User:Shiftchange/Wikipedia and the Web 3.0

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
Semantic Web, Web 3.0
Preview release
Platform MediaWiki
Type User-generated content

This page constitutes my design specification for a semantic technology, a user-generated content management technology, that best solves the content monetisation problem.[1] For Wikipedia, the objective is to reward from within, instead of allowing payment from outside.

I call this semantic technology the real Semantic Web or Web 3.0. It provides meaning to imminent function, specifically to the process of content curation. It serves to produce an invention, something with endless possibilities, that solves many problems experienced at Wikipedia. It solves the problem of paid editing and error avalanche. It should speed up the overall production rate by bringing new users into an economy we foster. It should also encourage attention towards difficult tasks. More generally, the Semantic Web will create a means to enforce quality control at web sites based on user-generated content. The outcome will be the disintermediation of advertising-centric monopolies in which in authors don't monetize.[2] For decentralized users around the world to be both incentivised and gain full control of their content, a digital token system is required for reward.[3]

The ideological basis for this invention is founded on holistic management and functional finance with some elements of protectionism in the system. If it is successful the Semantic Web may be used to counter mass unemployment due to automation. This invention has the potential to evaluate Gresham's law. It is my invention because I created it by myself. I started the process that led to its origin. I have devised a uniquely new technology from the combination of others. This is called combinatorial evolution.[4] This is not a marketing gimmick and not the development of a virtual economy. It follows human-centered design principles. The concept includes an expansion of the informal sector pursuent to libertarian goals.


No payment for contributions due to disintermediation on the Semantic Web so no trade for work hopefully coming soon to Wikipedia

Wikipedia's editing model has not accounted for the change in value transfer that Bitcoin forces.[5] The distribution of rewards using a blockchain or digital currency to define quality and build value[6] will be a better method than trying to force a machine to acquire knowledge for us. Managing knowledge will always be a social activity. Novel user centered approaches to evaluating semantic knowledge management have been demonstrated to enhance the working environment by improving both the efficiency of the work and the quality of the work.[7] There is no known way to prevent artificial intelligence systems from reflecting the biases found in society.[8]

A reward and a better website provide incentive for new participants

Wikipedia is failing. It lacks integrity. Co-option creates conflict of interests. It is illogical to believe that a declining active editor rate with an increasing number of articles can suffice at a time when internet becomes the predominant media. The volunteer model used by Wikipedia is in terminal decline. This must be addressed because the problem can only worsen. The longer that Wikipedia's failing model stays the same, the worse it will get. Just like Youtube has to determine what is pornography we will determine what is not knowledge and remove it for the sake of purity. In the case of knowledge, this function will never be automated. Just like the attention is on Twitter and Facebook for fake news and comment shills now, attention will turn to Wikipedia. At some point, as the terminal decline continues, more people will be seeking my solution. By then the basic design of the Semantic Web will be mostly complete.

We are compromised. We are unable to identify bad actors or paid operatives adequately. Trial and error does not generate knowledge.[9] Our contributors need more of a chaperone or a guardian to ensure that our manufacturing process is continuously improving and not degrading. We seek purity but we can't rely on that emerging ad hoc.

Machines will never ascribe meaning therefore what I have outlined here is what will become the Semantic Web. This system aims to create the ultimate knowledge machine through vastly improved collaborative knowledge management. The methodology is to create incentive that will hopefully escalate intelligence through participation. It should facilitate enhanced ability to decipher information from knowledge. In this article we can see that an author for Wired is mistaking knowledge for information.[10] Only humans can acquire knowledge because machines cannot think or understand the same way we do. Meaning will never be truly machine generated. We are therefore obliged to embark upon the creation of the sum of all knowledge by using machines, not by relying on them.


Name Date Dif
Inception 11 September 2016 [1]
Invention 2 October 2017‎ [2]

The project began with elements of artistic inspiration, outsider art, brainstorming and ideas from surrealist automatism, swarm intelligence and pay to play.

