V Please find document status at the bottom of this document

Please note, that this document is not only a review, but some concepts regarding Blockchain For Science have been shared here for the first time, so it is Original Work. Please give appropriate reference, it has some weaknesses, but is the most compelling document on what blockchain has to offer for Science and Knowledge Creation for the time being. Older versions can be found here: 1 (earliest thoughts:)), 2,3 and 4.

A completely redesigned version is currently being created.

Blockchain for Science[a] and        Knowledge Creation[b][c][d][e][f][g][h][i][j][k]


PD Dr. med. Sönke Bartling (corresponding author, initiator and maintainer, @soenkeba, soenkebartling@blockchainforscience.com)

Founder of Blockchain For Science &

Associate researcher at the Humboldt Institute for Internet and Society

Many more contributors - please see document history.

Isn´t “really good science not always a break with orthodoxy – and how could the orthodox then fairly assess it?”

 (Michael Polanyi - potentially wrongly attributed or cited by the maintainer)





What is Blockchain?

Blockchain in Science and Knowledge Creation

Current projects and Implementations



Maintained from May 2016-March 2018


Blockchain is a computer protocol involving cryptography, a new way to look at databases and a socio-cultural-legal-political-economic (r)evolution and knowledge creation will be affected by it.

Blockchain has the capacity to make digital goods[l] immutable, transparent, externally provable, decentralized, valuable, and distributed (and potentially permanent). Besides the initial experiment and data acquisition, all remaining parts of the research cycle could take place within a blockchain system. Attribution, data, subject anonymity, data post processing (e.g. via smart contracts) & archiving, publication, research evaluation, incentivisation, and research fund distribution would thereby become time-stamped, comprehensible, open (at will) and provable to the external world. Currently, scientists must be trusted to provide a true and useful representation of their research results in their final publication; blockchain would make much larger parts of the research cycle open to scientific self-correction[m]. This bears the potential to be a new approach to the current reproducibility crisis in science[n][o][p][q], and could ‘reduce waste and make more research results true’.[r][s] Beyond that, blockchain could be used to reduce overhead and accelerate the scientific process and incentivise true innovation. [t]



Currently, blockchain is at the peak of its hype cycle (goo.gl/6DbyPC). Many claim that the blockchain revolution will affect not only our online lives, but will profoundly change many more aspects of our society [1–4]. Some foresee these changes as potentially being more far-reaching than those brought by the internet in the last two decades. If this holds true, it is certain that research and knowledge creation will also be affected by this. So, why is that the case, and what is this all about? More importantly, could knowledge creation benefit from it? Adoption of new technologies is good, however, it should not be an end in itself - there should be problems that can be solved with it. Currently, there is a credibility and reproducibility crisis in science [5–15].

In this article, we will first provide some abstractions and technical points of blockchain, then discuss application examples, and finally, identify problems in the research world that might be solved by means of blockchain.[v][w][x]

Blockchain - the data structure

In a literal sense, blockchain is a computer data structure, a list of data blocks that are linked through a cryptographic function. The earliest description of this data structure dates back to 1991 [16]. If one changes the content of one block, all following blocks need to change as well.

Blockchain became widely known as the data structure (= ledger) that underpins Bitcoin [17,18]. Bitcoin is an online payment processing tool that lacks centrality and trusted third parties such as banks or companies (like Paypal). It is distributed, the blockchain ledger is stored on many computers, and there is no single point of failure. In Bitcoin, long known concepts have been successfully implemented together and found wide use for the first time, as they are:

  • Cryptographic tools such as public key cryptography and hashes
  • Consensus mechanisms (=ways to settle discrepancies within same data sets that are stored on different computers) [19,20] 
  • Proof-of-work (=methods that uses laborious computer calculations to prevent a system from being flooded with ‘spam’ or fake identities)[y] [21] 
  • Economic incentives (miners are paid with Bitcoins) to agree upon the correct state of the blockchain ledger

Bitcoin continues to function reliably, despite several billion dollars worth of value now within its network. Breaking Bitcoin could potentially make large portions of this money accessible to the attacker.  

Blockchain - the (r)evolution 

Payment processing is just one application of blockchain systems. To differentiate the characteristics of the upcoming online (r)evolution from the payment processing tool and implementation of Bitcoin itself, the term ‘blockchain’ is nowadays used in a much wider context. It describes a system for organizing all kinds of digital things, be it files, databases, or assets, in ways that were first widely perceived in Bitcoin. Attributes of this system include: 

  • Decentralized
  • Distributed
  • Immutable (≈’append only database’)
  • Transparent (provable to the external world)

Before we explain in more detail what this means, let us first take a look at how we use computer services today:

Nowadays, it is clear that whoever provides online services, be it a cloud storage service, a bank, an email provider, or a scientific publisher, needs to be trusted to do what they are supposed to do. We know that the provider could technically alter our accounts, change scientific results, or indeed our emails and files at will. We rely on those trusted third parties not to do so (Figure 1). Furthermore, we know that once data is digitized, it can be arbitrarily changed at will without leaving a trace (e.g. by researchers). 


Figure 1: Today the owner (or researcher, academic publisher, data-repository etc.) has full control over their computer, data, and services they run (e.g. a database)[z][aa][ab][ac] and could technically alter the content in arbitrary ways. After the blockchain revolution, this is no longer the case, as decentralized trust providing systems provide ‘cryptographic power’ to ensure the integrity of a computer service and authenticity of the underlying database. 

After the blockchain revolution, this changes fundamentally. The technology has far reaching implications and so it is worthwhile understanding its language - it will be used much more often in the future.

Decentralization[ad][ae][af] means that there is no single point of failure: there is no one single computer system that can be switched off, censored, or otherwise blocked in order to stop a service.

Distributed means that there is no single hardware infrastructure holding the service. Often, this means that a copy of a database exists on several computers, however, it may also be the case that a database is split between many computers.

Immutability means that strictly speaking, data cannot be changed. However, in practice, this means that data cannot be changed without leaving a trace. Most of the time, this means that old versions can be recovered and that any changes will be protocolled in a system. It is like comparing an excel sheet in which values can be changed at will to a piece of paper. On paper a trace of every manipulation is left displayed (Figure 2). Another practical interpretation would be to call a database an ‘append only’ database. This does not necessarily mean that all data are immutable, e.g. in Blockchain for Healthcare that is an often uttered concern - not the patient data itself immutable, but the access rights to it. [ag][ah]

Figure 2: Blockchain can make research databases immutable, meaning that they cannot be changed without leaving a trace. [ai]

Transparent [aj](provable to the external world) means that a computer program is really running as is publicised (advertised). At the moment, we must rely upon others to calculate things (e.g. impact factor) or to apply post-processing tools to research data in the manner that they claim; after the blockchain revolution, this will be transparent and provable to peers.

In what follows, blockchain will refer to the data structure and blockchain will refer to a system that comprises the above features.

Blockchain - the database view point

Blockchain can be seen as a database with certain characteristics. When compared to current databases interesting correlations can be made (Table 1).


Accessibility to researcher



(Intrinsic) backup


Example use case in Research

Spreadsheet (e.g.Excel)






Workhorse in most researchers daily life

Digital Lab Book[ak]






Workhorse in collaborative lab environments

‘Databases’ (MongoDB, SQL, …)






Backend in data storage, (journal) webpages, libraries, cloud solutions, etc.

Bitcoin-like blockchain






Notarization functionality

Blockchainified database (e.g. BigchainDB)



XXX (ongoing debate)


Not yet described, assumed great potential

Table 1: Comparison of research database characteristics


Blockchain revolution - the technical implementations

Blockchain characteristics are being realized through cryptographic methods and consensus protocols. All of these are long since known, and were initially developed to handle hardware failures, e.g. inside big databases [19]. Nowadays, they are used to provide trust among sometimes unknown and distributed entities.

Blockchain systems rely on many discrete computers to secure the blockchain system and provide the trust or security that is today provided by administrators (Figure 1). These computers can be anonymous entities (miners) which are incentivized to do so by intrinsic value inherent to the system (e.g. Bitcoin, Ethereum) [22]. They can also be defined by a central authority. For example, the securing computers could be provided by trusted and independent research institutes [23] or governmental organizations. However, in contrast to what trusted third party administrators can do today, the blockchain-securing computers cannot alter data stored in the blockchain systems in an undetermined manner, even if someone wanted them to do so. They simply provide ‘cryptographic power’ so as to secure the blockchain. However, if a certain amount of them are compromised, data that is stored in a blockchain system becomes completely unreliable and mutable.[al][am][an] This is not a bug, but an inherent characteristic of the consensus mechanisms. If they are selected carefully and guarded, such an event would be very unlikely.

Blockchain revolution – beyond Bitcoin

There are many Bitcoin-like blockchain systems. Focusing on their ‘coin’ aspect, they are called ‘altcoins’ [24]. Many are just copycats of varying, sometimes questionable legitimacy, some are even scams - but others provide very interesting new features and functionalities that extend far beyond payment processing and hype [25]. A discussion of these is beyond the scope of this article, and would actually be difficult to provide, since innovations and interesting new concepts are being published on almost a daily basis [26]. A list based on current market capitalization can be found here. Here, we will mention some that implement concepts or provide an organizational structure that are especially interesting for research.

One such system, the Ethereum blockchain, goes so far as to provide its own programming language to run distributed, unstoppable, and provable applications [27]. This includes smart contracts [28] which can be used to realize distributed, autonomous applications and organizations [29].

Storj, filecoin, swarm and MaidSAFE are also interesting concepts[ao]. They can be seen as blockchain-based, distributed cloud services to store data, files (or to provide services…). Coins are used to incentivise resource providers who provide hard drive space and network bandwidth (‘the permanent web’ - ‘Web 3.0’ to stress some buzzwords).

Namecoin is one of the first Bitcoin forks and is purposely built to store key-value pairs, in the foremost case, this is being used to register domain (.bit) names without a central entity like ICANN.

