEO and co-founder of Orvium.
   KILL SCI-HUB !!!
   KILL Publishers !!!
   

   View Slide

6. Blockchain - X minute primer where X < 10
   

   View Slide

7. Blockchain
   Metadata
   Data
   

   View Slide

8. Blockchain
   Metadata
   Data
   Previous
   Hash
   Metadata
   Data
   Previous
   Hash
   Metadata
   Data
   

   View Slide

9. Blockchain
   Previous
   Hash
   Metadata
   Data
   Previous
   Hash
   Metadata
   Data
   Previous
   Hash
   Metadata
   Data
   Previous
   Hash
   Metadata
   Data
   

   View Slide

10. Blockchain
    

    View Slide

11. View Slide

12. Blockchain - tamper proof
    

    View Slide

13. View Slide

14. View Slide

15. 51% rule
    

    View Slide

16. 51% rule
    

    View Slide

17. To prevent a coordinated attack transactions are
    designed to be slow - “proof of work”
    

    View Slide

18. We need:
    Previous
    Hash
    Metadata
    Distributed writers
    Lots of copies
    1 2
    A network with no
    trusted intermediary,
    and shared copies of
    the database
    4
    3
    Proof of work scheme
    to prevent 51% hack
    Incentive to
    participate
    ⏳
    

    View Slide

19. In STM we don’t have these things:
    1 2 4
    3
    Distributed writers
    Lots of copies
    A network with no
    trusted intermediary,
    and shared copies of
    the database
    Proof of work scheme
    to prevent 51% hack
    Incentive to
    participate
    Hard to get to scale
    across all publishers
    We have a high trust
    environment, lots of
    trusted intermediaries
    Anonymity is eventually
    pointless in STM -> low
    incentives for attack
    Existing incentive
    schemes too
    entrenched to be
    supplanted
    

    View Slide

20. Finally - idea is simple, but implementation is hard - case study - Bitcoin
    95% of transactions on the bitcoin network may be artificial
    Not secure
    Not egalitarian
    Not efficiently distributed Over 80% of mining is preformed by six mining pools
    $2.7M stolen from exchanges per day in 2018
    

    View Slide

21. Risks
    It will get attacked and hacked
    Vendor lock in
    Lack of technical capacity or attention within our organisations
    Who owns the transaction? Who owns a peer review?
    Reward systems in academic publishing are hard to shift
    

    View Slide

22. Opportunities
    Let’s us all talk about metadata
    Submissions as transactional events
    A global append only store
    Multiple copies, by design
    Transactional history of articles is interesting
    

    View Slide

23. Alternatives
    AWS S3
    DAT
    Kafka
    Solid Web Apps
    

    View Slide

24. Alternatives for your attention
    Metadata
    Standards
    Distributed
    Systems
    Open
    Publishing
    Systems
    

    View Slide

25. View Slide