Hive is a decentralized public forum. And the "costs" for supporting it are spread across multiple organizations: 1) witnesses nodes who operate blockchain software, 2) api nodes that run blockchain software + other software that requires more hardware resources, and 3) front end website operators (peak.com, ecency, hive.blog, actifit, etc, etc).
It is relatively difficult for witnesses to censor blockchain information (it requires 75% of the witnesses to agree to the censorship). On steem, before the takeover, this happened only once, when witnesses censored transactions from the steemit account to prevent it from taking over the chain.
After the takeover, however, censorship became quite easy there, because two entities (steemit+some Korean organization) gained enough combined stake to control the entire set of witnesses. After this, massive censorship resulted. So this is the primary censorship weakness of Hive: if someone gains enough stake to control all witnesses. But even in this case, as happened with Steem->Hive, the code can be forked to avoid the censorship.
There are fewer API nodes than witnesses (maybe around 10 nodes), and API nodes can also censor, but this form of censorship is even more limited than what is possible by witnesses, because it only takes one API node that is refusing to censor information in order for the information to be easily available to all Hive users.
Also, we've dramatically lowered the costs of running an API server nowadays, and there is no election requirement to run an API node, so any information freedom advocate can decide to run a Hive API node if they see censorship at this level.
Similar logic applies to the frontends: it's basically the same situation as for API nodes (and most frontends also run their own API node).
From my point of view, the above are the only points available for actual censorship of information on Hive.
However, there is a related issue, which I believe is the one you're concerned about: "downvoting".
Downvoting does not actively prevent information from being read. You can still read any post, no matter how much it is downvoted, on every Hive frontend. On some frontends like hive.blog, you may have to go thru an extra step of clicking on a button that says you want to read it (but on ecency for example, there is no such need). The "hiding" isn't a blockchain level thing, it's just an optional thing done by the frontend software. But to my knowledge, there is no Hive frontend that doesn't leave you with an easy option for still reading the post in such cases.
But Hive is more than just a public forum. It also has an "incentive" system to encourage information sources to post information on Hive. And downvoting will potentially dis-incentivize some posters from posting, especially if they are primarily posting for the reward and not just because they have some reason to share the information. So, for information sources where the reward is a primary motivator, voting both positive and negative can affect what information gets shared. But I don't consider this censorship. This would result in some censorship if we lived in a world where people's primary income source was income from posting articles on Hive, but I don't think we live in that world.
To get down to specific cases, I don't think that whoever is posting RT articles (I don't really know if it is RT or not) is particularly motivated by the rewards, and so they keep posting despite the downvotes. If they stop, I think it will just be because they don't think it is worth their time, not because of getting posting rewards or not.
Now, there is one argument still left that could be related to downvoting as a form of censorship. Another way to view Hive is as many "public forums". In this view, every community is a public forum, trending is a public forum, hot is a public forum, and everyone's "feed" is yet another public forum (this is a really cool feature is it means literally anyone can easily establish their own public forum).
Downvoting a post lowers posts in the rankings of many of these forums (it doesn't affect feeds, however). So a downvoted post is not likely to be read in the trending forum.
But I view the trending forum as an "overbooked" public forum. There's not enough " attention time" to allow every "speaker"/poster to have his equal say on the trending forum. So somehow such speakers have to be prioritized. This is an unavoidable problem: it is a hard limitation on human attention.
Upvoting/downvoting is the mechanism that Hive uses to make this determination. Not everyone gets the chance to "speak" on the trending forum whenever they want, and unpopular posters may never do so, but Hive still supplies plenty of other public forums for those posters, and it isn't difficult for people who want to listen to them to find them.
First of all: I am glad you are taking that time; as mentioned before, communication has been tried repeatedly by me and many others, and if there were answers they were often pretty offhand, usually not answering the question, and if so then often with arguments that on closer inspection did not hold true. So yeah, thanks!
On the topic: I am not concerned about downvotes themselves; I actually think it is a great concept, it's only that people coming from a world full of likes and void of criticism might need to learn how to handle it.
I am concerned about the way power is being wielded by a few, and having quite an impact on the many.