There are several projects out that develop platforms that build an incentivisation/rating/reputation system around providing content (including liking/commenting) using a blockchainified attribution and incentive distribution mechanism (Steemit, userfeeds.io and Synereo). More so, there are systems out that `pay` revenue to work at a project (Comakery.com). Certainly, these are very interesting concepts with respect to scientific communication, attribution, work/idea/content sharing incentivisation and have been described as such [30](pevo.science).

Most altcoins work on their own blockchain. However, to make things really confusing, all concepts could technically be implemented in one single blockchain, e.g. the Bitcoin blockchain.

Hyperledger project is a cross-industry collaborative effort, started in December 2015 by the Linux Foundation to support blockchain-based distributed ledgers. The project aims to bring together a number of independent efforts to develop open protocols and standards, by providing a modular framework that supports different components for different uses. This would include a variety of blockchains with their own consensus and storage models, and services for identity, access control, and contracts.[ap][aq][ar]

Open Document Repository[as][at][au] (ODR) by Kubrik is a global network of document repositories run by public libraries. All repositories share a data storage system based on IPFS where they publish the open data, open access articles and all corresponding meta data. All data updates are tracked on a public permissioned ledger (blockchain) that is run between nodes. All participating research publishing entities will have voting power on this blockchain, so that instead of the energy and cost intensive “proof of work” model, the security of this blockchain will be based upon the trust in all participating public academic institutes. ODR Demo: http://kubrik.io/demos/odr/search[av] login for upload available on request) [31] 

Scientific sensemaking itself is much deeper integrated into the protocol itself in dsensor.org and it is designed to evolve peer review to a computational consensus model. Using Dsensor [32] if a scientist creates a thesis and wants to test it the scientist enters the hypothesis in computational form (called a Dmap in Dsensor speak). The Mapping protocol then automates the testing of the science, starting by trawling the Dsensor network for relevant data from other peers. That data is then sampled and ‘scored’ based on its prediction power to verify or challenge the thesis until a computation consensus is established. Science attaining this status then becomes ‘computationally active’ in the network meaning any peer has the ability to tap into the collective living knowledge network and feed in their own unique sensor data get the insights from the science working for them.[aw][ax]

Blockchain revolution - and beyond blockchains[ay][az][ba]

In the blockchain revolution, other systems that show characteristics of blockchain systems, such as being distributed, without a single point of failure, decentralized and immutable, but that are not based on a blockchain (the data structure), would exist.  Actually, they could play a much larger role in the long term than actual blockchain systems.

IPFS (interplanetary filesystem) “is a peer-to-peer distributed file system that seeks to connect all computing devices with the same system of files. In some ways, IPFS is similar to the World Wide Web, but IPFS could be seen as a single BitTorrent swarm, exchanging objects within one Git repository.” Research data or publications that are being stored in IPFS would be available without a centralized server and be very effectively distributed among re-users (See Open Document Repository[bb] by Kubrik).

There are database systems that have blockchain characteristics. For example, BigchainDB is a “big data distributed database and then adds blockchain characteristics - decentralized control, immutability and the transfer of digital assets.” (https://www.bigchaindb.com/whitepaper/bigchaindb-whitepaper.pdf). Many other companies exist providing similar solutions (e.g. ERIS).

Which blockchain for science and knowledge creation?

Blockchain databases may show different characteristics which can be used to divide them into different groups (Table 2).

First, they can be divided by who secures the blockchain database: Can everyone secure the blockchain (permissionless) or only certain entities (permissioned). Permissionless blockchain databases use the above described Proof-of-work or Proof-of-Stake approaches together with an incentivisation through an intrinsic value token to prevent attacks to the network. Permissioned blockchain databases don´t need this, because there are defined and trusted entities that provide ‘cryptographic power’ to secure the blockchain database. Furthermore it is very important to mention, that a permissioned blockchain does not mean that the ‘cryptographic power’ providing trusted third parties have any control over the content that is secured within a permissioned blockchain. They cannot censor or approve beyond the defined protocol in the blockchain system (Not like trusted third parties in the current sense such as service providers (e.g. journal publishers, universities, centralized data repositories, libraries, etc.)).

Secondly, they can be divided into public and private blockchains. This differentiation refers to who can actually use the blockchain database. Is everybody (public) allowed to use the blockchain database or are only certain parties allowed to use it (private)? However, this differentiation is somewhat coarse, because the access and user rights can be much more differentiated depending on the actual use cases. Furthermore, please note that public/private says nothing about who will be able to read the content. For example, a public blockchain can still be used to secure non-public research data.


So, for Science and knowledge creation a blockchain that is secured by trusted third parties’ computers (permissioned) [33] and to which everybody has access (public) seems to be most suitable in the opinion of Soenke Bartling and other peers (Table 2). There are only very few reasons why it should be permissionless, since trusted third parties exist (research institutes, government agencies) (Extance 2017). Please notice, that those trusted third parties would have no control over what is actually stored in the Blockchain for Science. It is hard to believe that even under the worst circumstances a government or other entity would try to infringe blockchain securing computers in a Blockchain for Science to censor research results.[bd] However, this needs to be discussed carefully by the community.



Table 2: A permissioned, public blockchain seems to be most suitable for science and knowledge creation.


Blockchain and the research cycle[bk][bl][bm][bn][bo][bp][bq][br][bs]

In this section, we collect and propose applications of blockchain in science and knowledge creation [34]. We organize this around the research cycle (Figure 3).  Copying of ideas, concepts and text from grey literature (e.g. blog posts) about blockchain for science and its unattributed reuse has recently caused controversy [35,36]. We expect established journals and authors to give appropriate credits in their upcoming articles about blockchain for science that include all means of current publication methods.


Figure 3:[bt][bu] Large parts of the research cycle can make use of blockchain (yellow arch); only the experiment/initial data collection itself cannot. From data collection onwards, the rest of the research circle would then become immutable, comprehensible, and externally provable. This would make larger parts of the research cycle open to scientific self-correction and may make more research results reproducible, true, and useful.



  • Blockchains provide a ‘notarization’ functionality. Through posting a digest (e.g. cryptographic hash) of a text, data, or general purpose file to a blockchain database, it can be proven that this file or text existed at a certain time point. From this digest, one cannot conclude on the topic or content of the text or file, but the owner of the text or file can always prove that he or she was in possession of the file/dataset at a certain time point. The time point is defined by the time the block was created in which the digest was posted. This concept is also named ‘time-stamping’ and ‘proof-of-existence’ [16]. One easily accessible implementation can be found here and a free and accessible version here[bv] [38]. Researchers could post their ideas, research results, or anything else to a blockchain system to prove their existence at a certain time point [39][40]. The company factom is leading the socio-cultural-legal changes around that.
  • For innovations, instead of sending faxes to the patent offices, one could provide a proof-of-existence by posting it to a blockchain database [41]. Strong ‘prior use’ or ‘first to invent’ claims can be made by the Bitcoin blockchain notarization functionality. Bernstein is working in this space.
  • Lab books could post digests to a blockchain system to make them immutable by means of time-stamped entries. A use case is described and potential implications for IP are discussed [42].


  • A study design can be pre-registered to a blockchain, so that it would prevent the arbitrary alteration of study design after the experiment [43–45]. This can also prevent the arbitrary suppression of research studies from being published in case the results do not meet certain expectations (publication bias) [46]. A registration of studies is recommend to increase the value of research [5,9].

Experiment / data acquisition

  • Using blockchain technology, data integrity for approval studies for novel therapy or drugs can be proven to auditors [47,48].
  • All research data that is acquired could go to a blockchain database. All data that is acquired during an experiment could then be available first to a certain audience. It could become openly available and could be reused by other researchers. However, this must not necessarily be the case as a researcher could control who may access the data. For example, they could send research data (or representations (e.g. hashes) of it) to a blockchain system after initial acquisition, time-stamp it, and still keep it secret up to a certain time point. After this time point (e.g. final journal publication), they could then release cryptographic codes so as to make the research data publicly available.[bw][bx][by] This could address one issue that is a reason for ‘Why Most Clinical Research Is Not Useful’ and could restore trust in research, which is currently low, because `research is not transparent, when study data, protocols, and other processes are not available for verification or for further use by others [49–53].
  • Clinical trial consent for protocols and their revisions can be made traceable and secured on a blockchain system [45,54].

Figure 4: Blockchain to connect Internet of Research things. Lab equipment, microscopes, blots, MRI scanners, digital lab books (IoRT ‘Internet of Research Things) could store the data in a blockchainified database. This would leave an immutable, time-stamped proof of data and its acquisition (limitations: see ‘Challenges’).

  • Research data could be acquired by a ‘blockchain-ready’ sensor (microscope, MRI-scanner, Western-Blot scanner, etc.) in an internet-of-things [55] (‘Internet of research things’). Such a sensor would directly encrypt the data (potentially on a hardware level) (Figure 4).
  • As soon as the data is stored in a blockchain database it can be rendered immutable. This means that it cannot be manipulated without leaving a trace (Published at the same time [56]). This can prevent arbitrary data manipulations, be it conscious or inadvertently (e.g. by biased researchers). For example, researchers can prove that they did not drop ‘outliers’ from the initially acquired datasets, or if so, they would then need to explain as to why they dropped them. Research result manipulations (resulting from whichever motivation it may be) at the level of the initial raw data acquisition would require much more effort than data manipulation in a post-processing sheet - which might only require changing a single digit or image. This could improve scientific reproducibility and may make more research results true.
  • Blockchainified research data handling significantly extend the ideas and motivations of open data research, since the integrity of the research data can be proven by means of blockchain (Bell et al. 2017; Huprich 2017; Extance 2017).
  • Blockchainified research data handling could become mandatory for approval studies of novel therapies or drugs, because here truthful data handling, post-processing, and analyses is especially critical. (for example, the FDA cooperates with IBM blockchain).

