The Double Spending Problem on Steemit

in #steemit7 years ago (edited)

Hey everyone, this post has been rattling around in my brain for quite some time, and I finally got around to finishing it. I think I've really distilled my thoughts on this subject, and it's an important one, as it in a way forms an underlying foundation of my own viewpoint. This post is a bit of a follow up in my series of "thought provoking posts", the last being Which came first, the Engagement or the Reward? if you haven't read that one yet.

So, lets dive right in, I hope you enjoy!


The Double Spending Problem

If you don't already know, the "Double Spending Problem" is one of the fundamental problems with digital systems, and is the raison d'être of blockchains. In digital systems, if you have a piece of data (such as a song from your favourite artist stored as an mp3 file), and you sent it to your friend Alice, there is nothing stopping you from also sending it to your friend Bob. While this has caused an amazing explosion of the internet in popularity akin to the printing press's ability to distribute content quickly and reliably, it has an unfortunate downside when you want to transfer something that you don't ever want to be duplicated. For example, if you want to send Alice the deed to your house, you most certainly don't want Bob to see it and claim that the house is now his. And when you boil that down to a value system, or a monetary system, you want to be able to transfer value without it being able to be duplicated in the same way in which all other digital content is.

Bitcoin was one of the first implementations to really address the "double spending problem" of value on a large scale. We have already had tools like cryptography for a long time before Bitcoin, and (sometimes people in crypto forget or don't know this...) we have also had distributed consensus for decades before Bitcoin. The real novelty that Bitcoin provided was that is was an implementation designed from the ground up to prevent "double spending", by using cryptographic proof of work to achieve a single state consensus of an accounts transaction history in a global ledger.

Nowadays, there's several token systems that throw out the blockchain in favour of different approaches. For an academic, I was totally expecting this to happen: we know of other ways to do distributed consensus, so it was only a matter of time before the other options were tried. The IOTA Tangle is one of the most commonly known systems in the community, but more are getting popular, including work being done on the Hashgraph design, and one of my new favourites, the already fully functional zero-fee DAG block lattice, Raiblocks. All of these systems still solve the "double spending problem", using one mechanism or another to achieve distributed consensus.

Let's dig further into the problem. An important facet with the prevention of double spending in a currency or reward system, is actually also the inability to double spend "coinbase" (instantiation) rewards. That is to say, if you were to create a coinbase transaction (for example, you solved the Bitcoin proof of work and produced a block) which mints coins into existence, it is fundamentally important that you can not re-use this transaction and mint yourself some more coins. No, you only get a mining reward once, and then you need to work on a new one. This way, the system is provably fair: the inflation rate (or token production rate) is consistent, reliable, and can be determined in advance. Tokens cannot be duplicated, and they cannot be spent twice.

So what's this all on about, and how does this relate to Steemit? Yes, don't fret, the Steem blockchain is just as secure as any other.
But, I am here to tell you why Steemit has in fact not solved the double spending problem in it's entirety.


The Double Spending of Value

On Steemit, we have this unique concept of "Proof of Brain". The majority of the rewards are not minted into existence by miners, but are rather distributed to the community by way of the reward pool. If you think about it abstractly, it is a unique and amazing system. You produce something of value (an interesting post), and you get some reward based on a multitude of factors, and in return you recieve tokens that represent value. This is an abstraction of the "Proof of Brain" system that Steemit has designed.

However, on Steemit, there is nothing immediately stopping you from double spending your Proof of Brain "work". On a blockchain, time moves forward only, the distributed consensus knows you cannot double spend Proof of Work that you created, so you can only ever get that reward once. On Steemit, it is all too common to duplicate your "work", and be rewarded twice. This can come in many forms. As the most obvious example, you might repost something you posted on Steemit a month ago, and perhaps claim "it is now reaching a new audience". If that repost gets a new value, you just double spent successfully.

On the extreme, we can see even worse behaviour. Imagine if on Bitcoin, as soon as a miner produced a block, you could send out the same block and recieve the same reward that the miner did. How crazy would that be? The blockchain simply wouldn't work. Well, on Steemit, there is fundamentally nothing stopping someone from taking a successful trending post, and posting the same thing, and getting a reward for it.