The first permutation of my invention was dismissed in late November 2016. In that suggestion, I made the mistake of including the word bitcoin, triggering alarm instead of interest. At this point I knew something was required along the lines of my suggestion, however I had no clear vision of how a design might propagate.

On 1 October 2017, I began discussion on my proposal here at the Village Pump. From this discussion emerged a rough conceptual framework for a new invention. The following day I began development of the Semantic Web by creating what I call the Genesis Edit. After that I continued gathering feedback, elaborating on my proposal and building up the details. I then reached out to Jimmy Wales again. As expected, the community did not receive my advice well. Rather than act as a deterrent I realised I had to be more credible, articulate and give wide consideration to the feedback I had obtained.


Reward for participation, through assessment and trust consensus.

I am developing a mechanism that will create a reward incentive for Wikipedia contributions using an abstraction layer. This convoluted gearing appears rambunctious and neurotic and yet it should be sufficient. It comes with a number of redundancies and countermeasures. When deployed this mechanism is to be in total compliance with Wikipedia policy and guidelines. The system will be robust. Antifragility should not be required. Participation is purely voluntary. Nothing is to be earned, only acknowledged and then rewarded by a system after rating, consensus and approval.

Gifts or rewards are granted on the basis of merit. The only thing a contributor needs to know is that the better they do they work the more likely they will be rewarded with a higher amount and likewise for edit rating. This can only mean one thing - work or edit as hard as possible for a probable gift. There can be no specific payment agreements. No payments will be made to contributors for the specific manufacture of articles, under any circumstances. The currency will be the incentive, the motivation to participate, as it is everywhere. Reward should not be generally tradable or transferable.

Reward or micropayments for work directly on specific pages will not be provided. At first, no-one should expect a reward for a single edit because not every edit will even be rated. The ability to select only participating contributions will assist in making assessments. Contributions are to be rated according to their value. Value is assessed by how well does the contribution add to the sum of all human knowledge. Reward allocations will create consensus on both knowledge and reputation with a trust metric. The less consensus that is generated by edit rating then the less reward is created. The focus is on merit not on reputation. Negative reward should immediately exclude participation so that initial contributions will be highly incentivised. Reputation is only visible during and considered in tertiary assessment. Reputation should not directly affect reward, only assessment. The reward committee should compulsorily consist of the top 10 reward earners and act as a final system-wide review. This should act as a guarantee that value has been added and transferred. The wider community will not want the reward approved while the top earners will be the most eager to prove and receive the reward for their meritorious work.

All rewards will remain in the ownership of the WMF until dispersal. No reward can be redeemed arbitrarily. Prior knowledge of the actual amount of reward should be an indication only. The dispersal time should not be known to contributors. Dispersal should occur chaotically so that contributors may enjoy a gift (no matter how small) at any time. Upon dispersal notification to receiving contributors should also be provided. The highest rewards should be paid to the contributions who bring the most knowledge to Wikipedia. The system may attribute greater reward for any number of reasons determined by the community. These would vary according to the type of information users are generating.

Reward Likelihood Assessment
(out of 10)
Value criteria
-25 Whenever Invalid (bad faith) 0 None
-2 Whenever Poor (good faith) < 5 None or some
+1 Often Good 5 - 7 Some
+10 Sometimes Excellent 7 - 8.5 Majority
+100 Hardly ever Exceptional 8.5 - 10 All

It should not be possible for other users to see the amount reward received by other contributors unless their reward is higher than their own. The reward should start at an extremely low amount. Increasing the amount given as a reward should be formulaic and open. Declining trust metrics may lower the reward. Knowing that the reward increases as the strength of the consensus emerges will increase incentive. Knowing that the reward will eventually increase means that the learning curve for editing will be worth tackling. Until the system is proven to be effective no reward need be dispersed. The digital currency used for dispersal should remain secret until the system gets final approval. There should be an option to postpone reward collection. Trials and bootstrapping may require the preselection of suitable candidates. Likewise an oversight committee should monitor the reward and its allocation as well as any parameters that need adjustment. The committee should also be required to demonstrate how trust has been generated to the community as part of the reward approval process. This is the only political aspect of the Semantic Web. If trust does not emerge from this group, re-initiation may be required. A separate donation fund for this project may be setup to allay concerns over inappropriate funding. A separate digital currency payment provider should be used to avoid reward custodial legislation.

