A question for the Steemit community about the mechanism of downvoting and the dangers of charlatans

in steemit •  7 years ago  (edited)

snake oil salesman

I have a question for you, Steemit. But first, I want to share with you a true story that I learned from reading "Charlatan" back in 2010.

In 1885, there was a man born by the name of John Brinkley. Despite no actual medical education, he purchased a degree and became a doctor. His specialty became removing the balls from goats and inserting them into the scrotums of men (and even women), for the purpose of curing a wide variety of ailments, most especially sexual dysfunction. He did this for many years and eventually became a millionaire (goat balls aren't cheap you know), despite his practices being utter bullshit and actively harmful to people to the point of being responsible for their deaths. He even went on to diagnose people over the radio and prescribe them bullshit potions to buy, harming people en masse.

This man was able to lie to people who didn't know any better for decades and got rich of it. With that in mind, my question is the following:

How should this community best handle the spreading of views that cause physical harm, and even possible death?

The tricky part of this predicament is that we live in a world where the internet has allowed us to create bubbles around ourselves such that erroneous views can survive and propagate within digital echo chambers. If you've heard the terms "post-truth" or "fake news", these are the results of these times in which we live here in the 21st century. Technology allows us to plug our ears and hum like never before in history. It's a new Golden Age for the spreading of falsehoods.

I love how Steemit has a very free speech attitude to things. I think freedom of speech is extremely important. At the same time, there is a downside to free speech where lines are usually drawn around communication that has the potential to cause harm, and I don't mean the proverbial "sticks and stones may break my bones but words will never hurt me" kind of harm. I mean "drinking bleach is actually good for you and here are 10 bleach recipes" kind of harm.

So the question becomes how to best handle those kinds of posts. On Reddit, the community is divided into subreddits, and moderators are able to moderate such posts. I'm actually one of those moderators for a subreddit called /r/Basicincome. If a post doesn't follow the agreed upon rules and guidelines of the sub, it is within my power to remove them. But Steemit isn't divided into subs, and there are no moderators aside from admin.

On Reddit there's also a downvote option, to enable a mechanism for the cream to rise to the top, and the shit to sink to the bottom. But the downvote mechanism doesn't exist here either.

Here then comes the idea of using the flag button as the missing downvote button, but I've noticed that because downvoting isn't its actual purpose, and because the effect of this artificial downvote is public instead of anonymous, that the poster of something downvoted will pursue an eye-for-an-eye strategy of flagging everything posted by the person who flagged their post. I'm fairly new here and this has already happened to me in retribution for a downvote, so I imagine the problem must be widespread.

So what should Steemit as a community do about potentially harmful posts?

The laissez faire free market kind of approach is to consider Steemit a free market of ideas, where what's upvoted has value and what isn't upvoted doesn't. But sometimes it takes awhile to calculate value. Take for instance a health blog where someone recommends putting one drop of dimethylmercury in every meal for the purpose of extending life and increasing well-being through the magical power of metal. Let's say hundreds of people start doing that and thanks to the placebo effect, they start feeling great within a few days, so they not only give the post a 100% upvote, but they also send the person Steem and even money to purchase a supply of poison to sell to others as part of a new poison pyramid scheme.

Free market thinking here says that eventually people will begin to figure out that what this person is doing is killing people, and people will stop buying his stuff. Let's assume this is true. In that case, how many people needed to die for the market to respond? Additionally, will that number fall to zero? Or will there continue always being a sucker who didn't get the memo?

It seems to me that Steemit needs to upgrade itself in this area. Either moderators need to be created and given power to remove such things, and/or an official downvote button needs to be added to help lower the visibility of harmful posts, and/or the flag button needs to be officially recognized and explained as the proper response to such posts, with the further possibility of retribution flagging removed.

If people feel downvoting posts by flagging will lead to retribution, that itself is a chilling effect. All the charlatans out there who wish to profit from the way Steemit currently functions will cry out that downvotes are censorship, but you tell me, should we or should we not reduce the visibility of physically harmful, downright counterfactual posts?