There are a few places where we run into a morally grey area, specifically when cross-posting. Imagine, if you wrote an article for a news site, and they paid you for it. Or you wrote an article for your personal blog that is hosted outside of Steemit, where you receive ad revenue. You may have even done work for someone pro-bono (in which case, the value you are receiving is not monetary, but social — such as the case of an artist doing pro-bono work in exchange for exposure), and had it utilized somewhere in the real world or on the internet. In all of these examples, you have already spent your "Proof of Brain" and received some form of value in return. The value you received was outside the Steemit blockchain, so it may seem ok to double-spend your “Proof of Brain” (or in other terms, "intellectual property”) by posting your work and also receiving value for it from Steemit.

However, this means that you would be receiving value twice for the same piece of work — this is like if a novelist sold the same book to two publishers and didn’t tell them. Imagine if that happened in the real world. It would be a disaster, right? It’s both the morally correct thing to do and common courtesy to let someone know if they are buying an exact duplicate of a piece of work. So if both publishers were told that they were buying the same book and were fine with it, then there’s no problem. The issue only arises when both value-adding parties believe they are buying an original work when they’re not.

So with this in mind, let’s examine a situation where you sell an article to a news site, but also let them know that you will post it on Steemit. And when you post it online, you let people know that they can also find the article on the news site. Since both value-adding parties involved (the news site and Steemit readers) are aware of the double-spend of your “Proof of Brain,” then they can make an informed decision about how much value they want to give you. No harm, no foul — as long as everyone involved is aware of what’s going on. In practice, this just translates into letting people know when your post can also be found elsewhere. This is the correct way to go about cross-posting.



Distributed Intellectual Consensus

Hopefully by now I have convinced you that, unlike blockchain "Proof of Work", where the distributed consensus prevents you from double spending "Work" implicitly and as part of the system, the Steemit "Proof of Brain" has no built in mechanism to prevent double spending of the "Brain". So how do we actually prevent double spending of ideas?

While the Bitcoin system is proactive in that the transaction will be rejected immediately, right now, the only system we have is retroactive. The only way to reject value going towards a double spend, is for the community to retroactively remove the value on that transaction. This is a round-about way of saying that the post must be downvoted to remove the reward. Upon this, the transaction in a sense was not rejected, but rather simply not rewarded.

This method is what I would call "Distributed Intellectual Consensus". In this system, all actors must make an informed decision, and reject double spends of value. In order to make an informed decision, we use can investigate, consider reputation, use tools, and use many other methods to help identify if something is indeed a double spend or not, or has otherwise earned value. Eventually, we can come to a distributed conclusion on the value of the post.

This, as you can imagine, brings me into talking about @cheetah. When it really comes down to it, this problem is exactly why @cheetah exists, and why I created her. It is a tool to help identify if a post is (potentially) something has is being attempted to be spent twice. It's not perfect, and it's not 100% accurate (its actually about 99.8% accurate -- in a metric of how many posts are false positives -- according to my calculations), but it is one of many tools that we can use to help critique a post. The idea in this post is precisely why I created her, and this is why I believe she is so necessary.

But note, on Steemit, not everyone evaluates and votes on every post. While this would allow us to have a true, complete democratic (by stake) view of a posts "worth", what really happens is that we get a sampled version of this. Only a select sample of users view a post, and therefore it is up to them to determine the validity of it. With a tool like @cheetah, and a group like @steemcleaners, it becomes more feasible to evaluate efficiently. Just like how Delegated Proof of Stake improves regular PoS by delegating the responsibility, we can use systems like @cheetah and @steemcleaners in order to help evaluate a post. But we still need users to help us too, by reporting questionable posts and even taking action whenever possible, and therefore I encourage everyone to take these problems seriously.



So that was my ramble. And yes, this is strictly my own personal viewpoint on Steemit, and the "Proof of Brain" system. It's subjective, and I'd always love to hear the opposing view to refine how I think about my own views.
So, let me leave you with this parting "ultimatum" of an idea, and let me know what you think in the comments below:

If you believe the Bitcoin blockchain would not work if miners could spend the reward of blocks as many times as they would like, you should also believe the same of the rewards given to Steemit posts.



