Earning from Authorship based on impact/use

JasperJasper Eindhoven, the NetherlandsPosts: 514Member ✭✭✭
This essentially came up in this thread https://forum.ethereum.org/discussion/298/trinitycoin-creating-value-through-values, and i decided the subtopic of rewarding authorship was worth a separate thread.

@Donal posted this document there https://us.v-cdn.net/5021640/uploads/FileUpload/63/55e81a90b81caed6e4ab1e65c4ba29.docx which is interesting, but i dont see how it gets money in the system in the first place. I mean, the whole problem of releasing information is that people can just copy it into any other system, and never get anything back to the user.

What I see online, is that money is made by creating platforms like youtube, twitter, etcetera, and people go to the platforms out of convenience and habit. The users are then shown advertising, which makes the money. Some of the money indeed does go authors, but it is not very effective.

We could make a decentralized platform, and do that same trick. Secondly, a system could additionally take donations/tips, not only directly(the content could display adresses to donate to), but also possibly utilizing the system itself. I.e. if you use the system to donate, it automatically sends downstream aswel. Or possibly you could pay to get rid of advertising. Weakness is that it is trivial to fork the software to remove the advertising, leaving only donations. However, reminding back that it is about a 'platform of convenience'; if the ads are not obnoxious, people might not bother.

Still, other ways to get money into the system are appreciated. One is to just use coin creation as a supplement. However, the marketing sets its value because people that want exposure feel the viewers of particular content are real. Coin creation does not have that aspect.(the 'identify real people' problem has been discussed in other threads, and itself off topic here, however it may be able to reinforce this system)

Identifying authorship is a requirement. For a document, someone can claim H(D) and then publish D, but that does not prevent minor variations being published. Since the *first* would be the original author, if we can identify derivatives and the extent of the changes in derivatives, minor variations wont pay, only variations that really do something.

I think essentially humans will have to judges to figure out if things are derivatives of each other. Once decentralized, trustless consensus on that is solved, the biggest problem is creating the platform and then popularizing it..

Comments

  • SatCaSatCa Posts: 29Member
    @Jasper : Yes, humans will have to judge, these issues. I do not see computing saving us any work on that. I understand a "First Information Reporting" protocol could be used on a central platform that to verify 1st Authorship.

    It'll require awareness and transparency. Measures to ensure people serve and not blockade authors need to be in-place too. This is why I was talking about a good team of bankers.

    A good banking team can help evaluate, IP that is highly valuable. The conduct of the banker may be regulated via a whoofie like currency. Or honestCoin, going to release when enough Bullish crypto creators have tried and failed ;)
  • DonalDonal Posts: 11Member
    my system generates ad revenue. thats not the final draft of my paper.
  • JasperJasper Eindhoven, the NetherlandsPosts: 514Member ✭✭✭
    @Donal sorry if i misrepresented the paper. I am such a clut.. Perhaps i should just have laid out the last two paragraphs about identifying authorship. That was a problem i wanted to lay out there. But probably should have refrained.
  • maciejolpinskimaciejolpinski Posts: 11Member
    this is exactly the problem that 'Let's talk Bitcoin' community is trying to solve right now with the LTBCoin concept. "How to distribute tokens among the content community members based on their valuable contributions to the community?"

    It raises tons of fascinating questions such a 'how to measure content's value' , 'what is the relationship between creators and curators in generating value' , 'can value be measured objectively?' , 'how content's value changes in its lifecycle i.e. is Gangnam style still as valuable as it was when it had only 10k views'?

    In one of the threads, I've proposed a very rough sketch of a potential 'ContentCoin' system which allows users to invest their tokens in a content piece through a 'Proof-of-Tip/'Proof-of-Burn' scheme and get rewarded when others invest after them.

    The system is designed to incentivise users to predict the future, subjective value of various pieces of content.

    If you're interested you can read the whole thread here:
    http://ltbcoin.com/discussion/13/idea-to-encouraging-tipping

    Would love to connect with everyone who's working on solving similar problems.
  • JasperJasper Eindhoven, the NetherlandsPosts: 514Member ✭✭✭
    edited March 2014
    Cant seem to get it into my thick skull today.. Basically each website has its own coin, and the value of the coin is the ability to push things higher in the way of attention? Each website is also centralized, and might have moderators/only accept particular articles?

    So website creators are incentivized to keep the coin at a good value, by being effective warding off spam, for instance. Authors want to live on good websites? I dont think we have carrots/sticks for consumers directly, other than convenience.