If it's wrong to downvote posts that recommend eating raw uncooked pork for its "natural vitamins", how do we help prevent people from getting trichinosis? How do we protect the uninformed from being hurt by those who are either ignorant themselves, or are hucksters like John Brinkley out to make a buck without giving a shit about who they hurt in the process?


Who am I? Read my Steemit intro post to learn more about me.

Subscribe to my blog | Follow me on Twitter | Like me on Facebook | Follow me on Steemit
Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

You have reasonable examples here, like with bleach. In actual dangerous cases, for example, if an author tells someone to mix 2 chemicals together--and i will not even mention the chemicals as an example--these chemicals can kill someone if breathed in. This is not a "view", it is clearly a malicious direction that is known to cause harm and has clear, concise, indisputable evidence to cause harm. Right?

"How should this community best handle the spreading of views that cause physical harm, and even possible death?"

"Views" (or opinions) do not cause physical harm or possible death. Someone sharing their perspective, their witness on an event, their testimony, etc. also does not cause physical harm or possible death. However, it can cause cognitively dissonant people quite the irritation.

We can all agree that preventing others from seeing another person's message, even if we strongly disagree with the author's "views" or if their views are simply not popular, is purely censorship.

Another example, someone having a hard time accepting an author's testimony and research on the negative consequences of dairy and high protein diets .... well, that's just trolling. It is no different than the Salem witch trials--these people would easily silence people through murder; after all, if you believe you are saving the lives of many others, isn't that "viewed" as beneficial for society as a whole? As for trolling, people are absolutely within the right to express their view :) https://steemit.com/nutrition/@robertgenito/protein-for-muscle-milk-for-bones-and-other-lies-we-believed

So how should "flagging" happen in the real world?

I think a good start is to have flagging on the side of the publishing website itself. This means that steemit.com or busy.org would have the ability to flag and hide unpopular content. This "flagging" behavior would be saved on the website's database end, not on the STEEM blockchain. The STEEM protocol and blockchain itself would not have a "flagging" and reward removal built in. This way, when people get sick and tired of the censorship on steemit.com (FOR EXAMPLE ONLY), they can always "vote" with their opinion by discontinuing to use steemit.com and begin to use a less censored website like busy.org (FOR EXAMPLE ONLY).

DISCLAIMER: I am not accusing steemit.com or busy.org in acts of censorship.

In large part, the concern you raise is being addressed is the creation of 'communities'. This was announced in the 2017 roadmap. Communities are on track to be rolled out in the next couple of months, as far as I am aware.

Nice! I hadn't come across that information yet, so thank you for sharing it. I think communities would definitely be an improvement, although I think only one positive step among multiple steps that would be helpful to be taken. Cheers.

Sure thing, @scottsantens. There have been tons of good conversations here on how to handle challenges like those you bring up ... but most of these are nearly impossible to find at present as they are spread out across several months of posts in our still-nearly-unsearchable platform.

As of now, especially when it comes to complex subjects, there is still plenty of work to be done on how to prevent users from posting harmful lies without creating some kind of (inherently biased) 'truth police'. Your articulate and balanced perspective on the matter can doubtless help this effort along: )

Hadn't heard about communities and I thought I'd read the roadmap. Apparently I missed that. Good to know!

Easy to miss amidst the day-to-day noise here. But based on stuff like @ned's AMA from a few months ago, communities could become quite a thing.

Flagging is just the way the Steemit.com User Interface has chosen to represent what is essentially a downvote. If you go over to Busy.org you'll see a downvote option with no flags... but it's the same backend.
I agree the flag representation makes people more apprehensive about using it, I prefer the downvote icon. But in the end the functionality is fine. Retribution is unavoidable if the user decides to respond in that fashion. It's a fact we just need to live with and each decide individually, (or through collective organizations) when it's worth sticking your neck out and fighting for accuracy, against spam, against plagiarism, etc.
If someone downvotes (flags) potentially harmful medical advice to properly curate it as junk, I support that use 100%. They should absolutely take the time to post refuting information in the comments I'd hope.