Data management / analysis

  • Bitcoin and many altcoins use large amounts of computational power for the proof-of-work algorithms. The mining incentives could be set in a way so that some of it is also being used for laborious scientific calculations [57].
  • The recommendation to reduce waste in science which reads: ‘Public availability of raw data and complete scripts of statistical analysis could be required by journals and funding agencies sponsoring new research’ [5,49] could be realized through blockchain.
  • The analysis of the data, post-processing, and statistics can be protocolized in the blockchain database and proven to peers (Figure 5,8). Potentially, statistical analyses and other post-processing steps can run on a blockchain system and become provable to the research community. Hashed and time stamped data publication have been suggested.
  • Data Management Hub (DaMaHub) is a distributed platform for the scientific data workflow that enable scientists to organise and share research data and outcomes in an easy to use, secure and reputation building way. Data is managed in the normal file system environment and synced between different research partners securely and privately. All users share a data storage system based on IPFS where they publish the open data, open access articles and all corresponding meta data. All data updates are tracked on a public permissioned ledger (blockchain) that is run between nodes. [58]


Figure 5: In closed science, scientists just publish a description of their research data and results in their final publication (I). Currently, researchers can publish their research data in data repositories, but that leaves no trace of the data collection or handling process (II). Of course, the final repository could be a blockchain based, decentralized database (III). However, blockchain technology could take the whole process one step further: the whole data handling process could take place in a blockchain system and would therefore be provable and open to scientific self-correction (at will) (IV).

  • Research data can be post processed and analysed in planned, published and reviewed manner. It can be set in stone and realized as a smart contract (Zach Ramsay, in personal discussion). Ideally a smart contract can result in decision with respect to a research hypothesis (in personal discussions with Zach Ramsay, James Littlejohn). This concept should be called ‘Smart Evidence’ (Figure 6). It could be a great way to prevent ex-post-facto hypothesizing. Furthermore, it would be a great way for approval studies, e.g. for drugs and novel therapy concepts. The conditions for the approval of a new drug would be set into blockchain stone before the study commences.

  • 2.png

Figure 6: Smart evidence - research data postprocessing and analysis are set in ‘blockchain stone’ before the data is acquired, post-processing and analysis is automated and it may result in acceptance or rejection of the research hypothesis. This would prevent ex-post facto hypothesizing.

  • Above´s concepts allows anyone to propose (and demonstrate) a different way of doing an analysis. This provides the opportunity for science to act more like a "free-market" where there may be a lab that is really good at producing hypotheses and methodologies, another that has the capacity to run the experiments, and yet another that excels in statistics (Zach Ramsay, personal comment).
  • Smart contracts can be used to prove that data postprocessing is done in a certain way and only in a certain way, even without revealing the whole transaction process on the blockchain [59]. T[bz][ca]his opens up novel possibilities to maintain data autonomy and subject privacy in e.g. healthcare or public health research. E.g. Subject data could be sent to smart contract that is openly (/widely) available and that was reviewed by an ethic committee. The smart contract releases data only after a privacy preserving amount of subjects has been reached [60] or only after a certain time period [61], etc. Furthermore, the fundamental problem of identity information that is being contained in the data itself (face, genome, etc.) is solved, because the smart contract won´t look for it [62][cb][cc][cd][ce][cf][cg][ch][ci][cj][ck][cl][cm]. The privacy and data autonomy could become so convincing that it might become ethically justifiable that all patient/subject data (even unconsented - under current understanding) could automatically contribute to public health research. Applications are humongous! An example workflow would be: Send all blood-pressure data of all patients to a smart contract, the smart contract averages the patient data with respect to a certain region and time. The smart contract assures that only after reaching a privacy assuring mixing of data the average blood-pressure is made available (Figure 7).
  • This will shift privacy related questions to: Who do we trust some data to do all with TO which smart contract do we trust all data to do something with it.

Figure 7: Privacy preserving patient data processing through smart contracts. Patient/research subject data is loaded into a blockchain, the subject data (of multiple subjects) sent to a smart contract (Icon from here). The smart contract is reviewed by a committee and/or public - it is assured that it will only release privacy preserving results, e.g. averages and/or time delayed results. Even if the data itself would reveal the subject´s identity (whole genomes, faces, etc.) the smart contract won´t look at it. Potential in public health, life science research are humongous (unconfirmed idea of the document maintainer).[cn][co]

Data sharing[cp]

  • Through blockchain databases, data can be stored and shared. Blockchain technologies can provide a redundancy and availability of data, e.g. IPFS. This would be a great way to realize open data research (Figure 4, III).
  • Associated cryptography can assure that the data is only available to certain people, groups and from defined time points onwards. If subject anonymity is of concern, this can be organized by means of using strong cryptography, e.g. in case of healthcare data [63], even without a trustee.
  • Blockchain technology could also be used to ‘store’ grant money for research and only release it after the publication and/or reproduction of research data/results [44,64].


  • Publications can be notarized in the blockchain, meaning they can be time-stamped. This idea can be extended to many other, science related processes as lay out herein [65].
  • A decentralized peer-review group (DPG) has been proposed to assure that quality of research [66] [cq][cr][cs][ct]or peer-review can be organized using blockchain [40,67].
  • Ideally blockchain systems will be used to timestamp and attribute contributions to dynamic publications and especially low-threshold dynamic publications [68] or granulated publications (e.g. https://www.sciencematters.io/), such as wikis, in which every change (or single scientific observation) can become time-stamped and attributed in blockchain (many publications, including personal communication with Lambert Heller)[cu].
  • Publications and comments can be shared on a social-media platform and likes, comments, or other interaction can then result in pay-out of coins to incentivise research result sharing [69].
  • A whole open access journal system can be built in a decentralized and distributed form (see aletheia[cv] or Pluto.network)
  • Blockchain systems make it possible to publish research anonymously [70] or with a second online identity - and yet one could still get money or other research impact appreciation for it [71,72]. This may make sense if very controversial results are generated and scientists are afraid that this results are ‘too disruptive’[cw]. Due to the fear of suppression by peers in the complex research social network, they might be afraid to publish such research results or interpretations with their full name [73].
  • In the form of a ‘whistle-blowing’ function or anonymous commenting [74], this could also contribute to the internal self-correction of scientific misconduct. If wanted, publications can be claimed later, and the researcher can replace a name placeholder with their real name.
  • Blockchain technology could be used to ‘sign’ anonymous publications with credibility providing ‘signatures’. For example, the publication could be signed with ‘An english professor in physics with a Hirsch factor of 15’ or ‘A German medical doctor’. A research institute could issue cryptographic certificates to do so [75,76].

Research evaluation

  • A blockchain (e.g. Namecoin) can be used to register and maintain unique research identifiers like (ORCID) or links to publications or datasets[cx] (like DOI) [77].
  • A social network community that incentives content creation and curations can be used to incentives idea, data and results research sharing [30].
  • The quality of research is currently assessed using impact factor and other altmetrics (like RG score, Altmetric). One has to trust the third parties issuing these to correctly calculate such metrics. With blockchain technology and smart contracts, this could change so that the way the metrics are being calculated is externally provable.
  • A ‘research currency’ as an incentivization system to ‘make more published research results true’ as described in [6] could be realized using blockchain technology and without a trusted third party, also described as micro-credits [78] .[cy]
  • Science reputation systems can be built using blockchain without a trusted third party.
  • As such, a Decentralized Autonomous Academic Endorsement System has been proposed and interesting implementation ideas and next implementation steps have been disclosed [72].[cz]


Figure 8: Overview of what parts of the scientific process that are open to scientific self-correction. Blockchainified research may make the whole research process traceable and open - at will.

Research funding

  • Prediction markets [79] to confirm results and to incentivise research could also be used in science [80–83] and could be implemented on blockchain (see Gnosis and Augur).
  • Blockchain could be used to realize a `money-back` functionality for irreproducible research results [56].
  • Blockchain would seem to provide a good mechanism for realising the "credit" systems being proposed for using shared infrastructures like NIH Commons and European Open Science Cloud (Proposed here by Eoghan Ó Carragáin).
  • New methods of research fund distribution could easily be realized with blockchain technology and smart contracts. For example, a system in which researchers redistribute 50% of their research money among peers [84] can be realized using smart contracts [85]. Research funds could be sent completely anonymously, without trusted third parties.
  • Similar to a DAO (distributed autonomous organization) [29,86] that could complement functions provided by companies, a research-DAO (or DARO: distributed autonomous research organization) can be used to complement research funding agencies (Figure 9) (example projects: Collider-X.org, Space.coop or Replication foundation)[db].

Artboard 4.png

Figure 9: A DARO (distributed autonomous research organization)[dc][dd] allocates and distributes research resources.