Like Cheetah and Steemcleaners, and want them to stick around? Like what I do for Steemit? Please vote for me as a witness here!

Sort:  

Very cool discussion.

I do find a slight possible flaw in your thinking when comparing to the bitcoin blockchain. The bitcoin blockchain has a set amount of rewards created per block. Replaying a block creation would change the inflation rate of the currency itself which would be a very serious thing. In this DPOS system, the 3 second block production is trustworthy and unchanged, regardless of how many double spends of content happen. To me, that's an important distinction. To me, there are two levels. The first level creates the token and that's the really important one that should not be possible to "double spend" or it changes the fundamental economics of the token. On a second layer, we have the proof of brain distribution of the rewards pool. Ideally, yes, we want to avoid double spends here as well, but I don't think it's at the same level of concern as a block production which creates new tokens out of nothing.

As I was reading through this, I was reminded of a post I voted for recently which was a copy of a post which was vendetta flagged to $0 by a whale. It was a valuable post (IMO) and deserved a portion of the rewards pool and since the original version was flagged to $0 for irrational reasons (IMO), I was happy to vote it up. To me, it's an example of how all value is subjective. In that case, I was happy to support it because the author disclosed the duplication (and the reasons for it) and the post was personally valuable to me.

That reminds me of a post I did last year: Why You Should Care About Plagiarism and Fair Use. From that perspective, it's about determining who in the community is a good, trustworthy actor, and who is not. If someone republishes super valuable content, I'm okay with rewarding it if it's new for me and I know this person isn't purposely trying to scam anyone such as reposting something which was supposed to be exclusive content elsewhere.

From a certain perspective, every time a movie plays in the theater, they are double spending the original work. It's not about some labor theory of value, more about who is willing to subjectively assign a value to the exchange in that moment and pay for it.

Good analysis on the blockchain comparison, and you are right: in this sense it's not creating new tokens, it's purely about the distribution. The creation is secure, but the ethical dilemma remains.

From a certain perspective, every time a movie plays in the theater, they are double spending the original work.

Not exactly. You paid for the ticket (I hope) and some of the money goes to the people who created it. You are an informed purchaser: you know that you are buying something meant for mass consumption and that other people are going to view it.

The example that resonates with me more is the artwork example. If you create a painting, and sell it to someone under the assumption that it was the only painting of it's kind, the purchaser would be pretty unhappy if they find out that you are actually mass producing the exact same painting. The purchaser would feel cheated (and rightly so) as they thought the value that they bought was unique.

Most of my point in this post boils down to the problem of being an informed purchaser, or an informed voter. You may change your voting behaviour if you knew it was a copy paste job, or if it was a steemit exclusive.

I agree fully. The issue is fraud. If people are acting fraudulently, that lowers value for the whole network.

"There are a few places where we run into a morally grey area, specifically when cross-posting. Imagine, if you wrote an article for a news site, and they paid you for it."

Yes, but it's also very common for a publisher to buy limited rights, for instance first electronic rights, and those vary depending on the publisher. Otherwise we'd never get anthologies in science fiction. I agree that disclosure of those terms are important, and the key to avoiding conflict of interest.

In my case, my IGMS columns have a one-year blackout period before I can re-post them. I have so far only posted links back to them, because I want to support the magazine that supports me, and Steemit offers little friction to linking to an outside source. But because IGMS doesn't do much graphically with its website, I could totally see reposting them here (after one year) with linked drawings and photos. It would make no difference to the letter of our agreement, but I would want to add value on the second platform.

http://www.intergalacticmedicineshow.com/cgi-bin/mag.cgi?do=columns&vol=randall_hayes&article=_index

Congratulations @anyx, this post is the second most rewarded post (based on pending payouts) in the last 12 hours written by a Hero account holder (accounts that hold between 10 and 100 Mega Vests). The total number of posts by Hero account holders during this period was 311 and the total pending payments to posts in this category was $6235.46. To see the full list of highest paid posts across all accounts categories, click here.

If you do not wish to receive these messages in future, please reply stop to this comment.