    The 'early tips pay' idea i get instantly. Might try to do make a script playthingy for that. little writeup. (maybe cute enough for cll-sims/examples.. edit: i am not claiming it or something, just wrote for this discussion, but then thought it was a digression) The advantage of using the amount is that you can also profit from finding stuff that was underappreciated.
    Post edited by Jasper on
  • maciejolpinskimaciejolpinski Posts: 11Member
    edited March 2014
    Thanks @Jasper for taking the time to dig deeper into the topic. Indeed, my ramblings in the ltbcoin thread could be difficult to digest (they were aimed at ltbcoin community which is familiar with the 'contentcoin' topic + it was a late night, 'brain dump' type of a post on my side ).

    So let me try to clarify these issues again, in a much (hopefully) clearer way.

    Let's start from the beginning:

    Ranking the importance & value of content on the web has been historically an extremely difficult and centralised task.

    With Web 1.0 'links' were the signals of value and Google emerged with their PageRank model - 'the more inbound links you get , the more important you are'. You are therefore rewarded with more attention in the form of higher search rankings. However, it's easy to create links and fake sites for almost no cost therefore we got link spam, 'keyword-rich sites' etc. Of course, Google evolved and now their algorithm is way more complex. But essentially what they operate is a centralised, closed ledger of 'key-value' pairs with a PageRank value attached to each piece of content on the web.

    The application built on top of that ledger is Google Search.

    With Web 2.0 'likes / followers' became the signals of value and the social web emerged with the dominant Facebook. With Facebook's OpenGraph every document on the web could be defined not only by their link structure but also their 'like' structure. And similarly to links in the 'Web 1.0' era, getting more likes/shares/followers meant that you get more attention in the form of increased traffic. Because the FB's Edgerank protocol favors content that people like,share,comment on. And like Google , FB is a centralised, closed ledger of 'key-value' pairs with a certain value attached to each piece of content in their database (it's more complicated than that but for the sake of simplicity let's put it that way).

    So both Google and Facebook sit on top of the massive flows of 'value signals' created by users and extract enormous benefits from keeping their 'value protocols' secret. They parse these proprietary, centralised ledgers with massive computation and AI technologies to evaluate content's value. I don't need to explain how profitable is that :)

    Combine that with an advertising based funding model and you get the web as we know it today. Full of spammy headlines, link baits, fake profiles, sensational content, essentially a tabloidisation of everything.

    What I'm proposing is a next layer - let's call it a Web 3.0 or the 'value web' or the 'subjective layer'. I believe that using the decentralised ledger technology and digital tokens, we can design incentive schemes to let people subjectively rank value of content and apply subjective human judgement to problems that were so far only solved mechanistically and computationally.

    Creators and curators would be rewarded if they can successfully predict future popularity of certain content or application types. I'm writing 'applications' because the mechanisms could be applied to forums, websites, app stores and create incentives for curators to promote their investments further. They would be no longers readers and users but 'micro-investors' or 'micro-shareholders'.

    This popularity wouldn't be just measured by links or likes but by future investments by people who believe certain content would get even more popularity.

    So it is economically valuable to you to invest in stuff that you believe others might find valuable. Furthermore, you're incentivised to promote that stuff further but only to groups of people who are most likely to get the biggest subjective value out of the content.

    Metaphorically speaking, each user becomes a node on the network which creates, curates and evaluates value of each signal that passes through them. And they are inclined to forward this signal to other parts of the network for which this signal is most valuable. Rinse and repeat until the nodes no longer can assume that they can find any additional nodes that will benefit from the signal.

    As each nodes 'attention' is a scarce resource - people have only so much of it during the day, creators and curators would have to commit a certain amount of scarce token in a 'proof-of-tip' or better 'proof-of-burn' process.

    With the new token issuance cycle they would be rewarded with new tokens proportionally to the accuracy of their predictions i.e more people invested after them.

    What I'm proposing then is a hybrid 'Proof-Of-Burn/Tip' model where nodes/users are rewarded for accurate predictions of future value. So instead of mining as a purely computational effort you get mining as a inherently human effort as each evaluation would take some of brainpower.

    Combined with objective layers of links and likes, this would give us a powerful method for ranking content online and transfer the value back to the edges of the network.

    'Early tipping' is just one approach and thanks so much for taking the time to put it into a mathematical formula. I love how things can develop organically when people bounce ideas off each other.

    To not overcomplicate things I'll end here but if anyone is interested we can continue to discuss this topic.

    I wrote a blog post some time ago on this topic (less in-depth but could be interesting to you): http://www.maciejolpinski.com/blog/why-the-next-facebook-will-be-owned-by-you-hint-its-because-of-bitcoin/

    Let me know if this makes more sense and let's start working on it :)))




  • JasperJasper Eindhoven, the NetherlandsPosts: 514Member ✭✭✭
    That did help a lot. Yes, lets try figure this out. Note that i am interested in making 'no-go' theorems, to try narrow down the possibilities. Dont be demotivated, i am just trying to avoid spending time in directions that dont work.