  • Concepts similar to colored coins / cryptocurrency tokens could be used to relate research funds to some conditions, even if the distribution mechanism is anonymous and ‘black boxed’ on blockchain. For example, a funding agency could direct the research funds to certain research fields, locations, or institutions. Only researchers that fulfill those requirements would be able to claim those coins.
  • Blockchain could provide many novel ways to distribute research money. For example, research funding provider could pick a combination of characteristics of different kinds of researcher behavior that they want to support. E.g. the amount of patents, citations, tweets, likes, blogs, datasets shared by a researcher, combined with age, location, academic rank, early citations, etc. The problem with the current system is that novel ways of research money distributions aren´t easily employed and system gamers can easily adopt to a constant funding environment. If one asks for patents, there will be patent applications since nothing stops one from writing another rather meaningless application. If funding distribution is under constant and unforeseeable mutation system gaming will more look like gambling than gaming and researchers might come up with an overall behavior that is best for knowledge creation [87]. At least this might hold true for third party research money. Blockchain will also prevent the potential allegation of arbitrariness for research money distribution since the process can be make completely proofable on the blockchain.
  • Blockchain for science and knowledge will not only aid researchers to better conduct and publish their inquiries, but could also engage the public through a more transparent research process. Ultimately this technology could open up the academic process to the public for inquiry and even participation, while simultaneously safeguarding the integrity of their research. This open access could inspire and enable amateur researchers to collaborate with professional researchers in an effort to crowdsource research using the principles of citizen science.
  • The blockchain (r)evolution launched a new economical field - the ‘token economy’. It evolved out of more and more often occurring ICOs (initial [de]coin offering) to ‘crowdfund’ projects and companies. Here, an established blockchain token (aka coin, e.g. Bitcoin, Ether, etc) is exchanged for a novel token that is related to a project. Many noteworthy projects reached astronomic investment sums within record breaking times. Whether this economy can continue to  thrive at its current pace without corrections remains to be seen. There are a couple of differences to traditional crowdfunding [88]. First, the token is in many cases directly tradeable after the initial funding round (if no vesting is employed), which may incentivise early investors[df]. Secondly, in best practice tokens reticulate the product itself (e.g. network tokens). Lastly, it may pay a dividend that is assured via smart contracts. The market value of a token typically rises corresponding to the proliferation and success of its project. Based on these observations, it is claimed that the token economy is the business model of the web 3.0/open source projects. Cashing out on network effects doesn´t rely on creating a centralized single points of failure (e.g. Facebook) anymore, therefore it is considered a breakthrough in open designs [89,90]. ‘Cryptocurrencies (aka tokens) are the spiritual heirs to Linux and Wikipedia’ [91]. This extends so far as to create ‘mememarkets’ that allows the monetization of all information and its network effects [92]. So, if this token economy could be applied to scientific[dg] ideas a[dh][di][dj][dk][dl]nd groundbreaking discoveries, we could incentivise scientist to spread ideas early and strongly. This stands in contrast to the current system, which incentivises scientists to remain inert until publication, patent application, etc. It has been claimed that this could also be a solution to the famous innovators dilemma [93]. [dm][dn]
  • More and more ICOs are announced in almost all fields [94,95]. Astronomic sums are collected in record breaking times. Some ICOs are scams, others lack team, product or even clear project plans. Other say that ICOs are a great way to finance open and onboarding businesses. Without doubt, ICOs can be used to finance research projects. The first ICO for a research project is Arna Genomics[1]. ICOs are considered a democratization of investing. One can ask why the public should be interested in directly investing into research projects that might be far away from applied use. It is questionable whether the lay investor can properly assess the potential value of a research project. This needs to be discussed and critiqued in the mid term future as soon as it becomes clear that ICO are a working concept for science funding. Many small investors might have interests and passions different from those of governmental and institutional research funders. The current research funding system isn´t perfect – it creates a massive workload for grant applications, reviews, etc. that prevents scientists from spending time at their research. The entanglement of decisions committees with benefactors is unavoidable, most of the time useful, but it might also support the low-risk, low-gain research projects and not the outliers. So ICOs add interesting aspects to the portfolio of research funding means. The scientific community should closely watch this development and guide it to take useful and constructive pathways as early as possible. We believe that the cryptoeconomy (ICOs and other token systems) can have a solid standing in science funding in the future and should coexist with established methods, once legal, cultural and structural frameworks are worked out (The author started a community project to develop good guidelines for projects that apply for ICOs in science and research). The wider public needs to be convinced that a proposed research project is valid. So by this exposure science culture might enter a new era of transparency and here the blockchain could add convincing methods (e.g. immutable data trailing as Arna Genomics is showcasing to proof widely that their test works as proposed).
  • What should token for science/research projects represent? One thing would be ‘asset-backed’ tokens as described here [96] as a bet on the commercially usable part of the intellectual property. However, is that enough? Couldn't there be tokens that represent a purely scientific value of a research project? Imagine you would have invested in the ‘relativity theory’ at a time when it was still a crazy idea? Ultimately this might make the onboarding effects of the token economy available to support the dissemination of really novel, scientific concepts. Well, these constructs might seem far out, and might look as a brain twister at first sight. For sure, they will always look like Ponzi or pyramid schemes, because a pure scientific value will never create a commercial revenue stream that pays dividends. But these constructs might also be a way to create a completely new asset class that might support risky and really innovative science projects, because in early states the pyramid like dissemination will allow huge margins and people might be incentivised to invest early and strongly in crazy ideas … Another problem arises: How should we deal with ‘negative results’ in terms of the commercial applicability of a project that is still a valuable scientific finding?


  • One fundamental challenge of blockchain is the real-world/blockchain interface problem. How can the blockchain world learn about real-world facts? One instance of this problem is the fact that one has to trust the researcher[dp][dq], sensors, etc. to correctly collect the initial research data. Another example of this problem is the question as to how individual researchers/subjects are recognized within the blockchain world and how their identity is confirmed[dr]. This could be done by research institutes (often an institutional email is used to this end, or indeed cryptographic certificates) or other entities that already have a large database of researchers (ORCID, online social networks or publishers).
  • The current legislation did not foresee the blockchain revolution. Many legal and tax questions remain currently unresolved. These challenges are not specific to blockchain for knowledge creation, but they also exist in other applications of blockchain technology, and are an exciting, evolving field. This is especially intriguing when new funding models (ICOs), anonymous research money distribution, etc. will find widespread use.
  • The scalability of most blockchain implementations, e.g. the amount of transactions per unit time, is limited compared to other, centralized technologies, which is kind of obvious, because a status has to flow through a much larger network; the optimization of this scalability is a part of ongoing blockchain research. Sidechains / local blockchains, etc. are one option, many more are discussed on a technical level. [ds]
  • Implementation! How will a blockchainified research workflow look like? Blockchain needs to be highly integrated into current research workflow and tools.
  • Business models! Current business models are arranged around the prospect of creating a single point of failure/container - blockchain changes this - interesting concepts are currently evolving.

How can blockchain help both ‘kind of sciences’?

  • Provable, immutable data acquisition, post-processing and storage
  • Smart evidence
  • Research subject privacy, crypto-assured study blinding
  • Connection the IoRT
  • Transparent approval studies
  • Unconventional, innovative, but still transparent means of research money distribution
  • Early, simple and strong incentivisation of ‘crazy’ concepts / ideas through blockchain token
  • Researcher anonymity for whistle-blowing / revolutionary standpoints

More reproducible and true results

More innovation / discovery

Figure 10: How blockchain could help both `kinds of sciences’.


The blockchain revolution is a game changer and hence chances are that this can be used to break with inappropriate cultures. Indeed, blockchain technology could be used to ‘Increase value and reduce waste’ [5], by opening the research cycle to scientific self-control beyond the final publication and might therefore be a fix to the current reproducibility crisis in science.  Furthermore, it could provide new means for the `machine room` of science (e.g. attribution, assessment, research funding, etc.), which could ideally be used to support really innovative research. So blockchain could improve both kind of sciences (Figure 10).

1.         Swan M. Blockchain: Blueprint for a New Economy [Internet]. O’Reilly Media; 2015. Available: https://books.google.de/books?id=RHJmBgAAQBAJ

2.         The trust machine. In: The Economist [Internet]. 2015 [cited 22 Jun 2016]. Available: http://www.economist.com/news/leaders/21677198-technology-behind-bitcoin-could-transform-how-economy-works-trust-machine

3.         Meta-Council on Emerging Technologies. In: World Economic Forum [Internet]. [cited 6 Jul 2016]. Available: https://www.weforum.org/communities/meta-council-on-emerging-technologies/

4.         Ulieru M. Blockchain Enhances Privacy, Security and Conveyance of Data. Scientific American. Available: http://www.scientificamerican.com/article/blockchain-enhances-privacy-security-and-conveyance-of-data/. Accessed 6 Jul 2016.

5.         Ioannidis JPA, Greenland S, Hlatky MA, Khoury MJ, Macleod MR, Moher D, et al. Increasing value and reducing waste in research design, conduct, and analysis. Lancet. 2014;383: 166–175. doi:10.1016/S0140-6736(13)62227-8

6.         Ioannidis JPA. How to make more published research true. PLoS Med. 2014;11: e1001747. doi:10.1371/journal.pmed.1001747

7.         Ioannidis JPA. Why Most Published Research Findings Are False. PLoS Med. 2005;2: e124. doi:10.1371/journal.pmed.0020124

8.         Young NS, Ioannidis JPA, Al-Ubaydli O. Why Current Publication Practices May Distort Science. PLoS Med. Oktober 7, 2008;5: e201. doi:10.1371/journal.pmed.0050201

9.         Goldacre B. Are clinical trial data shared sufficiently today? No. BMJ. 2013;347: f1880. doi:10.1136/bmj.f1880

10.         Engber D. Think Psychology’s Replication Crisis Is Bad? Welcome to the One in Medicine. In: Slate Magazine [Internet]. 19 Apr 2016 [cited 25 Jun 2016]. Available: http://www.slate.com/articles/health_and_science/future_tense/2016/04/biomedicine_facing_a_worse_replication_crisis_than_the_one_plaguing_psychology.html

11.         Freedman LP, Cockburn IM, Simcoe TS. The Economics of Reproducibility in Preclinical Research. PLoS Biol. 2015;13: e1002165. doi:10.1371/journal.pbio.1002165

12.         Angell M. Drug companies & doctors: A story of corruption. New York Rev Books. 2009;56: 8–12. Available: http://www.fondazionedibella.org/cms-web/upl/doc/Documenti-inseriti-dal-2-11-2007/Truth%20About%20The%20Drug%20Companies.pdf

13.         Schneider L. Voinnet aftermath: ethical bankruptcy of academic elites. In: For Better Science [Internet]. 7 Apr 2016 [cited 7 Jul 2016]. Available: https://forbetterscience.wordpress.com/2016/04/07/voinnet-aftermath-ethical-bankruptcy-of-academic-elites/

14.         pubpeer. A crisis of trust | PubPeer [Internet]. [cited 7 Jul 2016]. Available: http://blog.pubpeer.com/?p=164&utm_source=rss&utm_medium=rss&utm_campaign=a-crisis-of-trust

15.         Munafò MR, Nosek BA, Bishop DVM, Button KS, Chambers CD, du Sert NP, et al. A manifesto for reproducible science. Nature Human Behaviour. Nature Publishing Group; 2017;1: 0021. doi:10.1038/s41562-016-0021

16.         Haber S, Stuart H, W.Scott S. How to time-stamp a digital document. J Cryptology. 1991;3. doi:10.1007/bf00196791

17.         Nakamoto S. Bitcoin: A peer-to-peer electronic cash system [Internet]. 2008. Available: http://www.cryptovest.co.uk/resources/Bitcoin%20paper%20Original.pdf