Interesting post. You're calling attention to technical concerns, but also some ethical concerns as well. Right now, it is all too easy to exploit the system and "double-reward" yourself by reposting your old work. And to be honest, I wouldn't be interested in doing that, but plenty of people would have no problem with it! There is an ethical dimension to all of this, and it will be a challenge to reel in our darker inclinations that all human beings have.

It seems more likely that we will generate technical solutions, rather than convince everyone to act ethically. So you are wise to approach the problem from that angle.

While I am very new here, I already can see how I will potentially lose out on a lot of rewards if I refrain from using "vote for vote" or "follow for follow" tactics. But I really want to follow and vote for people that I feel deserve it! I've been fascinated by the ethical choices the platform makes us make, but you're making me realize that there are still fundamental technical issues to address too.

Steemit really is the Wild West of the internet right now, but I am happy to see that there are people like you out there that are trying to implement changes that improve the system overall.

Thanks for posting.

too easy to exploit the system and "double-reward" yourself by reposting your old work.

As a new user I do not see this as an issue. It would only be a double-reward if you upvoted yourself on the old post and then upvoted yourself on the new post. That is the only way an Author can "Double-Reward" themselves. If I saw an Author post 4 months ago, and then the same post today, and I remembered that post, then is it not for me to decide if I want to reward the post again? I mean face it 4 months ago my vote was worthless, $0.001 if that. Today at full power it is $0.070, I now have the chance to give a semi-Real Reward to them. Why should steemit be different than any other type of publishing company. If you were to go into any music store or any place that sells music I am sure you can buy a copy of Elton John's "Goodbye Yellow Brick Road" Album. If you went into a bookstore I am sure you can buy a new copy of Anne McCaffrey's Dragon Riders of Pern. Why should I view steemit differently. The reason you can still buy those items is marketing to a new audience. I hope that steemit continues to get new audience members.

So in conclusion I myself do not see reposting information from the past as a bad thing.

While I am very new here, I already can see how I will potentially lose out on a lot of rewards if I refrain from using "vote for vote" or "follow for follow" tactics. But I really want to follow and vote for people that I feel deserve it!

The marshmallow experiment comes to mind. I think that a lot of people using the former tactics are in the camp of taking the one marshmallow. The quick, easy, and gratifying reward, but the one that turns off followers. Whereas what you want to do is wait, play the longer game, and build up a base: and in the long run, get a greater or more consistent reward. Two marshmallows.

Steemit really is the Wild West of the internet right now

I've said this for the past year. :) I wonder if it will actually change, as solutions (if they exist) certainly don't appear to be trivial.

Good point. I studied psychology in university, so I am familiar with that experiment. F4F or V4V is a quick fix, whereas building a real fan base is a long-term challenge.

Yes, the solutions that need to be generated are not simple. There will perhaps be more complexity and exploitation before there is clarity.

Thanks for the post and the replies.

Repost that consistently repeats value creation is likely because its reaching NEW readers and the content is still the same and good. So its getting to its true value, as most people don't curate the same post twice.

Hello @ anyx. Just browsing through, and checking out the witness that I voted for. Happy new year.

@anyx this is a very important point. Thanks for making this noticed by many more users. Perhaps there are developers lie to thinking of new ways to continue improving this systems, with bots like cheetah and I am sure the algorithms will be better every day, and they will be more efficient to get even better results. Upvoted!

Regards, @gold84

I think SMT's have amazing potential for artists and content creators to form communities that target readers and potential clients. Whether you are an amateur or a professional. The communities themselves will have more control over content, membership and rewards.

And.. I didn't put in my original thought...

This would allow communities to define their own payouts. With a better UI and the ability to control certain aspects such as membership to keep out the spam & scammers I see no reason content couldn't receive rewards years later.

Something that would be a nightmare here on steemit.com.

