Algorithmic Crowdsourcing on Steemit as a way to grow demand for Steem Dollars
The Problem:
Steem Dollars are available to promote e-commerce and other activities on Steemit yet there are few ways to spend Steem Dollars on Steemit. In addition, the main way to earn Steem Dollars is by blogging. What if Steemit could become a crowd sourcing platform on top of e-commerce and blogging?
A Solution: Algorithmic Crowdsourcing
Steemit wants to grow and information processing is a source of growth for Steemit. At the same time, it provides new avenues for Steemians to acquire and spend Steem Dollars. Implemented right, the reputation system would become smarter, bots could become more beneficial, as a result of the addition of human computation.
E-commerce could benefit, as could many microwork activities. Translation for example could be broken up into microtasks where an algorithm auto translates and then human beings improve on it until it is perfected. Each human being would be assigned a part of the task and earn a percentage of the fund for the job.
What is algorithmic crowdsourcing you might ask?
What does Steemit gain from algorithmic crowdsourcing?
This quote below elucidates on the benefit of algorithmic crowdsourcing:
In an experiment with 1000 movies, the crowd was asked to
judge whether each movie was a comedy or not. Even with 10
judgments per movie, only 64% of the movies were correctly
classified, incurring a cost of 20 USD. However, when a
hybrid process is employed, higher quality can be achieved at
significantly lower costs: in another experiment, a workflow
employed crowdsourcing to train and boost a machine learning
classifier. The learning classifier operated on a high-
dimensional so-called perceptual space, which was created by
aggregating simple crowd-judgments like star ratings. Interest-
ingly, the same quality as reported in the first approach was
reached by just spending 2.80. While spending the same
amount of 20, boosted the quality up to 80%. Furthermore,
the costs for this technique are nearly independent from the
size of the input data, and even significantly larger movie col-
lections can be labeled by only slightly increasing costs. This
example clearly demonstrates the potential of hybrid
crowdsourcing workflows
What Steemit could gain is efficiency. Information has to be processed and some information can best be processed by humans. For example, when people post in the wrong tags, a crowd of human beings could be given the incentive to reclassify these posts, acting as human computers in a reclassification algorithm for the benefit of Steemit. If these human beings could be paid in Steem Dollars if and only if the activity gets enough votes, then you can fund this algorithmic crowdsourcing activity through a post on Steemit. The post would have to use a smart contract and be designated as a crowdsource request post, and anyone who upvotes it would fund the human computation. In this way, the bounty would be set and the crowd could fulfill the requirements.
More complicated crowdsourcing could come into the picture later on but this is just to present the idea to the Steemit community for discussion.
References
Lofi, C., & El Maarry, K. (2014, July). Design Patterns for Hybrid Algorithmic-Crowdsourcing Workflows. In 2014 IEEE 16th Conference on Business Informatics (Vol. 1, pp. 1-8). IEEE.
Zhang, H., Horvitz, E., Miller, R. C., & Parkes, D. C. (2011). Crowdsourcing general computation.
Thanks, this idea seems like one great way to put our SBD (Steem Backed Dollars) to work! :)
That is exactly the point. Having SBD just sit around isn't very useful but there isn't much you can do with it yet.
Before to put our SBD to work - need earn this SBD)
It seems that there are, potentially, many spin offs for Steemit, algorithmic crowdsourcing being only one, that could turn this into so much more than a social media platform. Another post suggested using Steem Dollars for some Ebay Steembay platform. The future's looking bright, where are my sunglasses?
The e-commerce platform idea is known but if you saw my other post, the first issue is to put in privacy protections. We can't rate service providers or review products if we cannot be guaranteed that the reviews are anonymous yet verifiable.
At the same time, crowd-sourcing is something you can do easier than e-commerce and you can even use the crowdsourcing to improve the quality of the e-commerce platform. A lot of people for example will not have any physical item to sell yet might still be willing to rent out their human computation. In parts of the world where people live on $1 a day, this crowd sourcing capability could tap into these resources.
To put it simple, we can't all be bloggers.
very well written
I read this idea in the depths of steemit before(Not as well thought as this though), but it didn't gain any foot hold. I'm hoping this does as i find this an incredible idea.
It's necessary. Not everyone can be a successful blogger. Soon Steem will have 500,000 or even 1 million bloggers,all trying to compete for attention and votes. In order for Steem to grow it has to expand and crowdsourcing is a big market with the lowest possible barrier to entry. Everyone would be able to spend their Steem Dollars this way or earn Steem Dollars from which to buy Steem Power for example.
This is an interesting idea in theory, but, speaking as a person who's been there, a lot of the freelancing and crowdsourcing sites that are featured in the graphic end up sponsoring races to the bottom. Often on these sites, quality is drowned out by budget. So, while these sites do provide a place to share ideas, these ideas are often low-value (and I understand this is not always the case) because they are produced and manufactured at extreme discounts. One of the things that is so great about Steemit is that the inherent value in the content drives the monetary value of the product. That's why this platform is such a game-changer, it reverses the dynamic's of fiverr, freelancer, and upwork and creates an environment that encourages creative work for the good of the platform and the mutual benefit of the participants. I am NOT saying that there is not a way to implement what you are proposing to avoid these pitfalls, just that I would personally like to see extreme caution taken if the community moves in this direction. Great post!
Some people just aren't going to ever make consistent Steem Dollars to pay bills by blogging. That doesn't mean Steemit can't make use of their information processing capabilities. For example, certain images might belong under certain tags but only a human can determine which tags.
As an aside I always wonder what people expect by posting references in non-academic contexts.
Is it just a marketing tool to confer legitimacy?
Since only an interesting minority will inspect quoted sources to check whether an excerpt makes sense in the context in which you're using it.
Interesting post though.
I don't see a need to separate by context. If you think in a logical fashion then you might want to know the references so you can trace back the dots which connect thoughts together. For example if someone posts an idea, the idea is not going to have much weight if it's not based on something from either academic literature, or provides a proof in the form of source code or demonstration.
You are right that only a minority of people will check the sources but it's precisely these people who will be the ones who develop Steemit. Kind of like not everyone is going to check the source code on Github but the people who tend to do that are developers. So where you are correct is that this post is geared toward developers and researchers because they'll be the people who write the code to make it possible.
Funny, a few weeks ago, before I had even heard of Steemit, I was thinking about how best to write blog posts on one of my websites. You know, writing to get a high Google ranking, which basically means writing good content nowadays. And about my Adsense ads, which hardly make any money. And the question crossed my mind, "Am I working for Google? Are the majority of people writing content working for Google?" Because that's what it seems like. We make content, a lot of people earn a little money, a few people earn a lot, and Google makes truckloads.
So I love this idea. I love the idea that I'm working for Steemit AND Google simultaneously, making a little money now for myself and others, and hopefully more in the future and I hope there's more work on the way through crowdsourcing ideas like this.
Great post!
good point
I'm agree with you
I love this idea!!! Make some posts "crowdsource" posts.
And ... you taught me a new word. I'd never heard of "elucidate". Google helped me clarify.
The strongest indicator of a healthy and growing platform is all the new ideas it's users are proposing. Give me more of this.