18.         Tschorsch F, Scheuermann B. Bitcoin and beyond: A technical survey on decentralized digital currencies. ieeexplore.ieee.org; 2015; Available: http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7423672

19.         Lamport L, Shostak R, Pease M. The Byzantine Generals Problem. ACM Trans Program Lang Syst. New York, NY, USA: ACM; 1982;4: 382–401. doi:10.1145/357172.357176

20.         Lamport L. The Part-time Parliament. ACM Trans Comput Syst. New York, NY, USA: ACM; 1998;16: 133–169. doi:10.1145/279227.279229

21.         Jakobsson M, Juels A. Proofs of work and bread pudding protocols. Secure Information Networks. Springer; 1999. pp. 258–272. Available: http://link.springer.com/chapter/10.1007/978-0-387-35568-9_18

22.         Buterin V. On Public and Private Blockchains. Ethereum Blog. 2015;

23.         Science B.0 on Twitter. In: Twitter [Internet]. [cited 22 Jun 2016]. Available: https://twitter.com/science_b0/status/712126802064449536

24.         Wisniewska A. Altcoins [Internet]. Institute of Economic Research; 2016 May. Report No.: 14/2016. Available: http://ideas.repec.org/p/pes/wpaper/2016no14.html

25.         Wang S, Vergne J-P. Buzz Factor or Innovation Potential: What Explains Cryptocurrencies’ Returns? PLoS One. 2017;12: e0169556. doi:10.1371/journal.pone.0169556

26.         Hurlburt G. Might the Blockchain Outlive Bitcoin? IT Prof. 2016;18: 12–16. doi:10.1109/MITP.2016.21

27.         Buterin V. Ethereum: A Next-Generation Smart Contract and Decentralized Application Platform. 2013a. URL $\{$http://ethereum org/ethereum html$\}$.

28.         Nick Szabo -- The Idea of Smart Contracts [Internet]. [cited 22 Jun 2016]. Available: http://szabo.best.vwh.net/smart_contracts_idea.html

29.         DAOs, DACs, DAs and More: An Incomplete Terminology Guide - Ethereum Blog. In: Ethereum Blog [Internet]. 6 May 2014 [cited 22 Jun 2016]. Available: https://blog.ethereum.org/2014/05/06/daos-dacs-das-and-more-an-incomplete-terminology-guide/

30.         (6) T, (7) J, (4) P. A Proposal For An Incentivized Synthetic Biology System Subchain On The Steem Blockchain Platform — Steemit. In: Steemit [Internet]. [cited 6 Aug 2016]. Available: https://steemit.com/science/@transhuman/a-proposal-for-an-incentivized-synthetic-biology-system-subchain-on-the-steem-blockchain-platform

31.         Engineering K. Scholarly Publishing & Knowledge Preservation [Internet]. [cited 17 Jan 2017]. Available: http://kubrik.io/projects/odr/

32.         NickLambert. Developer Case Study – Dsensor. In: MaidSafe [Internet]. 8 Feb 2017 [cited 20 Mar 2017]. Available: https://blog.maidsafe.net/2017/02/08/developer-case-study-dsensor/

33.         Archaeological Blockchain Launches Token Exchange Campaign - KAPU [Internet]. [cited 30 Sep 2017]. Available: http://cryptopressrelease.com/archaeological-blockchain-launches-token-exchange-campaign-kapu/

34.         Blockchain Science: An interview with Gemma Milne, Science Disrupt - Unblocked Events. In: Unblocked Events [Internet]. 1 Nov 2017 [cited 15 Nov 2017]. Available: https://unblockedevents.com/2017/11/01/blockchain-science-interview-gemma-milne-science-disrupt/

35.         Plagiarism concerns raised over popular blockchain paper on catching misconduct - Retraction Watch. In: Retraction Watch [Internet]. 14 Jul 2016 [cited 15 Jul 2016]. Available: http://retractionwatch.com/2016/07/14/plagiarism-concerns-raised-over-popular-blockchain-paper-on-catching-misconduct/

36.         Satoshi Village [Internet]. [cited 27 Mar 2017]. Available: http://blog.dhimmel.com/irreproducible-timestamps/

37.         Extance A. Could Bitcoin technology help science? Nature. 2017;552: 301–302. doi:10.1038/d41586-017-08589-4

38.         Bradley J, Bradley J. Scientific Research Needs a Trustless Blockchain Architecture to Be Trusted - CCN: Financial Bitcoin & Cryptocurrency News. In: CCN: Financial Bitcoin & Cryptocurrency News [Internet]. 13 May 2016 [cited 2 Jul 2016]. Available: https://www.cryptocoinsnews.com/scientific-research-needs-a-trustless-blockchain-architecture-to-be-trusted/

39.         Redman J. Clinical Trials Show the Blockchain Can Stop “Fraudulent” Science. In: Bitcoin News [Internet]. 19 May 2016 [cited 2 Jul 2016]. Available: https://news.bitcoin.com/clinical-blockchain-stop-fraud-science/

40.         Astroblocks Puts Proofs of Scientific Discoveries on the Bitcoin Blockchain [Internet]. [cited 22 Jun 2016]. Available: http://insidebitcoins.com/news/astroblocks-puts-proofs-of-scientific-discoveries-on-the-bitcoin-blockchain/31153

41.         Cawrey D, Wolinsky J, Rampton J, Wolinsky R, Palmer D. Bitcoin’s Technology Could Revolutionize Intellectual Property Rights. In: CoinDesk [Internet]. 8 May 2014 [cited 22 Jun 2016]. Available: http://www.coindesk.com/how-block-chain-technology-is-working-to-transform-intellectual-property/

42.         Blockchain and Smart-Contracts applied to Evidence Notebook [Internet]. [cited 18 Oct 2016]. Available: https://healthcaresecprivacy.blogspot.de/2016/08/blockchain-and-smart-contracts-applied.html

43.         Irving G, Holden J. How blockchain-timestamped protocols could improve the trustworthiness of medical science. F1000Res. 2016;5: 222. doi:10.12688/f1000research.8114.1

44.         Carlisle BG. The Grey Literature, apparently — Proof of prespecified endpoints in medical research with the bitcoin blockchain [Internet]. [cited 15 Jul 2016]. Available: http://www.bgcarlisle.com/blog/2014/08/25/proof-of-prespecified-endpoints-in-medical-research-with-the-bitcoin-blockchain/

45.         Benchoufi M, Ravaud P. Blockchain technology for improving clinical research quality. Trials. 2017;18: 335. doi:10.1186/s13063-017-2035-z

46.         Blockchains For Science: Aligning Research Incentives. In: Doing Distributed Business [Internet]. [cited 22 Jun 2016]. Available: https://db.erisindustries.com/science/2016/03/14/blockchains-and-science/

47.         Nugent T, Upton D, Cimpoesu M. Improving data transparency in clinical trials using blockchain smart contracts. F1000Res. 2016;5: 2541. doi:10.12688/f1000research.9756.1

48.         Repalli J. Blockchain in life sciences [Internet]. [cited 24 Feb 2017]. Available: https://de.slideshare.net/JayanthiRepalli/blockchain-in-life-sciences

49.         Ioannidis JPA. Why Most Clinical Research Is Not Useful. PLoS Med. 2016;13: e1002049. doi:10.1371/journal.pmed.1002049

50.         Schneider L. False priorities at EU2016NL: Mandate Open Data instead of Gold Open Access! In: For Better Science [Internet]. 28 Apr 2016 [cited 2 Jul 2016]. Available: https://forbetterscience.wordpress.com/2016/04/28/false-priorities-at-eu2016nl-mandate-open-data-instead-of-gold-open-access/

51.         Dhillon V. Blockchain-enabled open science framework. In: O’Reilly Media [Internet]. 3 Nov 2016 [cited 16 Nov 2016]. Available: https://www.oreilly.com/ideas/blockchain-enabled-open-science-framework

52.         Bell J, LaToza TD, Baldmitsi F, Stavrou A. Advancing Open Science with Version Control and Blockchains. Proceedings of the 12th International Workshop on Software Engineering for Science. Piscataway, NJ, USA: IEEE Press; 2017. pp. 13–14. doi:10.1109/SE4Science.2017..11

53.         Huprich SK. JPA Promotes Open Science. J Pers Assess. 2017; 1–2. doi:10.1080/00223891.2017.1319711

54.         Benchoufi M, Porcher R, Ravaud P. Blockchain protocols in clinical trials: Transparency and traceability of consent. F1000Res. 2017;6. doi:10.12688/f1000research.10531.1

55.         soenkeba on Twitter. In: Twitter [Internet]. [cited 2 Jul 2016]. Available: https://twitter.com/soenkeba/status/697436268737777664

56.         Topol EJ. Money back guarantees for non-reproducible results? BMJ. 2016;353: i2770. doi:10.1136/bmj.i2770

57.         Wagner A. Putting the Blockchain to Work For Science! In: Bitcoin Magazine [Internet]. 22 May 2014 [cited 2 Jul 2016]. Available: https://bitcoinmagazine.com/articles/putting-the-blockchain-to-work-for-science-gridcoin-1400747268

58.         Engineering K. Data Management Hub [Internet]. [cited 17 Jan 2017]. Available: https://kubrik.io/projects/damahub/

59.         Kosba A, Miller A, Shi E, Wen Z, Papamanthou C. Hawk: The Blockchain Model of Cryptography and Privacy-Preserving Smart Contracts. 2016 IEEE Symposium on Security and Privacy (SP). 2016. pp. 839–858. doi:10.1109/SP.2016.55

60.         Twitter [Internet]. [cited 19 Mar 2017]. Available: https://twitter.com/soenkeba/status/842337875245056000

61.         Twitter [Internet]. [cited 19 Mar 2017]. Available: https://twitter.com/soenkeba/status/842338036553793536

62.         Twitter [Internet]. [cited 19 Mar 2017]. Available: https://twitter.com/science_b0/status/842299809826516992

63.         Scott M. The Future of Medical Records: Two Blockchain Experts Weigh In [Internet]. [cited 2 Jul 2016]. Available: https://btcmanager.com/news/the-future-of-medical-records-two-blockchain-experts-weigh-in/