    About the 'mining' thing, i have no idea how you would secure a blockchain with it. That said, blockchains are essentially an accomplished fact, and it is possible to can issue new coins without it actually being mining.

    Note that the formula is just one 'family' of ways to make 'early tipping pays' work. I think time dependence can also be a good thing. (but the weight factor on the timestamp is actually irrelevant :p)

    More important point about 'early tipping pays'; the connection to the article is dependent on non-investors; people donating because they appreciate the article. Without that, there is essentially no connection to the article, and it would essentially be a sort-of pyramid-scheme-with-foreknowledge sort of thing. It is true in general that such a connection needs to exist.

    I feel that if entity A has more power than B, and if both are in the same deterministic situation and act rationally, A will always keep it. Essentially by definition. This is the reason why for instance i dont at all think that you can somehow make any only-stake-sigil-proof system egalitarian by using the fact that little(less ether) people use more small transactions than a big powerfull one with more ether. Any system where the behavior of the little people advantages them will cause the big powerfull one to Sigil his way into the same advantages. Afaik the best you can do is try decrease the disadvantage of the small users, but never remove it.

    I say this because we have a only-stake-sigil-proof system here. If input data like 'likes' etcetera are used, if it is as simple as two parties, the more powerful one keeps that position. What i mean, is that it only keeps working as long as enough 'power' is in the hands of people that make it keep working.

    One way around that is, is to have the people-to-system relation such that the value of the coins in the system is dependent on the system working properly. If you have multiple of these, even if one party has a lot of stake, it is in his interest to use it wisely.(though you might still wonder how much leeway those people have)

    I also expect there are strong limits on how you can do issuance due to interactions between users. I expect that it can soon hit problems of abuse.
  • JasperJasper Eindhoven, the NetherlandsPosts: 514Member ✭✭✭
    Now i think about it again, people could just try guess which things will be popular, and just free-ride on the fact that it will be. That isnt necesarily bad if the guess wasnt very sure, because then it is supporting artists that are at that point obscure.

    But if the guess is basicly a sure thing, it isnt so good imo. It would be a pointless race to be on time.(much worse in my amount-based example!) If success is a sure thing, you dont need to pay anyone for promoting.
  • maciejolpinskimaciejolpinski Posts: 11Member
    Hey @Jasper,

    These are great comments and input - I was sidetracked a bit due to work but still thinking a lot about these topics. :)

    In regard to 'mining' and securing the blockchain - you're spot on, it would be difficult to 'rewrite history' with every block. That's why I'm considering that a mesh of Ethereum contracts with internal tipping/betting subcurrency might be the way to go. Don't ask me about the details though, still getting my head around the Ethereum's CLL and moving the building blocks in my head - a decent Contract Sandbox or GUI Simulator would be awesome to quickly test some hypotheses. That'll come soon I guess.

    Speaking of of 'guessing' and the 'pyramid' scheme - the whole point is to distribute value between creators and curators plus create incentives for people to devote their time & energy to these tasks.

    I totally agree that it might resemble a 'pyramid' scheme in some sense and people will act based on greed, guesswork, speculation etc. But as long as it injects value into a content ecosystem - then I'm ok with that. Similarly the entire social media industry is built upon people's vanity with millions of people optimising for 'likes' etc.

    In terms of the 'race to be on time' , I agree that it could create too much rush. But 'early investors' should be rewarded and a sense of urgency should be embedded in the model in some way. Otherwise there's no incentive to act.

    Anyway, I'll post soon again with more thoughts.


    PS. My impression is that some sort of a 'contract simulator' for these type of projects would be extremely useful.

    And I'm not only talking about a real time compiler but also a GUI that could help us explore & visualise certain patterns over time, potentially with some AIs that would be programmed to maximise their profits within a specified set of contracts.

    It's great to exchange thoughts but I feel like I'm trying to imagine and explain some complex spreadsheet formulas with writing :) and I'd rather show you some animated charts with sample data that would speak for itself.


  • maciejolpinskimaciejolpinski Posts: 11Member
    Ah and one more thing:

    I've found this which in some aspects might be relevant to our case (prediction markets, users try to predict future outcomes with weighted votes etc).

    I haven't fully digested it yet but pasting it here for others.

    TruthCoin - a whitepaper for decentralised prediction markets based on Bitcoin.
    https://bitcointalk.org/index.php?topic=475054.0

    Abstract:
    TruthCoin - a whitepaper for decentralised prediction markets based on Bitcoin.

