- From: Adam Sobieski <adamsobieski@hotmail.com>
- Date: Mon, 18 Dec 2023 11:00:00 +0000
- To: "public-swicg@w3.org" <public-swicg@w3.org>
- Message-ID: <PH8P223MB0675A8D51540A8157AE1D739C591A@PH8P223MB0675.NAMP223.PROD.OUTLOOK.COM>
Introduction Economic incentivization systems were broached during a previous discussion about advancing decentralized moderation, social media, and AI [1]. I would like to follow up and expand upon these ideas which I hope are of some continuing interest to the group. Economic incentivization systems can contribute to the mitigation of malicious and mischievous acts, spam, bots, misinformation, and disinformation on social media platforms. Models thus far considered include that end-users would: (1) pay to post, (2) pay for any boosts or prioritizations pertaining to content distribution, (3) receive monies back depending on the value added by their content. For the average post, monies paid to end-users should equal (or even slightly exceed) their initial investments. Better-than-average posts should result in profits for end-users. Malicious or mischievous posts should result in effective fees or fines. Revenue-sharing Models The distribution of monies to end-users, in a manner rewarding valuable user content, can be viewed as a form of revenue sharing with, beyond select influencers, broader or entire populations of end-users. Economic incentivization system models can distinguish social-media services from one another while enticing end-users to migrate to those services which are more financially advantageous for them to make use of. Measuring the Value of User Content Applause End-users could be allocated a number of points, per intervals of time, which could be distributed over other users' content. If an end-user applauds often, the value of each instance of their applause would be diluted. If an end-user applauds less frequently, then more weight would be assigned for each occurrence. Complaints End-users should be allowed to complain about content, e.g., spam advertisements, while indicating which service, group, or page rules were violated. When so doing, end-users would be signaling that content should be reviewed by a moderator or administrator. It should cost end-users money to complain. In the event that a moderator or administrator confirms the complaint to have been valid, complaining parties should be provided with refunds exceeding their initial investments. Valid complaints, then, would result in profits for complaining end-users. These monies would come from those fines or fees taken from the complained-against parties. Research Questions Predicting the Value of Content The value that posted content adds to a social-media system is seemingly best measured after it is posted and after the system interacts with it. Can the value of content be predicted before it is posted or after some initial interactions with it? Distinguishing Types of End-users Should some user privileges be available for purchase, e.g., digital press passes? Distinguishing Types of Content Should reportage be distinguishable from opinion and editorial content, from humor content, and from other freestyle forms of content? Perhaps end-users could utilize a drop-down menu when posting to indicate content type. With respect to news articles, schema.org currently includes types for reportage, analysis, ask the public, background, opinion, and reviews. Predicting the Veracity of Content Can the veracity of some types of content be predicted before it is posted or as it is interacted with? Incentivizing Speed While economic incentivization system models can be devised such that it pays to post true and valuable content, could it also pay to post such content first? Predicting the Results of Moderation Can the results of some moderation events be algorithmically predicted? Risk Management How should risk be estimated for end-users at instances of time? Should end-users' requisite investments to post content and their related rates of returns be equivalent for all end-users, at all times, or might these vary based on end-users' recent past behavior and algorithmically estimated risk models? Could models of risk enable more efficient algorithms for content moderation? Points versus Money In some points-based systems, end-users receive points which they can redeem for prizes or for entries into drawings for prizes. With respect to social-media services, it is easier for end-users to move between competing services with money instead of points; points might not be portable across services. There are also know-your-customer and other benefits for social-media systems from end-users transacting with money instead of points. Policies and User Agreements End-users would agree to policies pertaining to economic incentivization system models as they entered and joined services, groups, and pages including with respect to investments to post, profits from content value, fines and fees, moderation procedures, elections of moderators, and pertinent appeals processes. Conclusion I hope that these preliminary ideas inspire further exploration into economic incentivization system models with which to advance social media. I am optimistic that standards and recommendations can enable new possibilities in these regards. Thank you. Best regards, Adam Sobieski [1] https://lists.w3.org/Archives/Public/public-swicg/2023Nov/0032.html
Received on Monday, 18 December 2023 11:00:36 UTC