64.         Furlanello C, De Domenico M, Jurman G, Bussola N. Towards a scientific blockchain framework for reproducible data analysis [Internet]. arXiv [cs.CY]. 2017. Available: http://arxiv.org/abs/1707.06552

65.         Hoy MB. An Introduction to the Blockchain and Its Implications for Libraries and Medicine. Med Ref Serv Q. 2017;36: 273–279. doi:10.1080/02763869.2017.1332261

66.         Magazine B. Bitcoin Magazine | Bitcoin and Blockchain News [Internet]. [cited 22 Jun 2016]. Available: https://bitcoinmagazine.com/articles/from-bench-to-bedside-enabling-reproducible-commercial-science-via-blockchain-1464881141)

67.         Tennant JP, Dugan JM, Graziotin D, Jacques DC, Waldner F, Mietchen D, et al. A multi-disciplinary perspective on emergent and future innovations in peer review. F1000Res. 2017;6. doi:10.12688/f1000research.12037.1

68.         Dynamic Publication Formats and Collaborative Authoring - Springer [Internet]. [cited 22 Jun 2016]. Available: http://link.springer.com/chapter/10.1007/978-3-319-00026-8_13

69.         benjojo, pharesim. Independent scientists could blog their research onto the Steemit blockchain — Steemit. In: Steemit [Internet]. [cited 1 Aug 2016]. Available: https://steemit.com/research/@benjojo/independent-sientists-could-blog-their-research-onto-the-steemit-blockchain

70.         Teif VB. Science 3.0: Corrections to the Science 2.0 paradigm [Internet]. arXiv [cs.DL]. 2013. Available: http://arxiv.org/abs/1301.2522

71.         Science B.0 on Twitter. In: Twitter [Internet]. [cited 2 Jul 2016]. Available: https://twitter.com/science_b0/status/695886036447203328

72.         b8d5ad9d974a44e7e2882f986467f4d. Towards Open Science: The Case for a Decentralized Autonomous Academic Endorsement System [Internet]. Zenodo; 2016. doi:10.5281/zenodo.60054

73.         Bartling S. Science goes darknet? In: Alexander von Humboldt Institut für Internet und Gesellschaft [Internet]. [cited 29 Nov 2016]. Available: http://www.hiig.de/blog/science-goes-darknet/

74.         PubPeer and Anonymity in Science - Neuroskeptic. In: Neuroskeptic [Internet]. 12 Dec 2015 [cited 13 Mar 2017]. Available: http://blogs.discovermagazine.com/neuroskeptic/2015/12/12/pubpeer-and-anonymity-in-science/

75.         Lab MM. Certificates, Reputation, and the Blockchain — MIT MEDIA LAB. In: Medium [Internet]. 27 Oct 2015 [cited 22 Jun 2016]. Available: https://medium.com/mit-media-lab/certificates-reputation-and-the-blockchain-aee03622426f

76.         The Possibilities of Badges and Blockchain - DML Central. In: DML Central [Internet]. 11 Feb 2016 [cited 22 Jun 2016]. Available: http://dmlcentral.net/the-possibilities-of-badges-and-blockchain/

77.         Lambert Heller on Twitter. In: Twitter [Internet]. [cited 6 Jul 2016]. Available: https://twitter.com/Lambo/status/192725705556103169

78.         Casati R, Origgi G, Simon J. Micro-credits in scientific publishing. Journal of Documentation. emeraldinsight.com; 2011; Available: http://www.emeraldinsight.com/doi/abs/10.1108/00220411111183546

79.         Wolfers J, Zitzewitz E. Prediction Markets [Internet]. National Bureau of Economic Research; 2004. doi:10.3386/w10504

80.         Dreber A, Pfeiffer T, Almenberg J, Isaksson S, Wilson B, Chen Y, et al. Using prediction markets to estimate the reproducibility of scientific research. Proc Natl Acad Sci U S A. 2015;112: 15343–15347. doi:10.1073/pnas.1516179112

81.         Hanson R. Could gambling save science? Encouraging an honest consensus. Social Epistemology. 1995;9: 3–33. doi:10.1080/02691729508578768

82.         Almenberg J, Kittlitz K, Pfeiffer T. An experiment on prediction markets in science. PLoS One. 2009;4: e8500. doi:10.1371/journal.pone.0008500

83.         Park I-U, Peacey MW, Munafò MR. Modelling the effects of subjective and objective decision making in scientific peer review. Nature. 2014;506: 93–96. doi:10.1038/nature12786

84.         Bollen J, Crandall D, Junk D, Ding Y, Börner K. From funding agencies to scientific agency: Collective allocation of science funding as an alternative to peer review. EMBO Rep. 2014;15: 131–133. doi:10.1002/embr.201338068

85.         Science B.0 on Twitter. In: Twitter [Internet]. [cited 22 Jun 2016]. Available: https://twitter.com/science_b0/status/710412421987508225

86.         Popper N. A Venture Fund With Plenty of Virtual Capital, but No Capitalist. The New York Times. 21 May 2016. Available: http://www.nytimes.com/2016/05/22/business/dealbook/crypto-ether-bitcoin-currency.html. Accessed 22 Jun 2016.

87.         Bartling S. Research system gamers might become unlucky gamblers through blockchain. In: LinkedIn Pulse [Internet]. 16 Nov 2016 [cited 9 Mar 2017]. Available: https://www.linkedin.com/pulse/research-system-gamers-might-become-unlucky-gamblers-through-s%C3%B6nke

88.         Mersch M. The Rise of the Token Sale – OpenOcean – Medium. In: Medium [Internet]. OpenOcean; 3 May 2017 [cited 14 Jun 2017]. Available: https://medium.com/openocean/the-rise-of-the-token-sale-28f2d07651c9

89.         Dixon C. Crypto Tokens: A Breakthrough in Open Network Design. In: Medium [Internet]. Medium; 1 Jun 2017 [cited 14 Jun 2017]. Available: https://medium.com/@cdixon/crypto-tokens-a-breakthrough-in-open-network-design-e600975be2ef

90.         McConaghy T. Tokenize the Enterprise – The BigchainDB Blog. In: The BigchainDB Blog [Internet]. The BigchainDB Blog; 6 Jun 2017 [cited 14 Jun 2017]. Available: https://blog.bigchaindb.com/tokenize-the-enterprise-23d51bafb536

91.         Naval Ravikant on Twitter. In: Twitter [Internet]. [cited 14 Jun 2017]. Available: https://twitter.com/naval/status/870302461596020736

92.         de la Rouviere S. mememarkets [Internet]. Github; Available: https://github.com/simondlr/mememarkets

93.         Trent McConaghy on Twitter. In: Twitter [Internet]. [cited 15 Jun 2017]. Available: https://twitter.com/trentmc0/status/873860105535709184

94.         [No title] [Internet]. [cited 30 Sep 2017]. Available: https://people.cs.uchicago.edu/~teutsch/papers/ico.pdf

95.         Analyzing Token Sale Models [Internet]. 9 Jun 2017 [cited 30 Sep 2017]. Available: http://vitalik.ca/general/2017/06/09/sales.html

96.         Wong A. Cryptocurrencies for science: Asset-backed tokens for science. In: Medium [Internet]. Alternative Assets: Impact Investing For Science; 31 Oct 2017 [cited 28 Nov 2017]. Available: https://medium.com/impact-investing-science-tech-ed/cryptocurrencies-for-science-asset-backed-tokens-for-science-e21a50d98a71

Document status: Living document, comments, suggestions, examples, pointers to wrong attributions and missing references highly appreciated. It`s an open living document - it doesn't look shiny and perfect, but it is honest, free, collaborative, up to date, patent troll preventing and under constant peer-review. It is kept most of the time in a form that allows continuous reading. For interesting discussions, contributors and older viewpoints I strongly suggest to check document history and resolved comments. The time is right - in real world there is seldom a single inventor!

You disagree with something? Comment anonymously (or pseudonymously by leaving a public key or signature in the comment). The maintainer will never delete or censor any comment, only resolve them!

Published versions here: LSE Impact blog, Zenodo, Researchgate and irights.info (German). This living document (with full history, comments and attributions and hence post-publication peer-review) will stay here, open to public since the beginning of July 2016 (publication tweet). A version was time-stamped with proof-of-existence. This publication is an extension of an earlier, pseudonymous, cryptographically signed and hash-time-stamped open publication from February 2015. Attributed 1:1 figure and text reuse most welcome, notification to the maintaining author appreciated. Writing about Blockchain & Science - please provide reference to all current means of publication methods - let blockchain for science be an exemplar for open science - everything else is highly contradictory :)

Reference static versions as:

  1. Bartling, Sönke; Fecher, Benedikt. (2016). Blockchain for science and knowledge creation. Zenodo. 10.5281/zenodo.60223
  2. Bartling, Sönke, & et contributors to living document. (2017). Blockchain for Open Science and Knowledge Creation. 10.5281/zenodo.401369


[1] Disclosure: The maintainer of this document is advisor to the Arna Genomics team. He has financial interests in promoting the ICO. However, his interests are more in kickstarting the cryptoeconomy for science and to learn from their experience, educate about ICOs and guide the cryptoeconomy in science into a constructive direction as soon as possible.

[a]Add: http://www.digitaljournal.com/tech-and-science/technology/canada-s-national-research-council-trials-blockchain-technology/article/513117


[c]Add: http://globexsci.io/

[d]looks like scam, text in whitepaper about OA problems is stolen from our paper :(

[e]Add: CryptSubmit:







Peer Review Feedback




[f]Add https://moringa.pub/

[g]Add: https://www.timeshighereducation.com/news/what-blockchain-technology-could-mean-for-universities

[h]Add: https://matryx.ai/?utm_campaign=dg&utm_source=coin-telegraph&utm_medium=ntv


[j]ADD: https://www.scienceroot.com


[l]My immediate interest is in blockchain's potential for disintermediation of data reuse and data sharing. What role can it play in  the simplification and automation of efforts to ensure  end to end compliance of all parties to Open Science principles?