    Abstract.

    "Where Bitcoin allows for the decentralized exchange of value, this paper addresses the decentralized creation and administration of Prediction Markets (PMs). An alternative proof-of-work blockchain collects information on the creation and state of PMs, with the winning state of a market determined by a modified weighted-vote.

    An incentive mechanism attempts to guarantee a) that all voters vote honestly, and b) that PM-creators act as entrepreneurs, bearing the economic costs and benefits of the PMs they create. Bitcoin users can create PMs on any subject, or trade anonymously within any PM, and all PMs enjoy low fees and infinite market liquidity through a LMSR market maker.

    Scalability and customizability can be achieved via ‘branching’ (controlled-fork). The paper closes with a discussion of implementation details."


  • JasperJasper Eindhoven, the NetherlandsPosts: 514Member ✭✭✭
    edited April 2014
    @maciejolpinski: cll-sim is a decent simulator, without any visualization. Python≠.cll(/.hll) but probably close enough to easily convert later. Ethereum-simulator doesnt have as easy testing and only has a javascript-like analog, but it has some visualization. Edit: and there is the test-net of course, havent ever really tried testing contracts there yet.
  • maciejolpinskimaciejolpinski Posts: 11Member
    @Jasper thanks, I'm familiar with the first one but not the second ( haven't tried the testnet either.)
  • JasperJasper Eindhoven, the NetherlandsPosts: 514Member ✭✭✭
    The point of the first post i made here was that 'secure computers' + 'effective human judge system that can judge small facts' implies you can use humans to judge if stuff is 'trivial variation'. Link you gave seems to be about 'large' facts though. (havent read the truthcoin .pdf, not sure if i will soon)

    I dont know how to do judges for these equality-of-articles facts, if it can be done, it may well be preferable. Not sure what is from me and what is from what i read, but i want to specify it anyway:

    Essentially in this idea the viewers/readers/users police Platforms.(perhaps a precarious base!) They do so by selecting ones that are beneficial to their authors.

    Platforms and authors figure out a good deal, this deal involves setting a parameters. For instance:
    1. the 'derivative-work-ness', that passes a fraction of income downstream.
    2. The level of 'incentive-to-promote' with donate-first-pays schemes. But maybe also other schemes of similar nature.
    3. How to deal with on-article advertising. This may sound bad, for something like blog entries.. and there it is. But it could be maps in FPS games, which have advertising signs, posters in the environment. It could even be populated with computer monitors, cups with sides where it again could be somewhat natural to put ads. It may also include leaving them out, using A/B testing or having users vote/choose.
    4. What to do with income from objects(articles) that go into the article, like those 3d models again.
    Some of those are a little too far-forward looking, i say ditch 3,4 for now. These deals staying good for authors is essentially up to viewers. Authors may, and probably should still do things on the side like releasing H(Article) to prove they had the file first. Perhaps there should be limits on how much the caretakers get paid.

    Platforms and advertisers also form deals, very similar/simple, just coin for ads. Advertisers simply monitor platforms to see if the ads are dutifully placed. They also try estimate how many people actually go there. (Of course, 'proof of view' might still be interesting to consider)

    Advertising and donations are sent to authors and platform caretakes according to the rules. Donations may relatively send a larger fraction to authors than platforms, as viewers aim them and may choose a direct send to authors.

    Finally perhaps there should be more ways to fight back if a platform goes bad. To ensure evidence can easily be found, if people pay the fees for storage and computation involved they can link claims of misdeeds by the platform, including H(Article) proof-of-posession of files.(before the platform ever published them, authors may publish them in WIP phase in case they accidentally leak) Additionally, perhaps some kind of voting power given to authors to the end of disbanding, changing caretakers or splitting up the platform.
    Kindah missing from this is how the platform itself works. Preferably the link to the platform is identified by the contract adress,(with domain name via namecoin) and each article within it is identified inside it identifies logically with hashes aswel. The browser of the platform would figure out the commands to send to the in case of donation. Here is where 'proof of view' could potentially work aswel.

    Hell, the browser could indicate if anyone is trying to discredit a platform. Giving him the explanation of the disgruntled, the claimed stolen articles, and the article that is claimed that was stolen from.(which has to be older, as provable with H(Article)) Of course that may be missing and it may be about the quality of the deals between caretakers and authors.

    All ideas for organizations go.. You could allow for some investments and dividents.(good for getting it going initially, dont think so otherwise.) Or you could add co-operative style things, or egalitarize the incomes of members.(should do that, but not too far, i reckon) Could try 'organizations budding', or try add some growth modes. Growth of number of authors is a matter of accepting incoming articles.
Sign In or Register to comment.