Interesting. I didn't know about busy.org. I'll have to check it out.

As for the functionality of the downvote button here being fine, I have to disagree. It took me awhile to even figure out that that's what was even meant when the word downvote was used. The least that could be done is to add text within the button so that when you click it, downvote is listed as one of the reasons. A down arrow instead of a flag would be even more helpful.

Also, the point of this post was to get people to consider the full costs of the attitude that says we should just count on some people sticking their necks out to downvote people. Is downvoting even compensated and thus incentivized? And are we okay with the people out there getting hurt because of a desire to avoid coding an actual downvote button?

I agree with you on the downvote/flag clarity. I'm ok with the actual functionality, but presentation could use some changes. By functionality I mean that when the flag is used, it does exactly the same things as a vote, just with negative values. So I think the (+)(-) correlation is good. There needs to be less of a stigma applied to the flag and rebranding it as a downvote is a good first step.
I'd be worried about incentivizing the downvote. Collusive rings could then profit via downvoting while harming content creators. Right now flagging actually costs vote power with no benefit to the person doing the flagging. This is kind of a good way to reign in abusive flagging. It leaves it more in the realm of people utilizing it altruistically such as the medical use case you've described.

Check out how "flagging" can be a crippling flaw for Steem's long-term potential... https://steemit.com/steemit/@robertgenito/steem-s-1-crippling-flaw

I respectfully disagree.

The Steem platform is based on consensus and a voting mechanism. That doesn't mean that everyone's vote automatically delivers a certain level of reward. Consider a national election. 50 million people cast votes, a candidate wins with 28 million votes. Under this logic we've just censored 22 million people because their votes resulted in no net gain. I don't see that as the case.

Further, removal of this tool leaves no means of dealing with identity fraud, plagiarism, spam and abuse. For example, there is no systematic means of categorizing pornography and nsfw content. Perhaps you disagree with doing that anyway and would rather it be uncensored and automatically displayed next to other content. That was the case previously, and would be a major impediment to Steemits success. However this content is now voluntarily tagged and properly controlled only due to the fact that flags exist.

Good Steemians are using flags to combat pools wherein users simply post 100 comments with no more content than counting the numbers 1-100. Each comment receiving 3000+ up votes and takes thousands of dollars for these collusive rings.

I happily flagged a weeks worth of posts of an author who was posting illustrations found online and claiming they had drawn them. You're suggesting that people should keep those rewards, presumably because it's not wrong until you get caught.

Steemit is a website. It will not be free of censorship. Even now Steemit takes steps such as blocking accounts utilized by children because it is in violation of laws if it does not. And even so I hesitate to label flagging as censorship because it is community driven.

If a political rally is broken up by the military, that's censorship. If a political rally is drowned out by the chants of a counter protest, that is not censorship. How would you quiet the counter protest without censoring them? In a decentralized setting the loudest voice wins, and here volume = Steem power.

But the more important aspect is that Steemit.com is not the Steem blockchain. It is the blockchain that is decentralized and censorship resistant. Your content cannot be deleted. Anyone can make a website with different display rules. Someone can set up UnderSteem where the most flagged posts rank highest on a trending page!

People can already import a feed of Steem content into a Wordpress plugin. Certainly they're under no obligation to display all the content on the blockchain. Choosing not to display something is not equivalent to censoring it. Even on Steemit itself, those posts flagged to negative still remain accessible and easy to click. I find myself exploring them to see what's up! Lessened visibility is not censorship, and all financial rewards are subject to the consensus of everyone's aggregate voting. We may get the sense that our votes are worth a certain amount at a given point in time, but we're just seeing the consensus being calculated on the fly. No one ever has Steem they've earned taken away, there is no mechanism to claw it back from your wallet. Pending payouts are subject to change.