[m]Why? the blockchain is just a ledger. it records ins and outs and by means of smart contracts it makes possible for machines to do things on ins and outs. how would it make possible scientific self correction? other than by having all data in one decentralized storage facility.

[n]What reproducibility crisis in science? Maybe foodnote to DOI based document could help. Can we differentiate the crisis and where is the new thing to this approach?


[p]Interesting attitude to one third trying to improve reproducibility and more than that setting up workflows of senior scientists. Very encouraging article thanks.

[q]Whats the meaning of “Data points" to cut down on cherry picking?

[r]It would also be interesting to think how less developed countries without robust scientific infrastructures that also sometimes can't debate with global anglo publishing system, would benefit from this.


[t]add novel and innovative "market places"/"artifical competetion" ...

[u]The raw data collection via secured IoT device for the market to analyze and verify, and then a larger set of scientists analyzing the data for significance could be interesting.

[v]Blockchain became popularised as the data structure behind Bitcoin. However, the term ‘Blockchain’ is nowadays extended much beyond its initial meaning. Blockchain now commonly refers to a new way to organise computer services. The way computer services are organised today has become so second-nature to us so that we don’t question it anymore. We know that we have to trust our banks, online social platforms and every other online service not to manipulate our accounts and data, the amount of ‘likes’ or to delete our content. Blockchain approaches this issue differently. As the trust machine, a computer service becomes independent from the control of the underlying hardware. Nobody can just delete something or forge the output in an undetermined manner. Computer data becomes as trustworthy as a signed piece of paper, i.e. a contract. Computer services become provable and distributed among many systems, which are two of the major advantages of peer-to-peer networks.

Given the importance of science and research to society and humanity, it is very important that we are not required to trust a single entity or entities with securing data. From a more specific viewpoint, blockchain seems to be ideally suited to track research results, attribute findings to


researchers, make data-post processing externally proveable, and improve numerous aspects of research and communication workflows. The blockchain revolution brought up another interesting field: The crypto-economy or token economy. Switzerland has taken a lead in this field and has become the fastest growing crypto-economy in the world.

[w]It is clear that the immutability of blockchain makes it possible to generate trustworthy units/tokens with minimal effort. These can be associated with projects, used for crowdfunding campaigns (Initial Coin Offerings, ICOs) or to fund research ideas. Again, from a high level viewpoint, this bears the potential to bring widespread market interest and investment capital much earlier into the research process. In contrast to previous (and ongoing) evolutions in science and knowledge creation (such as Science 2.0 or Open Science) it is almost certain that funding regimes will be more widely affected. This has enormous potential. The time is right to structure this evolutionary step so that it takes the routes that are most beneficial to science and knowledge creation as early as possible, and to avoid dead-end silos and business models.

[x]I think a post-modern reference would come in handy before categorically calling a new technology 'hype'. That makes for very subjection lead-in. I know many writer's tools that would flag such a level of colour. Judging by the running context, this paragraph really does jump out at me. Perhaps the super-social analysis is best left for some later review in the conclusion?

Also, in general, '=' and '≈' as well as other such call-outs (), e.g., etc. these will disrupt readers and should be smoothed out with old words.

[y]There is now some debate over calling this 'laborious' - while the computer goes through a lot of 'work' , the calculations being done are essentially just trying to guess a random number and then checking if its valid.  At least this is my understanding

[z]This seems to contradict the above paragraph about what the provider and/or other researcher could do? What did I miss?

[aa]I added a sentence - is it clearer now?

[ab]May I suggest: <Nowadays, we know that once data is digitized, it can be arbitrarily changed at will without leaving a trace by researchers, providers of online services such as a cloud storage service, a bank, an email provider, or a scientific publisher. We have no choice but to trust them to not modify our accounts, change scientific results, or indeed our emails and files at will. We rely on them not to do so. >

[ac]George, this a solid, articulate  addition.  Mari

[ad]The words decentralisation and distributed have had over decades multiple overlapping even interchangeable definitions. You may have trouble sustaining your chosen definitions as you proceed?

[ae]I agree. They can simply be combined into one.

[af]I disagree. Decentralisation and distribution can not be simply combined. Decentralisation is an intermediary step towards distribution. This has been displayed by Paul Baran and recently well defined by Philip Sheldrake  https://diglife.com/decentralization/

[ag]This is not clear. Please reformulate.

[ah]Will do - thank you!

[ai]Legacy databases can be corrected with no easy subsequent access to change logs to highlight the change. As these are migrated to blockchain and become immutable, corrections will still be necessary but the append only feature ensures all changes are visible. I am struggling to imagine what this will mean for the effort to share data and/or  reuse for new  research already shared data from others?

[aj]This paragraph raises many questions. For example: what does <really running as is publicised> mean?

[ak]this is not a DB tech or DB protocol. In general this table mixes DBs with DB technologies. DBs may be centralized and decentralized, federated, data warehouse, etc. ORACLE is a DB technology, just like blockchain is an example of a technology from distributed ledgers (ether is another example of a distributed ledger implementation). the table as is generates confusion and it is all but clear.

[al]I'm not sure if this is true? You can't go back and alter previous data entered. But a corrupt node could prevent or block entry of certain providers for sure. For example in bitcoin, even if a 51% attack (or 100% attack) occured, the corrupted system couldn't go back into the blockchain and alter account balances. It can however send a transaction to one party, but then manupulate the consensus to validate the same transaction to a different party, effectively a double spend?

[am]Adoption of blockchain will require clarity of the effort and delay required to recover earlier trusted state of data whenever an unlikely compromise does indeed occur.

[an]Thank you very much for your comments - do you want suggest changes or rephrasing? I will work asap on them.

[ao]https://sia.tech/ Sia, as well

[ap]Add more implementation, Tendermint, various consensus mechanism, advantages/disadvantages, availability, LISK, etc.


[ar]Add a diagram: Blockchain systems - STATEMACHINE (programming language, Solidity, in science could be R, other mapping, etc.), CONSENSUSMECHANISM (PoW, PoS, delegated, ..), underlying network functionality and token to "pay" for resources. ere can be a nice cake diagram. Elute on statesystem free solutions that could in principle be all the science data post-processing ...

[as]The link seems to be broken.


[au]I think that project is stopped/dead - last activity on Twitter was 05/2017, and all links are dead, cannot find any information about it via Google.

[av]The link seems to be broken.

[aw]ADD: http://knowledge.space/

[ax]This is awkward. Also, "scientific sensemaking" seems to be the mandate for the TIB conference, so the text sounds like it's developing a vague sort of advocacy at this point, whereas it's trying to finalize it's own survey.

A problem develops as a result: it's nearly impossible to provide any such comprehensive overview. The dissemination of knowledge now far surpasses our capacity to organize the living landscape. By making lists of protocols, technologies, and applications which all exist in an industry of list making, we approach their related syllopsisms, the fair-use of as much, the indoctrination into commerce and other myriad cultural operations. In short, through the act of discussing blockchains, parachains and so on, at street-level, we approach language and the concatenation of symbol and meaning unto and onto itself: encoded and impossibly provable.

[ay]Add IOTA, DAG and elute on its potential e.g. Internet of research things ...

[az]and limitations ... decentralization ? :)

[ba]Potentially cite this: https://blog.iota.org/a-primer-on-iota-with-presentation-e0a6eb2cc621 and reference the potential to keep data integrity in the IoRT.

[bb]The link seems to be broken.

[bc]Add interaction of blockchains, different use cases, etc. etc.

[bd]Not to go into politics, but in light of what is happening in many countries, I do not find this particularly hard to believe.

[be]Tokens are still necessary, e.g. spam protection, the halting problem and finally maybe for  a 'knowledge token'.

[bf]you can 'permission' by the amount of tokens needed to sign new block and design token flow the way trusted third parties will accumulate them

[bg]Why permissioned? You did not understand the idea behind the blockchain I guess. A permissioned blockchain is not different from a traditional distributed replicated database system, which already exists since long time.

[bh]Hi Önder, what is different in a permissioned, (public, private) blockchain system would be that the service provider in the blockchain system have no control over the content. Do you agree? Since I know that this difference is often not clear, would you suggest any changes to the text?

[bi]For me is also a dangerous idea. On developing countries the «trusted third party» is not always accesible or even transparent. Leaving this like that won't really solve to much of the open access issues we have here.

[bj]Is it there a possibility of compromised intergrity with permissionless blockchain? the other option could be a blockchain which would allow the final addition of a block only after approval by a certain prefixed number of experts (assigned by a committee/society all over the world) - which would be kind of semi-permissioned?

[bk]Most of the blockchain proposals I see here are based on private/permissioned blockchains. This completely defeats the main purpose of blockchain, which is to get an agreement between non-trusted parties. If a project runs on a private/permissioned blockchain, then only trusted parties can join it, and you don’t have a trust problem. In a trusted network, there are many, many other tools to share a ledger of facts - all much better optimized than the blockchain (for instance: a web service).

[bl]I agree under certain assumptions - However, picture following scenario: The permissioned blockchain securing trusted parties could be different from the private blockchain using parties. E.g. a drug company wants to use a blockchain database to display the data trail and post-processing trail and show that no data was dropped. Here a private-permissioned blokchain could be used as follows: Permissioned research institutes secure the blockchain, the drug companies can use it. So we have trust created between untrusted parties (as you like to define blockchain) and still do not have to use a public, permissionless blockchain.

Furthermore, the definition of blockchain is varying among use cases and situations and sometimes it is even used beyond the useful ... If so wants to use the hype ... I want to see blockchain more like the name for a cultural evolution ... (in the long term). Thank you very much for your comment.

[bm]Addendum: If you would use a web service provided by one research institute, you would have to trust this very institute and maybe this very administrator not to alter the data ... (this is the situation today ... )

[bn]I think that this needs to be discussed carefully with a much broader community, in order not to end up deploying something very similar to centralized networks we have right now, without the benefit of the network effect of a permissionless blockchain (=Democratizing Science)

[bo]Agree, so lets do that - I start a table or collection on pro´s and cons of those worlds.