In my opinion (and most probably in the opinions of quite some critics of your downvotes) censorship is suppression of information by exercising power; manipulating the flow of information to create a 'narrative' that is more in line with a certain taste, a bias or some agenda. Call it censorship or not, but that's what most of this is about. In Hive's case this is done by a few who 'staked' their way to power, and they do have a proportionally huge impact. This is not about community-based guidelines or some consensus, this is not your private feed, this are heavy handed actions by you and a few others. You kind of rule Hive, and I do like a good ruler. But some activities really surprise me, as they do not make sense to me.
You are right, RT hasn't been 'censored' yet (only suppressed to some extend), but the effort is obvious. It was quite telling when a few months ago that account was bombed into oblivion from a reputation of 70+ by what looked much like a concerted effort or at least an interesting dynamic. Whoever posts from that account is astoundingly resiliant; maybe they aren't in it for the rewards, nor the fun, but to simply share/spread information. I don't care much about RT, but it's an ok addition to the spectrum.
Many others didn't take too well to being harassed with out of proportion dvs and left, and among them were world reknown artists and journalists. If you would have appreciated Hive (like I did) as an 'alternative' platform that also provides art/culture/information... from the other side of the coin, you would have noticed that as quite a loss of potential, esp. provided that wider adoption is something being aimed at. Which I don't see, and is one of the things not making sense to me.
And in light of that I also ask myself: To what extend 'de-centralized' is still the appropriate adjective for Hive? I see it on the blockchain-level being a distributed ledger, but to what extend is it true for the distribution of influence? Obviously zero. So maybe that slogan should be changed? Actually it would suffice if with power came responsibility ;)
Curious to read your thoughts on that
thanks
k
I wasn't responsible for designing Hive's "influence based on staking power", so I don't particularly feel compelled to defend it. It has some obvious weaknesses, but in the end I'm quite certain that Hive is still a much more decentralized solution than existing social media. It's not a trivial thing to fix the influence, because this design decision is baked into Hive at the core.
It is my plan to create a solution to the problems caused by wealth-based influence, but my plan for that will likely take a couple of years in the best case. I mentioned some details about this plan a long time ago in two posts and the ideas behind it have been under development for almost two years now. But even the design isn't yet finished, and we probably won't start coding of the actual software until next year.
As to my voting down of some high-reputation accounts, I do it to fix what I believe is a problem with the existing reputation system: it doesn't allow lower reputation accounts to lower the reputation of higher ones. So in the case of a few accounts that got high reputations from repeated voting by a few high staked accounts in the distant past, those reputations can't be lowered by anyone except an account with a very high reputation. For this reason, sometimes people contact me and ask me to help them downvote temporarily. This doesn't happen often, it's happened in two cases to my recollection: RT and world-travel-pro. There may have been one or two others in the past, but if so I can't recall.
I continue to downvote RT because I truly view them as a source of disinformation and I don't mind it being known.
To answer your final question, Hive is decentralized as a source of information.
But as to distribution of influence, its core design is staked-based influence, not democratic-based influence. This doesn't mean the system isn't decentralized, but it isn't particularly democratic. Nonetheless, anyone can continue to post whatever they want here, despite the opposition of "whale" token holders. The same cannot be said of centralized social media platforms.
Stake-based influence was probably an easy choice for Hive's design, because it matched the goal of the original coders to create a token that gave people influence by purchasing it, giving the token a tangible source of value, and equally important from a design perspective, it doesn't require the software to distinguish who are real people and who are bots.
Good to read that there is some plan to reduce influence of staked power; as the implementation naturally will take a long time, it would be even better if you already followed the idea with your personal DV-policy. I'll say it once (again) and then leave you to it:
(personally I can't fathom why you would do that in case of Hivewatchers).
If the few in power would respect those two points, Hive might be a much nicer place.
So the general drift I got out of this conversation:
You obviously play the long game, no quick 'mass adoption' or significantly expanding of the userbase. You rather take the time that it takes to work on a more fair and solid structure. If that's what it is, I am all for it, and even though I am kind of fading out of Hive, this makes me curious enough to check in again in a few years.
Thanks again for taking the time and shedding some light into that corner of Hive. I might direct some of the critical crowd to this comments as I think it quite a rare gem!