User interface[edit]

Imagine, WikiLoveCoin, a monetized form of WikiLove using a MediaWiki gadget. The front-facing user interface should be presented as a dynamic graphic. No software Installation is required. Ideally this would be positioned in the centre at the top of every page when a participating user is logged in. It will only display the current approximation of the reward amount in coloured numerals. Every order of magnitude that this figure increases corresponds to a reciprocal increase in the graphical intensity. The top ten reward earners will gain extra visual cues in effort to stimulate more of their work that others have recognised as meritorious. Participants may choose to disclose their reward figure.


Assessment is central to the incentive process. There are three kinds of assessment. It begins with the identification of edits made by participating users. Only those with higher rewards are able to conduct assessment at secondary and tertiary levels. Contributions made by participants should be clearly identified by a small symbol on the watchlist, for example. Upon viewing a dif, a scale bar should appear at the top of the page, under the reward graphic. Users should be able to assess contributions by dragging a slider. The scale bar graphic should change colour according to the selected rating. Content assessment ratings will need confirmation before entering the system.

Verification of assessments will be required to identify bad actors. This secondary assessment involves a simple pass/fail decision. A tertiary level of assessment is required to monitor secondary assessment and other flagged events. This could be triggered by an editor whose reward has reached a predetermined negative threshold or other anomalous behaviour. What action the community takes in such as situation requires discussion. This is the only human resource management part of the Semantic Web.


The is a dire need for some fundamental advancement of Wikipedia's editing model. Most the alternatives are profit-based and attempt to create online markets for content or include undesirable advertising. It is exactly these aspects which I have avoided.

See also[edit]


  1. ^ Ryan X Charles discussing Yours (7 February 2017). Yours. Youtube. Event occurs at 3:28. Retrieved 1 January 2015. 
  2. ^ "Flixxo Whitepaper v0.4" (PDF). Flixxo. October 2017. p. 2. Retrieved 28 September 2017. 
  3. ^ Itai Elizur (9 October 2017). "How To Monetize Your Online Video Content on the Blockchain". Small Business Trends. Retrieved 23 October 2017. 
  4. ^ W. Brian Arthur. "Darwinism theory of evolution applied to technology". ComputerWeekly. TechTarget. Retrieved 30 September 2017. 
  5. ^ Jon Matonis (12 July 2012). "Top 10 Bitcoin Statistics". Forbes. JPMorgan Chase & Co. Retrieved 27 October 2017. 
  6. ^ Daan Pepijn (25 October 2017). "Blockchain will make things even harder for blackhat hackers". The Next Web. Retrieved 25 October 2017. 
  7. ^ Davies, John Francis; Marko Grobelnik; Dunja Mladenic (2008). Semantic Knowledge Management: Integrating Ontology Management, Knowledge Discovery, and Human Language Technologies. Springer Science & Business Media. p. 231. ISBN 9783540888451. Retrieved 22 October 2017. 
  8. ^ Louise Matsakis (26 October 2017). "Google Is Sorry its Sentiment Analyzer is Biased". Motherboard. Vice. Retrieved 27 October 2017. 
  9. ^ James Vincent (18 October 2017). "DeepMind's Go-playing AI doesn't need human help to beat us anymore". The Verge. Vox Media. Retrieved 19 October 2017. 
  10. ^ David Weinberger (18 April 2017). "Our Machines Now Have Knowledge We'll Never Understand". Wired. Condé Nast. Retrieved 30 September 2017. 

Further reading[edit]

External links[edit]

  • Earn money for creating good content. A similar project where payment is specified, launched in May 2017
  • Basic Attention Token white paper for a blockchain-based digital advertising and services platform, published May 2017
  • Y’alls A Lightning Network platform for content creation, launched in October 2017
  • Joystream BitTorrent client, streaming and paid seeding, JoyStream v0.5.1 released in October 2017