Is it not funny? Is it not tragic?
All that content being generated by so many users...
And the vast majority of it fades away - to become data.
Upon one level such content achieves life eternal and yet upon another level it is seemingly tricked and cursed into a submerged obscured status of irrelevancy.
I personally consider it to be dreadfully sad. Who knows how many times the same topics have been treated over and over again by different users?
Let us look upon and define the true problem.
The problem is that while there is a LOT of content being generated, it quickly fades to obscurity. The 'value' of that information erodes. The degree of overlap is palpable.
Now, while I could point to StackExchange as an example of what 'could' be done about this (they have a pretty solid search infrastructure that alerts you when there are existing questions to the one that you are about to post) I am here to offer a different solution.
A flower...
If we were to look at posts and comments as 'threads' for which curators can attribute 'weave tags', then it becomes possible to have those pieces of content gravitate towards each other.
This process of gravitation exists over and above the current level of content generation and curation that we presently know.
As bits of information take on tags and these tags are allocated a weight depending on how many persons also tag identically, and their respective reputations, it becomes increasingly possible to re-arrange those bits of data into a cohesive whole.
The process of doing so, I shall call 'pollination'.
By now I think that you may already be starting to get a picture of what I am speaking of.
But to further enhance that mental image, allow me to paint a picture of the final result.
The Flower as Steemit-Compatible Cohesive Data Structure
Imagine, if you will, a flower. It has a number of petals (lets say 6) arranged around a pollen-rich core. Its stalk has leaves and might even have thorns.
The pollen core represents the core subject and its connection to the various bits of data.
The petals represent main areas that are related to the core subject.
The stalk represents time (a chronological representation of those same bits of data).
The thorns represent focus points of controversy, debate or critical alternative perspectives on the core subject matter.
Now in a Steemit context the type of flower may have many more petals - where some petals are bigger than others.
Anti-Sybil Considerations:
(Thank you for suggesting that I delve into such @lukestokes!)
As I approach speaking about reward incentives toward the implementation of this suggestion, I feel that it is important to address the specter of Sybils - fake accounts created for the purpose of undermining the structure of a network - usually for personal gain.
As such I see two reasons 'why' a Sybil attack might be used on Steemit.
Firstly - one could try to use multiple accounts to engage in a chain of circular voting - with the purpose of siphoning off rewards from the pool.
Where there is the prospect of rewards one may also find individuals who try to "game the system" through spawning multiple accounts, perhaps even botting them to perform tasks with little value but which siphon funds from the pool.
Secondly - One could try to use multiple accounts to seek to push particular flavours of information on Steemit while endeavoring to discredit others. In other words to create an artificial opinion-engineering mob. These may or may not also engage in a degree of circular voting - they are more likely to be subtle.
In terms of this latter category, unless an account responsible for the weaving of data flowers is also affected, the impact of political Sybils are more likely to have impact upon the quality and prominence of the 'pollen data' than the overall structural integrity of a data flower. Even so, that could influence the overall 'takeaway' of a flower and so like any good gardener it'll be important to filter out the pests before they can do much damage.
Similar applies to monetarily-motivated Sybils mentioned in the earlier category.
Also, in the same way that Witnesses are to some degree accountable to the whims of voters, so too would those responsible for Data Flower compilation (Delegates) be.
TL/DR Two reasons for Sybil attacks: Money or Agenda-Pushing Influence. Data flowers having central data weaving responsibility limits impact but its still a potential problem.
How this might be done:
I have noticed that there is a waiting list to getting accepted as a user upon Steemit - and I presume that one significant reason for this is Sybil attack prevention. Even so - if enough bogus users do make it through, they may still be able to get away with non-overt suspect activity. I also have a couple of further ideas but they go a little beyond the scope of data flowers specifically, so kindly keep that in mind.
One Possibility: New users need to build their reputation sufficiently so as to be able to claim their rewards.
I'd suggest a reputation level of 30 as an easily achieved target. Users under this level will still accumulate rewards but they need sufficient reputation to claim them.
This would mean that Sybils would require a fair bit of work to not only do their circle voting undetected but also to reap their gains.
An alternative that may reduce the need for user vetting: Have non-vetted users start with a non-modifiable reputation of 25 - without voting or following privileges. They may also not be followed.