Totally agree. If only communities could come faster. :)

@anyx -Wow - you do make an interesting point about Steemit having not completely solved the double payment issue. I guess the creators depended on the moral values being upheld by content posters. However, you are right, the blockchain, by itself does not prevent double spending that can be caused by re-posting by authors or plagiarism by some authors of old Steemit posts etc. In fact, when the Steemit interface had some issues couple of months ago, many Steemians ended up double posting content (they hit submit twice because the interface showed transaction error first time) and the exact same post got published and paid twice!
Thank you for the interesting treatise on double spending issue. Upvoted

Regards,

@vm2904

In fact, when the Steemit interface had some issues couple of months ago, many Steemians ended up double posting content (they hit submit twice because the interface showed transaction error first time) and the exact same post got published and paid twice!

This is an excellent example. If people realize that they can do this, and it gets rewarded, it might start to happen on purpose, and more often. We really don't want that!

great post. very interesting and thought provoking. what about re-purposing elements of content, re-contextualising and juxtaposing with new ideas?

there is a train of thought which says nothing is original. Although I agree with the sentiment of this article. let's say I posted a picture of a rainy day and wrote about what it's like to experience a rainy day. the image is my original one and so are the words.

perhaps later, I would like to post a curation of my work on the theme of water. is is fair and justified to then include my rainy day picture within the post? it's re-contextualised and I add new thoughts and words about my relationship towards water as subject matter. the previous post is juxtaposed against an new background, it's contextually sensitive. I don't think it's a morally grey area in this case. It's intellectually asking a question about what originality means and how strictly is it applied if we have diverged sufficiently from the original purpose of preventing double spend.

I believe in the normal world double spending is very common, accepted and a regular feature of life. proof of brain simply throws up a new question, what is originality, how can it be defined? and this article has made me think about what that means on steemit, in art and in life. Well done you crafty goat.

Th@anyx & A Merry Christmas. great work !

Good point. Where is the line? It will be difficult to draw one.

In my comment above, I mentioned how this problem is not only technical, but ethical as well. You seem to be picking up on the same idea in your comment.

Maybe we have to use our own discretion in the examples you provided. We of course know that a straight copy/paste of an old post seems ridiculous, and we have to begin to make our own ethical considerations as we walk back from the most flagrant abuses of the system.

Overall, I feel there is little issue with re-purposing when you have put effort into reworking the piece into something "original" again.

copy and pasting posts is one thing. Steemit's proof of brain is most interesting because it has evolved directly out of social media where ideas were rewarded with likes and popularity which meant no physical reward except for the most popular by way of corporate sponsorship, similar to top athletes. steemit takes the centralised (zuckerberg gets all the cash) model and distributes it. I find as a steemian with genuinely original content. ( I am a creative professional) with over 20 years of experience, my content doesn't get rewarded commensurate with it's merit. I see total crap making hundreds of dollars all time and I never score that highly. There is another force at work here and it's to do with the power of social networks and currency. In the normal centralised model, there is the opportunity for success through virality. However virality often takes more than seven days to achieve. Steemit will evolve a virality function (the ability to earn past seven days) and then it will reward more fairly content with artistic and intellectual merit. Merry Christmas. On the flipside, my meagre rewards can be traded for their real value with an extra twist to the equation. swap the tokens for ones which increase in value faster than the platform rewards. This is most satisfying. It's like finding pennies and turning them into more. magic beans indeed said Jack's mother.

@outerground, you're totally right. Something else is at play here.

It is interesting how Steemit reflects society to a degree. Those with monetary capital and social capital can exploit the system. The most talented people aren't necessarily getting rewarded. Steemit, in its current state, certainly isn't a meritocracy.

Now, that isn't to say I am even bashing the platform. It's growing and maturing, and these things are bound to emerge. But I do hope those aspects of Steemit can be worked out in the long run.

Merry Christmas to you too! And good luck on Steemit in the New Year.

thanks Chris, from a fellow Chris.. you have a new follower. I too am fascinated with steemit. I've been here every day since may and I certainly have got a lot out of it. It will evolve and it is full of talented people. Enjoy Christmas and I look forward to reading your posts. You'll soon be up past 50 rep ! ps thanks @anyx for the introduction

Thanks for the encouragement! Us "Chrises" have to stick together. I look forward to seeing your content as well. We'll talk more soon!

There's been some good experimental work on this issue, showing how social information alters our valuations.

https://steemit.com/steemit/@plotbot2015/experimental-proof-that-the-quality-of-your-posts-does-matter-some