If I send an article about flat earth theory to National Geographic and 90% of their editorial board chooses not to publish it, they have not censored me or taken away the money I should have earned for writing an article. They have simply 'flagged' it as unfit for their publication. We, all of us, the Steemit board, sometimes do the same.

There's a decided lack of easily accessible information about exactly how flagging/downvoting works. There's also conflicting etiquette regarding it. I've been piecing it together over the day or two since being here. As far as I could find out last night there is essentially a reward penalty for down voting. You use voting power as you would with an upvote, but you get no curating reward.

Glad you are bringing this up and it is getting some up-vote love. There is definitely a lot of spammy and low-quality posting on Steemit that could use more curation IMO. Any number of people desperate for you to upvote them and follow them, any number of people trying to promote some get-rich-quick kind of service and any number of people just spouting BS that wouldn't last 10 seconds on Wikipedia.

I also like how you linked it to the free-market and illustrated how the free-market can do a lot of harm before the supposed feedback mechanism kicks in. And I say supposed because in the real world there is seldom anything approaching the transparency and freedom of information to allow feedback to work properly. Wealth, market power, and legal intimidation can be used to bury information that would allow consumers, or the buyers in the market to correct their erroneous behavior based on negative outcomes.

In most of my reading of Steemit docs I have seen the Flag button described to be the same as a downvote for whatever reason. If this is the case I think the UI should be changed, if only to update the text describing reasons to Flag.

The problem with a downvote is it can easily be abused by a small but dedicated minority. Think of those cases where there is some negative news story about a restaurant or business and their Yelp page immediately gets inundated with one-star reviews from people who have never even been there. Similar things could happen here and I'm sure already have to some extent.

Before you know it we'll have warring camps of Trump supporters downvoting all left-wing posts and liberals downvoting all Trump posts. Gun vs. anti-gun. Flat-earth vs the world. One religion against the others. All religions against atheists. Etc. Etc. I actually come from Google Plus and Google deliberately chose not to have a downvote option to avoid this kind of downvote bullying. But they do support flagging, comment removal by post owner, and moderation within communities.

Possibly the voting-economics of Steemit could discourage that or at least make it less effective, but I don't know and wouldn't rely on it. Anyway, thanks again for bringing this up, it deserves some discussion and I'm sure this won't be the last time.

I think it is far too early in this experiment to create governance to protect people.

Start moderating the heck out of everything right now and you have a platform that has no real differenciation from any other except that it is built on a blockchain. Is that enough? Sure... Why not?

But I think we should look at how the concept of reputation as manifested in up- and down-votes can regulate these things and stay true to the experiment.

I have to notice that you did not give any concrete examples of charlatans. Worried about getting down-voted?

Let's let the experiment run on for a bit. There are spontaneously arising governance experiments here already.

If you want to call on some nebulous central authority to fix things... Well... Facebook exists already.

I tried to make it clear that I'm not calling for any one fix, just some change on what exists now, based on what I'm observing right now. I'm definitely not calling for heavy moderation or top-down governance. I'm simply suggesting that it might be a good idea to let flagging be actual flagging and to adopt an anonymous downvote button for the purposes of decentralized distributed moderation, or to at least make it clear that the flagging button is to officially be adopted as a downvote button.

Let's not make the slippery slope argument that one small change will lead to much larger changes.

Okay, I see.

On an unrelated note, I just took a glance at your other articles, and I see you are puttin gcontent out there on universal basic income.

I don't have a coherent position on this concept right now, but it is interesting. I will keep an eye out for the commentary you put out on this.

Best wishes!

What happened when you got into a flagging war with another user? As I was researching a post about the reputation calculation I found that any flagging war is certain to be won by the party that has the superior reputation. For example, if you flag this comment (please don't) it will have a negative effect in proportion to your voting power. If I tried to retaliate by flagging all of your active posts and it would have no effect on your reputation, but it would drain my voting power. If we kept it up long enough my reputation would drop below 25 and then I wouldn't have the ability to change anyone's reputation, up or down. The user with the higher reputation is guaranteed to win as long as the war is one-on-one.

