Since if you want to endanger the consensus, you need enough stake, but if you have so much stake, you benefit from not endangering the consensus more than you would if you try to game the system.

Except for the fact that is completely not the behavior we see, not the behavior that they intended according to the white and blue papers, and makes no sense in any context, sure. Aside from the fact that your description is absolutely wrong, it's fine.

Of course, I never said that POS was to make it "one process one vote," and in fact I said the exact opposite. Because that is impossible, as anyone who has ever tried to implement any kind of voting algorithm in any context would tell you. POS was never designed to implement that kind of approach. All POS is designed to do, when it comes to voting, is attempt to map "one dollar one vote." It doesn't matter where that dollar lives, it doesn't matter how that dollar got there, all it matters is that the dollar is recorded as held and that it's voting power is directed.

You amplify that by actually failing to understand what "the consensus" is. "The consensus" has nothing to do with what votes register and is purely about whether or not the database represented by the blockchain has nodes that agree with each other about transactions. That's all that it means. That's all that it can mean. So it's impossible for a voting action to "endanger the consensus" except insofar as it does any kind of action to the database, which gets registered by a node and that data replicated and verified by other nodes. Voting does not endanger the consensus. Stake has nothing to do with endangering the consensus.

(In theory, a larger stake should mean that you have more interest in the platform as a whole gaining value in order to increase the value of that stake, with the long-term view of eventually cashing it out in order to make use of it. In practice – because the underlying value of the platform doesn't actually matter anymore, if it ever did, large stakeholders only have an interest in making the mechanical process of expanding the proportion of their stake numerically larger at a rate faster than the inevitable degeneration of the value of the underlying token. That's the behavior that we see.)

Without KYC if you make it one account one vote, you'd be so overrun by bots, that probably it would be again about "who has the most stake to run all these bots".

Congratulations, that is an accurate statement of the way that it is and has been for the last several years. Not because that things are "one account one vote" or even because pseudonymity is easy, but because a larger automated base of execution is a more efficient means of keeping SP active when voting power is strictly bound to the temporal domain. With one account holding all of your SP, you are limited in how many 100% votes you can drop. By distributing your SP across a larger number of active accounts, you see a net efficiency gain in leveraging that SP because the decay rate of using it is distributed.

We have bots because actually using the platform as a social media system that is interested in content is deeply inefficient because of the temporal constraints placed on "one account." Humans sleep. They engage with only one thing at a time. They take time to actually consume content. There is an active pressure (since the last hard fork) to put any votes in exactly 7 minutes after the content has been posted to the blockchain. When timing and numbers are more important than content, machines are better players of the game and humans – and thus we see vast numbers of bots active on the system. This is one of the reasons that my voting architecture specifically avoids timing issues and sequentiality, because machines are much better at dealing with them then humans are.

Now, I completely agree that Steem should NOT be about rewards, and I think the biggest problem Steem had until today is that it wants to attract people with rewards and not with communities. Most people are here for the money and especially previous to HF20 but also pre HF21 90% of all interaction on this blockchain was about extracting the maximum amount of reward out of here.

While this reduced, I think we're still far from there. Maybe the community feature is going to improve this significantly.

In order for the Steem blockchain to not be "all about rewards," it would require a fundamental architectural change – one which is entirely unsupported by the software and design which underpins the whole thing. Or, more succinctly, "you can't get there from here." And that's not a problem "until today." That's an essential and elemental problem at both a philosophical and mechanical level. You can't do it. Or at least you can't to do it and still be a thing which is vaguely similar to the Steem blockchain.

Currently, 90% of all interaction on this blockchain is about extracting the maximum value of reward out of the blockchain, whether that be with posts which are strictly cryptocultist in nature or mechanized interactions between bots acting solely to keep SP active and thus direct the funds in the inflationary pool too high SP holders. SP which is not being used as part of a vote is effectively dead weight, increasing the rate of inflationary value decay without providing any sort of benefit to the holder. This is part of what delegation was supposed to improve, by removing an aggressive pressure on high SP holders to use all of their SP for automated voting mechanisms and instead delegate their SP to other users who were more active and could more efficiently use that SP for voting thanks to the aforementioned temporal restrictions. In practice – delegation just made bots warms that much more efficient than actual human beings because you don't actually have to transfer funds to another account and accept the risk that they will be used in a way you did not intend, you can delegate those funds as a sort of zero interest loan which can be revoked at any time. Zero risk and low friction means that it's much better to give it to a machine that it is a person.

While this reduced, I think we're still far from there. Maybe the community feature is going to improve this significantly.