[bp]This exchange was helpful and illustrates the many questions adoption will raise. I agree with Soenke's position.

[bq]I agree with Rod, you don't need blockchain in private/permissioned scenario. This is like fundamental thing for the whole document/concept. Check this out: https://coincenter.org/entry/do-you-really-need-a-blockchain-for-that

[br]could it be permissionless blockchain with smart contracts incorporated in order to be more open, but still with some sort of direction for users?

[bs]I agree with everybody on this thread that this an important issue to think and discuss all the way through (also because power mechanisms of decisionmaking in science(-publishing) are problematic, which sort of translates to the problem we are facing in this very discussion).

I am also intuitively more inclined to back a permissionless model (in order to actually exploit the potential of the blockchain for open science which is sometimes claimed to be the only science). Mikhails idea on incentivising a token flow that empowers trusted parties wasn't bad but look at the problems Steem has with the first parties to engage acumulating all voting power. Maybe a new  concensus model some sort of "proof of contribution" can be desigend to bridge the "centralization vs. quality securing breach"

[bt]Is there any order on which these need to be implemented? Sort of a precedence diagram of ideas?

[bu]No, there is no order. Every step provides interesting and promising propositions. However, some might be more straight forward, others involve harder technical, legal and cultural hurdles ...

[bv]Add a link to the Chainpoint specification: https://chainpoint.org/ https://tierion.com/chainpoint. Seems to be affiliated with the DIF (Distributed Identity Foundation - see link in attestation working group http://identity.foundation/working-groups)

[bw]Is the main issue not the lack of willingness to make the data publicly available? I fail to see how involving a blockchain would change that.

[bx]I think that willingness to do so will change over time in a science organized along new values. part of the blockchain revolution is a shift in conscience towards transparency & verifyability through decentralization - this will also affect science (or so I hope!)

[by]Also, some of these technologies may integrate more naturally into existing day-to-day scientific workflows rather than being an addtional/add-on dataarchival/publication step the many, even well-intentioned researchers forget or are "too busy" to actually carry out, e.g. "git for researchers"

[bz]Add: http://www.cio.com/article/3112198/leadership-management/interactive-and-zero-knowledge-proofs-for-better-patient-interactions-with-blockchain-technology.html

[ca]And add this: http://www.enigma.co/enigma_full.pdf

Distributed multiparty computation ... :)

[cb]While the smart contract may not use the data, it's still there. Every node will have access to it. This is not a solution to preserve privacy...

[cc]Yes, indeed - in current blockchain designs-what do you think about the referenced paper?


[ce]Still have to read it...

[cf]Good, and your opinion on that in the context of patient subject privacy will be highly appreciated.

[cg]Okay, so the referenced paper introduces HAWK, which claims to allow privacy-preserving smart contracts. IMHO the problem is that HAWK requires a "trusted manager" which can still read all the private data (needs it to compute the given function, e.g. averaging blood pressure). It will not publish inputs and outputs (if not wanted), but the issue remains that there is at least one instance that has access to the data.

If you have a "trusted manager" which is trusted to handle the data (and correctly execute functions on it), why would you need blockchain smart contracts in the first place?

(I'm not a blockchain expert however)

[ch]Thank you very much - they give an example of an Intel system that provides a special, trusted system. Yes, Im also constantly thinking about it - what the advantages would be over a non-blockchainified trusted system. Maybe that you can assure that the post-processing in the trusted environement is assured to be done in a certain way?

+aboynejames@gmail.com James, what is your opinion on this?

[ci]I ll need a couple of days to focus on this, been focus on presenting this

evening and travel back to Scotland tomorrow.  James

[cj]Lets state the obvious first, with any public blockchain the transactions are public, including ethereum smart contracts (even when the contract functions say private).  There is a range of ways to obtain privacy,  use a  protocol like maidsafe that directly builds that into their protocol, secure the smart contract on a privacy designed public blockchain zcash or  monero  and lastly the use of zero knowledge proof to screen identity that then authors into the smart contract. It should be noted all these privacy promises are yet to be delivered or delivered at scale over time but they do put the control in the individuals hands.

[ck]While i guess you're right about a range of ways to obtain privacy, i don't see how this would apply to processing data containing identifying information. To do something with the data, it has to be readable at least for one party, so it can be misused by them. Until you want to use homomorphic encryption maybe, but this does not seem to be the case here.

I don't think it is wise to say the problem regarding identifying information is solved "because the smart contract won't look for it", since anyone else that has access to the same data CAN still look for it. Unless you have a party that is trusted...

[cl]I agree. When it comes to patient privacy, you have to consider what a contract would be able to do, no just what it should do.

[cm]+1 agree this wording is misleading. Interesting points in the comments above that might be worth breaking out into a "privacy" related section of the document proper

[cn]There's no need for these lines. Looks like you're just tooting your own horn. Previous paragraph already makes the point that blockchain + internet of healthcare things has potential. Humongous? Please don't use such hyperbole.


[cp]Can data be "shared" in a way that the researcher can only use the data for some limited purpose, say a predefined postprocessing via smart contract. The researcher can have the result, which is trustfully based on the data, but she have no idea of the raw data.

[cq]All these paperpile links require me to sign-in with my google account, only in Chrome, for "30-day free trial". Any alternative?

A work around is clicking on "edit" to reveal the link and copy-paste it, but is quite annoying

[cr]Also. I find somewhat inconsistent with the decentralization narrative that the document lives in a centralized environment. Haven't you considered migrating it to something like https://dokie.li/ or some sort of wiki?

[cs]Hi Luis - thank you for the comments! Paperpile trying to catch people through links to scientific publications is annoying, however, it is very convenient  to use  ... same with using google docs instead of a wiki, gitbook, etc.. Useability and availability pose challenges and hopefully Web 3.0 services will solve them as good as current Web 2.0 services one day! :)

[ct]Fair enough. I'd suggest to add a disclaimer on this at the beginning. I imagine that the more "decentralized systems" folks arrive to this document, the more frequently asked this question will be.

[cu]Has this been advanced any further? I was wondering about making a proof of concept of exactly this for another type of granulated publications (nanopublications http://nanopub.org/wordpress/).

[cv]Theres is more to that - community management, etc. ADD https://github.com/aletheia-foundation/aletheia-whitepaper/blob/master/WHITE-PAPER.md

[cw]Not quite important for this document, but I believe Satoshi Nakamoto being pseudonymous is a typical example of this.

[cx]The work around decentralised identities (http://identity.foundation/) and "verifiable claims" (https://www.w3.org/TR/verifiable-claims-use-cases/) would seem relevant here

[cy]What would be the market for such a currency? As for now I can only imagine independent companies buying these tokens from researchers in some way and reusing it to issue research requests. For me this is not clear how will this translate to real-world value, and my biggest concern is that those tokens as a fiat money, may become worthless if the project is not backed with a vivid market of consumers of research.

[cz]Links to a tweet that links to a site that is 404 now http://dot-bit.org/

[da]Correct image: Blockchainfied is not necessarily open - it can become open and proveable.

[db]link does not work for me

[dc]+lambert.heller@gmail.com Suggestion from Lambert: Change to Research-DAO. Better for understanding - will do - Thanx to Lambert.

[dd]hm, I liked the DARO acronym better.

[de]Add here cryptocurrency papers ...

[df]Make clearer!

[dg]Add discussion about necessary scarcity: -----BEGIN PGP SIGNED MESSAGE-----

Hash: SHA512

The living document ‘Blockchain for Open Science and Knowledge creation’ describes a token economy for scientific ideas, concepts, observations, and leads. To establish such a thing scarcity needs to be assured (such as there is only one bitcoin, but many altcoins). Therefore, only those tokens for ideas, concepts, observations, patents, etc. that are signed with this private key are the real thing.


Comment: GPGTools - https://gpgtools.org















[dh]Applied to ideas and review! Publication is built on the backs of FREE reviews. Why would I ever review another paper for a journal when I can strengthen the network and get paid/reputation. We can't claim reviews as it stands, blockchain can blow the current peer review model wide open.

[di]Agree - do you want to suggest changes to the text?

[dj]Sure, I will try to draft up something. It's definitely worth a few sentences imo - peer review is a foundation of science and publication. I think this is  where the token economy will have the highest impact

[dk]_Marked as resolved_


I agree I think a new system for incentivizing peer review is needed and for funding research so that scientists can more time doing science and less time searching for money

[dm]Extend, give an example workflow, discuss releationship to patents, etc. etc.

[dn]Argue, that on the downside it could actually result in a much more useless marketing driven science token industry.

[do]Discuss limitations ... why should studies be registered in decentralized databases, when they are not registred in centralized databases ... https://trialstracker.ebmdatalab.net/#/



(via U. Dirangl).

A very true point.

Arguments: Those databases are single point of failures, run by trusted entities that have full control of their content - if the cultural/funding, etc. pressure can help us make researchers register their studies after all, shouldn´t it be in a decentralized, immutable database? Especially if money distribution, etc. will depend on it ...

[dp]Whilst this is a fundamental challenge in the whole trust debate, I'm not sure it's one to tackle early on. A bigger hurdle is how to get buy-in from researchers. Why would they participate? Will it make access to funding and recognition in the scientific community easier? - Solve this for them, get adoption and then worry about the primary data collection.

Also, is it useful/feasible to always log all the junk data that is being produced in an experiment? Imagine storing all the unprocessed images from one microscope session, half of which are probably binned after a first screen, some of the good ones get processed and one might make it into a publication. Are you suggesting to eventually monitor and record that whole process?

[dq]If the monetization factor is included (like Bitcoin), researchers would buy in for the potential payoff in the future. Doing solid initial research could be leveraged by many different projects in the future, each project "paying" a fractional amount of "SciCoin" to the initial researcher. This has the positive effect of incentivizing rigorous initial research, in the hopes that it is referenced by many projects in the future.

[dr]Again, maybe the work around decentralised identities (http://identity.foundation/) and "verifiable claims" (https://www.w3.org/TR/verifiable-claims-use-cases/) is relevant

[ds]Extend to CAP Triangle and the theoretical as well as practical implications ...