If you find any claim in the post that I linked above, please let me know while I have time to fix it.

Upvoted for discussion, but I disagree. Communicating on Steem is like talking in public. If one of your neighbors is giving dangerous advice, would you call the police? No, you'd tell them they're wrong, and you might tell your children not to listen to them.

One thing Steemit or any other front-end could do is showing a warning when the health tag is used, like the medical disclaimer that Wikipedia displays in some languages. "Don't take medical advice from a random blogger, consult a doctor if you're ill."

Making retaliatory flagging impossible is an interesting proposal, but it would be complicated and would have unwanted consequences. We may need to downvote some vandal who is flagging others just to annoy them. As soon as there's any conflict, an anti-retaliation rule would lead to preemptive strikes: I flag you before you can flag me.

When you're an innocent victim of heavy flagging, try to recruit others to help you by posting about it.

In general, nobody's promising that a free market in ideas will be perfect. The whole point of a free market is that it will be better than a system that tries to avoid any risk.

Yes! I am new and have come upon health posts with interesting titles but contain complete rubbish about the benefits or danger of certain substances. Ie: 'coffee is bad for you.' What? That is so last century LOL! Since I am new, I have trepidation about making a contrarian response in the comments. Public flagging is definitely taboo in my universe so I just move on.

This is exactly my concern. You should be empowered to help protect the community from harm, and yet you feel a trepidation to do so because there is no official mechanism in place to do so.

Everything is becoming more and more complex. I guess it also depends on the article to be flagged. I think hate speech, violence and abominable acts type of posts should for sure be flagged. I personally wouldn't flag a post that just has a difference of opinion to what I like or would say. Not everyone has a degree in writing. Some people are happy to post pics of their kids, while others will write a treatise, so I am more lenient in that respect.

I totally agree. Flagging should not be seen as a form of opinion disagreement button. Unfortunately there's quite a bit of confusion out there between what's fact and what's opinion, so no matter what, people will downvote things they think are harmful that in fact aren't, purely because they are misinformed, such that their downvote is based on opinion instead of fact.

However, I'd personally prefer things getting downvoted that aren't harmful to people, versus things not getting downvoted that are harmful to people, because the option doesn't exist, or does exist but in a form people don't use or are afraid to use.

THANK YOU for this article! I wrote on these same points about antisemitism and "holocaust denial" I've been seeing on Steemit-> https://steemit.com/steemit/@thatgrlbeth/call-to-help-steemit-is-a-haven-for-holocaust-deniers-and-anti-semites-read-for-yourself
I'm also a huge believe in basic income and future automation as a means to a better life for all. I'll be following!!

Im glad there is no down vote! In my short time here, I see so much jealousy over what peoples posts are earning. That and retaliation could lead to things spiraling out of control quickly, imho. Flagging is a better option. Cream will naturally rise to the top. :)

I think downvoting would require anonymity, and if flagging is to function as downvoting, that anonymity is required there as well, so as to avoid retaliation spirals.

Also, the cream is definitely not rising to the top here. I've found absolute garbage doing quite well, despite being the equivalent of snake oil. That's what inspired this post to begin with, the reality of shit rising.

So are upvoting and flagging currently not anonymous? On Hacker News, they certainly are. And there's both downvoting (for posts that don't contribute) and flagging (for spam and other posts that violate community standards). That makes the most sense to me. I'm new on Steem, but I've adopted the policy of flagging all "upvote/follow" spam. Especially when it appears multiple times in a thread.

Flagging is not anonymous. Under the payout it shows your name with a - instead of a +. I got a troll who flagged all my articles because I flagged what was a copy paste video that was very sexist and racist. Besides the content it should've been flagged.

Damn, that is arguably a design defect. Because it facilitates retaliation, and flagging wars. I've only flagged blatant spam, but now I'll think twice.