They also have the option to 'apply for vetting' - which would consist of a few fields, and the option to upload an ID photo.
Those who vet would have fewer applications, and would not only have the vetting application to base judgement on but also the post/ comment history. Their options for vetting: Accept, Insufficient (Keep posting or include ID), Warn (Improve post quality and re-apply) and Refuse.
Upon acceptance their adjusted balance in reputation and rewards begin to show.
The advantage of this is that potential Sybils need to jump through hoops to prove that they are legitimate. The downside is that some users may claim that they are being prevented from becoming users because.. 'reasons'.
TL/DR New users get through unless they are machine-recognizable (e.g. via IP) clones. However they cannot vote/ curate, their reputation is frozen and they cannot claim rewards until they apply for vetting. Less work and more material to work with for vetting team. However non-accepted users may get vocal.
A Second Possibility: Similarly, imposing a reputation minimum so as to have one's content eligible for data flower incorporation, as well as to curate data flowers through the attribution of 'weave tags' may also have the effect of mitigating their effect.
I'd suggest a reputation level of 35 as minimum for one's content to be considered eligible for consideration, and a minimum reputation level of 40 for 'weave tag' attribution.
And since'weave tags' also double as 'votes' for data flowers (basically voting for a piece of data on the basis of how many times its been tagged in a certain way), the effect that Sybils can have is mostly limited to the current level of Steemit curation.
TL/DR Use minimum reputation for being able to claim one's rewards to reduce convenience of Sybils in general. Also use minimum reputation level to determine who is proven to be trustworthy enough to contribute to and curate data flowers.
This Could Inspire a New Level of Reward
Reward flows for flower-incorporation never 'truly' die. This is the closest that one gets to a residual income (whether indefinite or not has yet to be decided). Flowers don't close. They undergo iterations. The regularity of an iteration depends upon activity. A highly active flower might undergo reward iterations every 7 days while a stale flower hardly touched may undergo an iteration over 6 months.
Activity is assessed via a consideration of "weave tags/votes" and development (with emphasis on the former). A flower with no new "weave tags/votes" or no development will see far fewer rewards than a flower with both. A flower with neither will not receive rewards for that cycle.
Rewards that are too small to quantify are either reserved for the next iteration, or grouped to the point where they can be divided and attributed authors and curators 'tagged' as having had preference this iteration, to be automatically 'skipped' in the next iteration such that other contributors aren't frozen out.
A relatively small portion of the STEEM pool is allocated to Data Flower rewards.
Highly active expansive data flowers will gain a bigger allocation that a less active, limited data flower.
That allocation will be distributed among:
- Delegates (Master curators that are voted on just like Witnesses)
- Content Incorporated Creators
- Curators
I admit that I am not clear on 'how' or what portion of rewards would be attributed in such a way - but I know what I am aiming for.
A Garden of Information
I am aiming for a garden full of data flowers, organized - easily traversed - a bit like Wikipedia but much 'tighter' on the quality of information.
Some among these flowers will cover areas that are well-covered - while others will focus upon more niche or fringe topics. Others still will focus upon gathering similar kinds of information or contents.
The result will be a living data mine.Not everything will be 'front-line information'. Some petals will need other petals to be peeled away to reveal them. But the data presented will paint an effective picture for anybody seeking out information about whatever topic.
This I consider to be a beginning - a seed of that which could be.
I do not pretend to know the full implications of that which I am suggesting - but imagine, if you will, a user being presented with a number of complete data flowers in response to a search term - or even as a result of association with a post that you clicked on and would like to learn more about.
It matters not the subject. If there are people writing about it then there will be a flower waiting.
Flowers are beautiful. They are one with their purpose in life - that of expression.
And is this not also akin to the purpose of data?
So this about covers my second Steemit suggestion. I must admit that I feel more convinced about the worth of this one, the other having had its obvious drawbacks.
I appreciate any feedback and critique that you Steemians have to offer.
Did you like this suggestion half as much as I did dreaming it up?
If so then there are two ways that you can help:
- Your feedback really helps me to identify flaws and improve upon the suggestion as a whole!
- Since I am follower-impaired your resteam could give this suggestion some of the exposure that it deserves.
And if you didn't like it then you could really help out by letting me know with a reason why (how else am I to improve?). :c)
Well thats all for now.
Sincerely,