I wouldn't bet on it. Look at what little documentation we've seen as regards to how the Community system is going to come into being. You are going to have to pay STEEM to create the group, you're going to have to pay STEEM on a per seat basis, and the Community is going to require an ongoing SP holding it order to cover its RC costs. Steem blockchain-based Communities are effectively a big economic push to try and drive a purchase of STEEM for the creation of SMTs under the magic argument that a Community-based SMT will allow you to earn enough Steem to keep the Community afloat.

Yes, I am saying that even more than the usual cryptocommodity pyramid scheme, Communities are doubling down on that.

That is not going to be a healthy aspect of the platform.

I don't think that downvotes always have to be bad, Stackoverflow has ways to give up and downvotes and I think it is an important feature.
Maybe some people were able to solve a problem in a certain way, but actually that way is not good for you, there has to be a way to push this answer below a "more valid" one.

A downvote system is inevitably an inefficient means of conveying your will. That is definitely true on Stack Overflow, slightly ameliorated by the fact that they do have a one account, one vote, all votes are equal architecture. However, in pretty much every case, you would be far better served by voting up the answer that you think is better than you will be served by voting down the worst answer. This leaves aside all of the other massive differences between the use and consumption habits of a content creation platform and a question/answer platform, which are vast and important, but all that aside – voting for things that you consider good is a much more efficient use of your voting power than voting against the things that you consider bad, because the latter doesn't ensure that you get more things that you want nor does it ensure that you will get fewer things that you don't want. Voting up good answers ensures that you are more likely to get things that you want. That is, answers which are accurate and at the top of the list when you search for them.

There is a way to push bad answers below better answers – vote for the better answers. If voting for the better answers is insufficient, you have a problem with the answers that you're getting and not with the quality of the answers.

Back to Steem, now while Steem should OF COURSE not be about the economic incentives, it should be about content, about communities, about fun, etc, but it should be also a part of Steem in the future as well. Making sure that the user gets a bit back of the platform he puts his data in is just fair.

What does that even mean? Seriously, what does it literally even mean?

If your argument is that the user deserves to get financial incentive for using the platform for his social media engagement, I would point out – again – that being obsessed with financial incentives is part of the massive failure architecture that we have seen lead to the current state of the Steem blockchain. The financial incentives architecture is exactly what is failing. You get what you reward. The value of STEEM is the value of the content on the blockchain. Or, more properly, the emotional reward of engaging with content on the blockchain, whether that be through posting it, reading it, or commenting on it.

No amount of obsession about the fiscal side of things is going to change the fact that until and unless the focus is on building a social media platform that people want to engage with, that ratifies them for their engagement, that provides them value in their standard usage case, there won't be any value to infuse the token with. It literally is just a way to keep score.

And my proposal is one of the ways I would solve this. Aligned with the fact that we have a PoS system and aligned with that doing it with KYC is not in the spirit of this platform (besides the possibility of gaming KYC too).

The issue is that it doesn't solve any of the problem. All it does is increase complexity, increase the already overwhelming vertical cliff of the user interface, inject a place for people to find more emotional dissatisfaction from their engagement, and do nothing about the underlying dysfunction which is why we see the crumbling infrastructure we do. We do have a POS system, and the fallout of that is not just occasionally problematic. KYC is impossible. Not just difficult, impossible, and the fact that money laundering is still a real crime that happens every day with fiat currencies that have thousands of years of people banging on those mechanisms to try and rub out the rough spots is proof that it's impossible. (I would not be hesitant to make the argument that the impossibility of KYC in a fully digital, even loosely anonymized token is one of its selling points.)

Your problem seems to be that the issue you think is most pressing is "the application of my voting power is inefficient," and to make it more efficient you want to signal to the system what you think the price/value of a given post should be, and then leave it to the system to try and invest/disinvest your SP/voting power to reach that point. But that doesn't help you get what you want. It is largely functionally identical to a straight up or down vote because unless you have truly magnificent amounts of SP, you will never be able to influence the results significantly toward or away from a given value. Never. And if you have that amount of SP, you won't care about a specific post making a specific amount to little or too much, the only thing you really care about is whether you think it should be more or less than what it currently is rated at having made, because your influence is best spent on upvoting your own content or that of a member of your bot squad, which gives you SP directly or by proxy.

What you want as described provides no benefit to you or anyone else, while introducing significant impairment in understanding, UI design, and likelihood of getting people interested. Rather than drawing people's attention away from an obsession with financial numbers, it doubles down on putting every single voting interaction front and center, concerned with "how much money do I think this content is worth?" Every interaction becomes about controlling the value of something else and not about engagement with the content and an allocation of resources that are yours.

That's a bad design. It's a bad design not just because it won't work but because it actively tries to undermine everything that you've described is what you want.